Proto-Indo-European comes from Eastern Europe

Why "Out of Armenia" theories don't hold up to razors, and Steppe Urheimat does.

In August of 2022, the largest and most controversial archaeogenetics study in a while came out, probably the largest since the initial floodgates of Indo-European debate broke in 2015. In this study, peppered with prominent names in genetics like David Reich and Iosif Lazaridis, the Steppe Urheimat seemingly cemented in 2015 was rejected, for a new “Southern Arc” hypothesis. In this hypothesis, Indo-European languages actually began somewhere around Armenia, spread northwards into the Yamnaya Culture through a peopled migration, and spread west into Anatolia. In this hypothesis, all Indo-European languages except Anatolian come from the Steppe. However, in more recent papers there have been much more arrogant in their claims, suggesting that Indo-European actually spread thousands of years before most linguists agreed, and that the old Anatolian Hypothesis is actually half-correct all along. These claims are not supported by the writers of the Southern Arc hypothesis, in fact the two are not compatible, and I think these claims are much more out-there than those made in the Southern Arc. So, for now, I’m just going to focus on my contra points for the Southern Arc Hypothesis.

The Southern Arc hypothesis rests primarily on two findings. First, that Anatolians experienced a large influx of DNA from the Caucasus, but not from the Steppe, around the early Bronze Age (around the time Anatolian is believed to have branched off). Secondly, that Yamnaya allegedly have some sort of particular Anatolian-Levantine type DNA in very small amounts that Eneolithic steppe cultures didn’t have, which implies a newer migration. First of all, the latter is probably not true. First of all, the authors insist that the Anatolian component in Yamnaya cannot be a product of local Farmers, because the farmers had excess Hunter Gatherer. This is strange because plenty of other studies modeled the Yamnaya just fine using Eneolithic Steppe and local Farmers. Anyways, this article I read a while back does a pretty great job debunking the idea that Eneolithic Steppe had a different southerly component than Yamnaya, and showing that Yamnaya can be modeled entirely using Eneolithic Steppe, Neolithic Ukraine, and Trypillian farmers. Some of you probably have zero experience with ADMIXTOOLS, so it might be a bit confusing of a read, but ADMIXTOOLS is formal software which is used in studies all the time, so don’t be turned off by its heavy use. Davidski also has a lot of articles attacking this claim about Yamnaya, if you want to comb through them. In fact, he’s written so many articles about this subject that most of my arguments here are reflected in his articles. But in my opinion the Southern Arc study used a very bizarre means of admixture modelling to try and model Yamnaya as something different from what it had always been modeled as. The insistence of such thing as “Anatolian-Levantine” ancestry was so ridiculous. Just because there’s a cline there doesn’t mean EEF and Levantine are at all worthy of being lumped in. WHG and EHG had something of a Cline. Now should we talk about “WHG-EHG ancestry”? And the study is relying on around 1% each Anatolian and Levantine for this, which is first of all very low, and secondly doesn’t align with prior studies which DO find extra WHG in Yamnaya. I see no reason to trust the ridiculously roundabout model Reich’s team is using, especially for such trace ancestry. Again, see the linked posts against the nitty-gritty of Yamnaya modeling. Even if, even if! Yamnaya DID actually have some sort of excess DNA from south of the Caucasus which cannot be explained by extra DNA from Eneolithic Steppe and European Farmers, it’s completely explainable by the Yamnaya being an extremely mobile people who engaged in long distance trade. Another explanation could be that Yamnaya did not mix with the WHG-rich farmers directly to the west of them, but with the almost entirely EEF cultures of the south like the Vinca culture.

Anyways, for this to be true rests on Sredny Stog and other previous Steppe cultures having a rather different profile from Yamnaya, which I am not confident in. We have some samples from the periphery of Sredny Stog which are rather genetically similar to Yamnaya, just with some extra ancestry from Neolithic Ukrainians. I have seen leaked samples of Sredny Stog further east from a preprint, which are quite typical of Yamnaya from Ukraine in terms of their ancestry. At least on G25.

The distances aren’t great, but I don’t think this is due to later Yamnaya having especially high CHG/Iran/etc. Afanasievo is basically identical to Yamnaya, the Sintashta one is an outlier. Poltavka is also basically Yamnaya.

Okay, but we’ll get back to the Steppe in a bit. the main crux of the argument is that Anatolians do not have Steppe ancestry, while all other Indo-European cultures have at least a bit in their initial stages. On the contrary, Anatolians did experience a large influx of CHG ancestry. But, there’s a few issues with this. First of all, one of the big things which bothered me first reading the paper, is that they didn’t actually attempt to measure Steppe ancestry directly in Anatolians, or anyone. They only used EHG (Eastern Hunter-Gatherer) ancestry as a proxy. Yamnaya were only around half EHG, but there is enough drift between Yamnaya and their Hunter-Gatherer components that there’s no reason not to do direct models involving Yamnaya in a formal study.

I’m quite skeptical that, even of our current samples from Anatolia, *no* Steppe ancestry is present. I’ve seen some evidence that some of them have trace ancestry. Either way, the issue with Anatolia is that Bronze Age western Anatolia is very poorly sampled compared to central and eastern Anatolia. There are no northwestern Anatolian samples from the Bronze Age, and only a few from the west in general. I will explain the importance of this later, but the more important issue here is assuming that the Indo-European migration into Anatolia would be something like later Indo-European migrations. The Indo-Europeans were a much less advanced people at this point in history. It’s unlikely that they had the wheel at this point, as Anatolian languages do not share the variety of wheel-related words that are cognate in other Indo-European languages. Wheel, axle, wagon, that sort of thing. The invention or adoption of the wheeled vehicle was one of the main reasons the Indo-Europeans were able to expand with such vigor in later cultures, and in significant numbers. Another thing to keep in mind is that Anatolia at this time was among the most densely populated and sophisticated parts of the world. It is perfectly plausible to assume that the Indo-Europeans, who were already admixed with locals in the Balkans, would have assumed a position of authority more akin to that of the Mongols or Magyars in Europe, leaving little to no impact. The study itself makes it very clear that such things continue to happen in Anatolian history. The invasion and periodic political hegemony of Phrygians, Scythians, Cimmerians, Galatians, Greeks, Persians only raises the amount of EHG-related ancestry in classical Anatolia very marginally.

“However, in individuals from Gordion, a Central Anatolian city that was under the control of Hittites before becoming the Phrygian capital and then coming under the control of Persian and Hellenistic rulers, the proportion of Eastern hunter-gatherer ancestry is only ~2%, a tiny fraction for a region controlled by at least four different Indo-European–speaking groups.”

However, when you analyze this group using ancient samples from Macedonia (not necessarily Macedonians. This was its own point of contention when the study was released), the homeland of the ancestors of the Phrygians, models come out with something much more substantial. Using EHG as a proxy is stupid for this reason, especially on ADMIXTOOLS where outgroups correct for some admixture component “absorbing” another.

Anyways, Anatolian culture also does not reflect a particularly heavy Indo-European transformation. Certainly it seems more influenced by Indo-European characteristics in the west, but the Hittites were much more influenced by the cultures around them which definitely were not Indo-European. There are much more deities among the Hittites which were of Hattic, Semitic, Sumerian, or Hurrian origin. And this would reflect the explosion in CHG/Zagrosian type DNA. The 4th millennium BC saw the rise of the Uruk culture, which probably brought genes into Anatolia just as much as they exported their culture and religion into Anatolia. Hattic is often proposed to have come from further east, along with the Kaskian language. The Hittites were most likely an elite group from the small kingdom of Kussara who were supported by Hattians. The Hurrians are well known to have come from the east. There was clearly a migration of eastern peoples into Anatolia completely divorced from the Indo-European phenomenon, but going on at around the same time. Due to the lower distance and higher population density of these peoples, they were much more capable of affirming their influence in Anatolia through demographic changes rather than through small elite groups forced to marry into the local elite in return for cultural influence in the way we saw later Steppe groups do in other contexts.

The biggest evidence of this, is that the same phenomenon happened during the Bronze Age for other parts of the near east.

It also happened in the ancient Aegean…

It even (probably) happened in Egypt!

OK = Old Kingdom. MK = Middle Kingdom. NUB = Theban from Nubian period (not a Nubian. An Egyptian under the Nubian dynasty).

And we know for a fact that Semitic did not diffuse into the Levant from the east during the Bronze Age. We know that Egyptian did not diffuse into Egypt from the east. And it would certainly be quite the shock if Minoans were actually speaking an Anatolian language (speaking of which, maybe that could be explained by the Steppe-rich sample(s) in Neolithic Greece...) So why is it assumed that this same transformation happening in Anatolia is indicative of linguistic change?

Linguists have always been in favor of Anatolian languages entering Anatolia through the Balkans, not eastward, because Anatolian languages are more diverse in the west than they are in the east. This implies that Anatolian languages had more time in the west than they did in the east, suggesting a movement from into the Balkans. Another (very much supplementary) line of evidence for this is the supposed “Hittite” substrate in Celtic languages. This could be residual influence from early Indo-European groups in Central Europe, who split off from the Anatolians in the Balkans. Another reason is that early Indo-European language does not have a great deal of terms related to farming. meanwhile cultures south of the Caucasus were heavily engrossed with farming. There are some archaeological ties between Northwest Anatolia (especially Troy) and some Balkan cultures along the Black Sea (which we now know had Steppe ancestry). Particularly the Cernavodă culture (4000-3200 BC) and the Ezero Culture (3300-2700 BC) to its south. Although, I don’t think these later Bulgarian samples pictured below belong to the Ezero culture. I’m pretty sure they are just some sort of colonists from the Yamnaya culture, and that site isn’t archaeologically classified as Ezero.

Anyways, back to the Steppe… There are other reasons why the Steppe, and prior to that EHG, were probably the origin of the Indo-European languages. Firstly, archaeological evidence has always connected Yamnaya and cultures related to them to other Steppe cultures, like Sredny Stog, Repin, and Khvalynsk. It doesn’t show some sort of introduction of material or burial culture which resembles something Caucasian let alone Transcaucasian. Certainly, any connection is not as great as the connection to those prior cultures. Secondly, and more importantly, the Y-DNA of Yamnaya men is almost certainly descended from locals, and not from the Caucasus. Yamnaya men are predominantly I2 and R1b-Z2103. Corded Ware and Bell Beaker men actually have different haplogroups. Corded Ware are predominantly R1a, and Bell Beakers are predominantly R1b, but not Z2103. Bell Beakers are associated with R-L51.

I2 is definitely a Hunter-Gatherer haplogroup. There’s really no question about it. R1a is also almost certainly something EHG in origin. EHG from over 10,000 years ago (Karelia, Peschanitsa) have R1a. Early Steppe cultures like the Eastern Hunter Gatherers at Samara and the Khvalynsk culture have R1a. I don’t really think anyone is arguing for R1a coming from south of the Caucasus anymore except Indians. R1b is a bit of a tougher nut to crack, since it probably actually originated among Western Hunter Gatherers and then was introduced to EHGs. But, it definitely was among EHGs in Samara and Baltic HGs (mixed EHG/WHG) in Latvia. Particularly, R-P297, which is a progenitor of both Z2103 and L51. R1b is also found in the Botai Culture, which is predominantly descended from Ancient North Eurasians and EHGs, with some Siberian admixture as well. R is found earliest among Ancient North Eurasians, who were ancestral to EHG. P and Q are also found among ANE samples (Yana RHS and Afontova Gora 2), which are ancestral to R. Q probably entered Siberian populations through ANE. Meanwhile, the Caucasus at the time were predominantly inhabited by J’s and G’s.

P.S. I’m pretty sure there were a few Yamnaya not in this sample with G2 and J2, but a very small minority at most.

Certainly, it is possible that some warlord in the Caucasus happened to have R1b by pure chance, and then expanded northwards. But, extraordinary aims require extraordinary evidence, yadda yadda. After all, Caucasian peoples had EHG ancestry themselves.

So… Why would a patriarchal, hierarchical culture known for spreading itself through Y-Haplogroup turnovers (if not autosomal turnovers) adopt a language of these Caucasian peoples who seemed to leave no paternal impact? Were they speaking the languages of their wives? You know, who they paid for or bridenapped, and described marrying as akin to breaking in a horse? Well… I believe Lazaridis (an author of the Harvard study) responded to this informally by pointing out the disparity in mtDNA suggesting that, if anything, the CHG ancestors of Yamnaya left less of a maternal impact than a paternal impact. But, this isn’t really that relevant in my opinion. It is not impossible or improbable that over time, for some reason or another, local matrilineages beat out foreign matrilineages. Another explanation could be a shift in mtDNA in the Caucasus from the time of the admixture event with the steppe (which is probably far earlier than Proto-Indo-European anyways, despite what the Harvard team is trying to push). And I don’t know what argument Lazaridis was even making, since prior studies found Steppe mtDNA as more similar to Caucasus mtDNA than Steppe Y-DNA was to Caucasus Y-DNA (see above). It still doesn’t explain the total lack of these CHG and Transcaucasian patrilineages, while there is an abundance of haplogroups which show no evidence of being from the near east and plenty of evidence for being from EHGs and other European HGs.

I don’t know why this paper was so quick to the trigger on these claims. Academics tried to move away from a European origin for the Indo-European languages for blatantly political reasons at the end of World War II. Its inevitable return certainly disturbed them at least on a subtle level. Not to mention, the significant and increasing influence of Indians in academia, a group which has always had a rabid hatred for the Steppe Urheimat theory. The obsession with Oriente Lux predates WWII though, many academics have always been interested in just tracing everything back to the Near East. Either way, the evidence provided wasn’t even considered conclusive by the authors, so people acting like its gospel are just retarded. Funny thing is, I’m probably forgetting a few studies that I was making at the zenith of this whole debacle, because it was a while ago.