search this blog

Monday, January 7, 2019

PIE Urheimat poll: two or three options left

If we let ancient DNA dictate the terms in the Proto-Indo-European (PIE) homeland debate ahead of historical linguistics and archeology, then, as far as I can see, there are two or three realistic options for the location of the said homeland. Here they are, in order of my own preference:

1) The Don-Caspian steppe around 4,300 BCE (see here). The ancestors of the Hittites and other Anatolian speakers also came from this homeland and entered Anatolia via the Balkans (or, less likely, the Caucasus) in fairly small groups sometime between 4,000 and 2,000 BCE. A lot of samples from Bronze Age Anatolia are needed to confirm or debunk the presence of steppe ancestry there.

2) The eastern Balkans during the peak of the ostentatious Copper Age in the region. Proto-Indo-European developed in the wealthy Chalcolithic communities of the western Black Sea coast and quickly spread both into the steppes and Anatolia via elite and trade contacts, and thus with minimal gene flow. Proto-Indo-European minus Anatolian, or PNIE, then spread from Eastern Europe during the Bronze Age with the mass migrations of the Yamnaya and closely related populations. A lot of samples from Chalcolithic western Anatolia are needed to confirm or debunk that people moved from the Balkans into Anatolia at this time.

3) Transcaucasia and/or nearby around 10,000 BCE. Proto-Indo-European, or rather Indo-Hittite, is much older than generally accepted, and came from the Epipaleolithic northern Near East. It was introduced into the steppes by foragers of the so called Caucasus Hunter-Gatherer (CHG) type, where it eventually became Proto-Indo-European minus Anatolian, or PNIE. Proto-Anatolian was spoken by closely related CHG-like foragers who stayed in the northern Near East.

Admittedly, that last theory is way out there, and at the moment, has about as much chance of being accepted by most historical linguists as Out-of-India. But the one advantage that is has over the other two proposals is that it doesn't need any additional sampling of ancient DNA.

I'll probably get grilled in the comments as to why I didn't include a proposal with the Maykop culture as the PIE community, or at least the Indo-Europeanizing agent in the steppe. Honestly, after seeing the ancient DNA from a wide range of Maykop remains courtesy of Wang et al., I think the chances that Maykop was an Indo-European-speaking culture are low. Indeed, both the Maykop genome-wide data and uniparental markers scream "Northwest Caucasian" to me.

Also, if the Caucasus was the PIE homeland, or even a major expansion point for early Indo-European languages, then considering its widely accepted status as a linguistic hotspot and refuge, it's fair to expect that it should still harbor at least one highly diverged Indo-European language. Is there any evidence that it ever did?

Below is an interactive poll. Please vote for one of the three options and feel free to let us know in the comments why you made the choice that you did. I might add more options to the poll if compelling reasons are given in the comments to do so.

PIE Urheimat poll
Don-Caspian steppe around 4,300 BCE
Chalcolithic eastern Balkans
Epipaleolithic northern Near East
Created with PollMaker
See also...

Yamnaya: home-grown

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

On the doorstep of India


«Oldest   ‹Older   201 – 231 of 231
Davidski said...


Any signs of such samples surfacing on the horizon?

I haven't been keeping my finger on the pulse lately in regards to leaks and rumors.

Dragos said...

@ Synome
Well I'm not entirely confident in their ascriptions, which is why i checked, and it seems they are not either, because there is missing detail (exactly who is what). A lot of the individuals come from multiphase burial sites. Some of the ascriptions are not properly elucidated. Others are from the BB period (2600-2000), but have no actual BB goods. My view is formed on the basis of my analysis of individuals by site, goods, date, autosomes and uniparentals.
But maybe the next couple of papers will elucidate it further.
All I can say, at this stage, is something is fishy. Farming spread by migrations, Yamnaya has a distinct signature, CWC spread by migration, but BB was a ''spread of ideas'' ? Hhmmm.

Ric Hern said...

Maybe the emphasis on Pottery Style is too much. In modern times we rarely compare our everyday use kitchen items. A functional cup or plate is functional. The only time something like this catch the eye is when it is made of something out of the ordinary eg. Gold rimmed or intricately painted. This is why I think there were so many regional types because they were not really special tradeable goods like Vases etc.

So maybe concentrating on more widespread extraordinary trading goods would be a better indicator of what really bound the Bell Beaker people together. eg. Metals..?

Bob Floy said...


"Farming spread by migrations, Yamnaya has a distinct signature, CWC spread by migration, but BB was a ''spread of ideas''?"

It happens. It's certainly happened with most of the major religions.

Dragos said...

Hey did anyone notice that I5117 and I5118 are from Kurgan burials in Hungary (33-3000 BC) ?
They're both G2a2. Isotopic analysis from an different study showed that earlier Kurgan burials had isotopic ''local'' signatures, whilst later ones seemed distant.
Seems like further evidence for Balkan origins of the Kurgan culture.

epoch said...


"Corded Ware Single Grave samples from The Netherlands"

Most of the Corded Ware burials in the Netherlands are from sandy soils and are completely dissolved due to acidity. There is only one male skeleton found, found really close near the the Tuithoorn where the Olalde Dutch samples are from.

The point is that this man is linked with CWC based on pottery and dates from the associated village that was uncovered. His burial was completely unique and totally atypical for CWC: Apparently he was laid on a wooden structure in the open as there are clear signs his legs were chewed on by canines.
(from page 175 on)

There is another female skeleton, and that's it from the Netherlands. No other piece of bone found as they were all buried is sands..

epoch said...

There is a paper on the late Neolithic transition in Switzerland in the make and I recall seeing some pictures from a presentation showing it has some Corded Ware samples. Anyone know when that one is due? Any rumours about it?

Samuel Andrews said...


Please get one of the Bell Beaker enthusiasts to edit the "Genetic studies" section of Beaker culture wikipedia page.

Dragos said...

@ Davidski
That's true, however even earlier mounds, kurgans & monuments are found near Moldova and Romania :)
My point was mostly related to local successions, and slightly tongue-in-cheek', but those R1b-V1636 arent very common anwhere.

I'm pretty sure more Hungarian & Balkan Kurgan genomes will appear this year.

Grey said...

Samuel Andrews
"Beaker R1b P312 tribes were large in number & loud. There were tons of them. They overwhelmed western Europe. This makes it hard to imagine their ancestors were quite & small in number when living in eastern Europe."

mu understanding is that at least some of the time BB lived as a separate population within pre-existing farmer settlements which to me implies they weren't competing over the same resources - a couple of possibilities might be:

1) different land i.e. BB using land the farmers didn't (e.g. pasture land adjacent to rivers too wet for crop farming but ideal for cattle?)(maybe similar to an Akkad-Sumer herder-farmer symbiosis?)

2) some kind of artisanal-trader type population like African blacksmith clans i.e. a third option to add to "pots not people" versus "people not pots" - "potters not pots."

although that would need an explanation for how they later came to dominate genetically (animal diseases?)

Grey said...

Bob Floy

"It's certainly happened with most of the major religions."

right but before mass communication mostly via (small numbers of) people: traders, soldiers, missionaries etc.

epoch said...


"Hey did anyone notice that I5117 and I5118 are from Kurgan burials in Hungary (33-3000 BC)? They're both G2a2. Isotopic analysis from an different study showed that earlier Kurgan burials had isotopic "local" signatures, whilst later ones seemed distant."

Where are they published? Can't find them in Olalde's BB paper or Wang's Caucasus preprint.

epoch said...


Never mind, they are in Olalde et al. Just not in the XL sheet.

Davidski said...


Beaker_The_Netherlands I5748, dated to 2579–2233 calBCE, is one of the oldest C14 dated Beakers in the Olalde et al. dataset.

This individual also belongs to R1b-P312. So he's a Beaker alright.

Is there any evidence that this sample might represent a totally new population in the Tui region that eventually replaced the late Corded Ware Single Grave people there?

If not, it looks like the so called eastern Beakers might just be an offshoot of the Single Grave culture from The Netherlands and surrounds, that then expanded back east, as far as Hungary, where they mixed with the Hungarian Yamnaya.

Here's a fairly tight qpAdm model for the most Yamnaya-like Beaker in the Olalde et al. dataset, who also just happens to belong to R1b-Z2103. Basically half Beaker/half Yamnaya.

Beaker_The_Netherlands 0.576±0.062
Yamnaya_Samara 0.424±0.062
chisq 11.469
tail prob 0.489238
Full output

I would suggest that while we wait for the Hungarian Yamnaya samples (which might come back R1b-Z2103 and I2 anyway) all of you guys who are really interested in the near and far origins of the eastern Beakers should start looking in detail at how the Single Grave culture came to be exactly.

Dopa said...


Then where are the Battle-Axes in Beaker graves?

Davidski said...


Then where are the Battle-Axes in Beaker graves?

I don't know? But keep in mind that I was talking about Single Grave Cored Ware not Battle-Axe Corded Ware.

So, did the Late Vlaardingen Single Grave groups from West Frisia use Battle-Axes as grave goods?

epoch said...


"Is there any evidence that this sample might represent a totally new population in the Tui region that eventually replaced the late Corded Ware Single Grave people there?"

Not that I know off. There aren't much finds from the area though. What I do know is that All Over Corded and All Over Ornamented beakers not only popped up in noted beaker sites but also in Vlaardingen sites. Olalde et al co-author Fokkens, while critical of the Dutch origin hypthesis of Lanting, still thinks AOO and AOC are typologically evolved from Protruding Foot Beaker, the local variation of Dutch Corded Ware beakers.

To give an impression of the local geology back then: The area nowadays called North Holland was split in a northern and southern part by a large estuary, with tidal meadows heavily under influence of the sea. Alongside of that area are higher sandy ridges where the BB and CWC settlements and barrows were found. They probably had cattle grazing on the tidal meadows.

South of that were large bogs and fens separated from the sea by dune areas. That's part of the area of the Vlaardingen culture, which focused on fishing and hunting, keeping pigs and agriculture on small plots of land on the strip of dunes. Their local pottery was downright dull, nothing like the prestigious beakers. The AOO/AOC pots do show up in their settlement, by the way.

The only hints of migrations are the Elp culture, which buried in supine position and had Kummerware pottery - "Miserable Ware" - that in my opinion may be tied to Vlaardingen pottery.

This baffles me the most. During the whole Beaker Era pottery was highly appreciated, apparently. Then at the onset of the Bronze Age people completely lost interest in them. Their pottery looks just like dull kitchenware.

Synome said...

I think the western CW fringe is a plausible origin of the Eastern Beakers. I'll do some more reading on it.

However, a CW origin of all steppe bearing Beakers is harder to square with the linguistic evidence. Beaker derived Unetice, Urnfield etc. are typically thought of as the ancestral cultures for Italo-Celtic speakers.

It's much tidier to have CW ancestral to Balto-Slavic, Germanic, Indo-Aryan based on shared features and divergence time estimates, and the same goes for associating an earlier branch off of Yamnaya with Italo-Celtic, which conveniently is found not far from where those languages were historically spoken in the Hungarian Yamnaya.

If some branch of CW is ancestral to all of the steppe bearing Beaker derived cultures of central Europe, we would have to explain the discontinuity between Italo-Celtic and the other CW associated languages.

epoch said...

@Dopa & Davidski

Yes, late Dutch Single Grave culture buried with battle axes. So yes, there is an issue to watch there.

Andrzejewski said...

@Synome "It's much tidier to have CW ancestral to Balto-Slavic, Germanic..."

Germanic is a mixed by-and-large Centum language from Bell Beaker R1b elements with Satem Corded Ware R1a1 elements, whereas a large proportion of up to 1/3 of its vocabulary derives from either the SHG/WHG lexome of the Ertebolle Culture and/or the farmer vocabulary of the LBK.

epoch said...


"I'd like to know stuff like population density, did Beaker folk live in isolated niches or did they dominate regions, nature of trade & immigration, how & why did Beaker tribes so often migrate and settle in new land."

I think every archaeologist would like to know that as well. We don't know much about the past. Which exactly is why ancient DNA is such a thrilling opportunity.

Matt Thomas said...

PIE is from the Don-Caspian Steppe and Pre-PIE is from Eastern side of the Caspian Sea, most likely modern day Kazhakhstan.

It is pretty evident, if you look at the ancestry of a modern day Norwegian or British Individual, they have higher Baloch/Gedrosian ancestry. If the expansion was from the South of Caucusus, the Baloch/Gedrosian type ancestry would not be present at high levels. The same is evident in Yamnaya individuals with a gradient from Caspian Steppe to Central Asia with higher Baloch type ancestry as you get closer to central asia and higher NE Euro ancestry as you get closer to Eastern Caspian Steppe.

Maciamo Hay from Eupedia has always been correct about all his predictions.

Davidski said...

@Matt Thomas

Reality check...

The excess Baloch/Gedrosian ancestry in Northwestern Europeans is an artifact of the analysis run by Dienekes Pontikos with the ADMIXTURE software.

The output from the analysis was wrongly interpreted by Dienekes and many others, because of their lack of understanding that ADMIXTURE runs were not formal mixture analyses.

The reason that the excess Baloch/Gedrosian ancestry appears in Northwestern Europeans is because there is no Northwestern European-specific cluster in this particular ADMIXTURE analysis run by Dienekes. The only Northern European cluster in the analysis is largely based on the allele frequencies in Northeastern Europeans, who generally have more European hunter-gatherer ancestry than Northwestern Europeans.

So to account for this and improve the statistical fit for Northwestern Europeans, the ADMIXTURE algorithm gives Northwestern Europeans a few more per cent of the Baloch/Gedrosian ancestry.

Also, Pre-PIE obviously didn't come from the Eastern side of the Caspian Sea; not from Kazakhstan nor anywhere near Balochistan/Gedrosia.

During the Chalcolithic, what is now Kazakhstan was home to the Botai and similar peoples who didn't contribute any significant ancestry, or probably no ancestry at all, to the Eneolithic and Early Bronze Age peoples of Eastern Europe, like the Yamnaya pastoralists. In fact, the Botai people, and even their horses, were mostly forced out of the Kazakh steppe by Yamnaya-related migrants from Eastern Europe. See here...

Major horse paper coming soon

Moreover, Maciamo's theory has always been, and, as far as I know, still is that the Yamnaya population formed as a result of the migration of pre-Yamnaya peoples rich in R1b from south of the Caucasus to the Eastern European steppes. This is false, and you can read about why it's false here...

Genetic borders are usually linguistic borders too

Big deal of 2018: Yamnaya not related to Maykop

Mem said...

But maybe CHG ancestry of Yamna came from kelteminar area from central-south kazakhstan.Botai was populate the west siberian steppe,not all east caspian lands.
Also,as I know,the genetic structure of kelteminar is still not analyze.

Davidski said...

Kelteminar was like Botai but with some Iran_N admixture.

There are samples like that already available from Gonur. They're obviously migrants from the north and might even be from Kelteminar.

Nothing to do with Yamnaya.

Halfalp said...

I dont really understand that fascination to link a Language with an Ancestry instead of a Lineage. Ancestry can be biased by sample bias and timeframe. Typical exemple: Sidelkino is the oldest EHG individual that we have and he shows CHG, then the Samara individual is a little younger and shows no CHG ancestry, then people are writing a story that CHG ancestry might have expand in a later time frame like in the times of Khvalynsk because it's then that the CHG ancestry start to pop a lot. I guess i follow the idea of Davidski that CHG might have been in Steppe Maykop way earlier than let's saying, Neolithic Caucasus. Also, i would not myself give any exeption for a PIE urheimat till, some challenger suspects that some people want credit for ( Shulaveri-Shomu / Leyla-Tepe ) are eliminated just like Maykop and KA. Also, if the origins of Bell Beakers R1b's and therefore all western IE languages are linked somehow with CWC or a CWC outlier and that PBS and PII languages are linked with CWC through Sintashta, what is Yamnaya even relevent for at this point? Tocharians, Anatolians?

Davidski said...


I don't want any discussions about Aryans on this blog unless this is firmly done in the context of linguistics.

Arza said...

@ Davidski


chisq: 8.271 tail: 0.763634757

best coefficients: 0.624 0.376
std. errors: 0.059 0.059



chisq: 6.785 tail: 0.816205595

best coefficients: 0.145 0.480 0.375
std. errors: 0.118 0.129 0.055

So Varna outlier shows not only the steppe ancestry, but also ethnic-specific Balto-Slavic drift. In 4711-4542 calBCE.

BTW don't high std. errors suggest that qpAdm has troubles with clear differentiation between Baltic_BA and Sredny_Stog ergo the non-WHG part of Baltic_BA comes from Sredny_Stog?

Arza said...

Baltic_BA - 14.5% (2.7-26.3%)

Global 25:
Ukraine_Eneolithic:I6561 42.4%
Trypillia 35%
Baltic_BA 22.6%

Distance 3.8771%

El Lurker said...

In Paraguay and Brazil (a case of 100% male conquest) the conquerors adopted the language of the tupi-guarani natives because the Iberians engaged in polygamy with the native women and the kids learned the language of their mothers.
Paraguay is still 100% Spanish Guarani bilingual and Brazil was Tupi speaking for 2 centuries.

epoch said...


There are actually, if I gathered the right info, only two skeletal remains available from the Netherlands. A woman and a man buried in a very, very non-typical way for SGC, but tightly associated with a settlement which is clearly Corded Ware. The rest of the Dutch CWC burials, again if I'm right, are all dissolved to corpse silhouettes, which do give info on the burial position.

«Oldest ‹Older   201 – 231 of 231   Newer› Newest»