The Yamnaya culture and it's legacy

Yamnaya was an Early Bronze Age culture (3300-2700BC) that expanded in Pontic Caspian steppe in East Europe. It is remarkable due to its homogenous population which contrasts to the diversity of the preceding Eneolithic (means Chalcolithic) cultures. In just 1000 years genes associated with Yamnaya culture reached what is now Portugal and western Mongolia, Britain and South-Central Asia. The scale of this expansion can be considered as one of the most important demographic events in West Eurasia after the invention of farming in Fertile Crescent. In 2022 Lazaridis et al. announced that all modern living IE nations languages can be derived from the Yamnaya.

In 2015 the first ancient DNA of Yamnayans was published and it became clear that they do not directly descend from Eastern European hunter gatherers. Roughly the half of their ancestry was from south, and the "Armenian like" term was used to label this southern part. Then in late 2015 the Caucasian hunter gatherers (CHG) were discovered. In 2016 Iranian Zagros Neolithic farmers were published and since then both those populations were used to model the Yamnayans.

Despite those statistical models the formation of Yamnayans gene pool was remaining an obscure subject. Finally in April 2024 two large papers dedicated to this subject provided more clarity. Various Neolithic and Eneolithic groups living in East Europe had both shared ancestry and differences. One group who was living from North Caucasus foothills to Lower Volga (site Berezhnovka ) was labeled as "Caucasus Lower Volga cline". CLV had both southern Caucasian hunters and farmers (Aknashen) ancestry.

I will dedicate a special topic to describe what is CLV but at this stage it's sufficient to say that Yamnayans had up to 80% of their ancestry from CLV. Not only Yamnayans but all Kurganic cultures (Sredni Stog, Khvalynsk, Cernabida, Usatovo) of eastern Europe and Balkans had the CLV ancestry mixed with various local European populations. For the case of Yamnayans the second part of their ancestry was the Ukraine hunter gatherers (UNHG), which made them similar to Sredni Stog. Even though the most important Y DNA's were missing in Sredni Stog which could mean that Proto Yamnaya's exact immediate geographic location is not found yet but shouldn't be outside from Don - Volga - Caucasus triangle. With more sampling we will learn the exact origin of the two most successful lineages. R1b-M269 and R1a-M417.

The samples from the new paper are not available yet but we have two samples from the forementioned CLV cline. They are from the Progress - 2 site in the North Caucasus (4500-4000BC, see the last map for the location). I used them to model various Yamnaya samples and got the same result as in Lazaridis 2024. More than 75-80% of ancestry in regular Yamnaya samples is from the Progress / CLV. Some outliers have extra ancestries which is normal given their location.

You may wonder how "Hittites from east, from historic Armenia" and "Yamnaya from CLV" are connected to each other? After all there is no evidence that Hittites and other Anatolian speakers descend from Yamnaya culture. What is the common shared ancestry between IE Anatolians and Yamnaya?

Another topic will be dedicated to this subject.

And at the end here is the list of Y DNA found in Yamnaya at this stage. R1b-Z2103 (west Asia Balkans), R1b-L51 (west Europe), R1a-M417 (Asia - Europe), I2-L699, J2b-L283 (west Balkans), R1b-PF7562 (Balkans Mycenean ) Q-L939 and isolated cases of J1 and R1b-V1636.

See also

• Population dynamics in North Caucasus

Hittites from east.

Before commenting the deeper origins of Indo Europeans proposed in the new Lazaridis 2024 paper I want to speak about the origins of Anatolians. Since the advent of paleogenetic it became clear that they can't be native to plain Anatolia. They could be either from east or from north-west. Since 2018 the accumulating data was giving clear preference to the eastern origins. From historic Armenia. But a doubt was still persisting, because the Balkanian route was not fully investigated. In 2023 Penske et al. has found a migration from steppe to eastern Balkans before Yamna around 4000BC but they didn't even discuss the possibility that they could be related to Hittites. It's because their autosomes were rapidly diluted having a limited impact in Bulgaria yet alone in Anatolia. Also the Y DNA associated to this migration the I2-L699 was missing from inland ancient Anatolian samples.

While a R1b-V1636 was found in Bronze Age Gaziantep quite close temporaly and geographically to the kingdom Armi mentioned in Eblaite texts around 2500bc. Hittites names have been discovered in this kingdom recently.

The V1636 was popular in CLV cline stretching from Caucasus to Lower Volga and was found in ancient Malatya and Sevan basin also.

Besides this R1b other haplotypes found in Bronze Age Anatolia also show self-evident eastern affiliation.

The J1-Z1828 who's subbranch BY94 can be an Iron Age Luwian lineage like the J2- L70 which itself is derived from the L25 a haplotype from east.

A J2-Z6065 which was found in Masis Blue in Neolithic Armenia. We have also bunch of of J2-M67 and G2-M406, even though this latter is related to Hattic people so it can't be interpreted as a prove of Hittites from east.

In conclusion all ancient Y DNA that we have in inland Anatolia do have an eastern origin, from historic Armenia and around. One of them is further from north of Caucasus region. Their autosomes are also shifted to east compared to the preceding periods.

One exception is the I2-P78 from MBA Yassi Tepe near Izmir, which is clearly from Mycenean Greece, so this person was almost certainly a Greek people representative who probably settled in western Anatolian coasts after the 2000BC. They are mentioned in Hittite texts as Ahhiyawa.

In sum we can now say with high certitude that Indo-European languages were spoken in Armenian Highlands at minimum the last 6300 years. The 4300BC is not a random number it's the date of the start of Late Chalcolithic associated with Chaff faced ware which had a tremendous impact on southern Anatolia. But not much on northern Anatolia where Hattic related people were settled since more earlier times. Probably since 5800BC.

But that would require another thread which I will do the next time.

PS a proposed location of Armi. Another location is in the northeast of Ebla.

See also

• One possible way to know the language(s) spoken in Kur-Araxian culture is to examine the onomastics in those places where there was continuity

Samples of Anatolian/Aegean origin from the Roman empire city Vimminacium in what is now the eastern part of Serbia

Samples of Anatolian/Aegean origin from the Roman empire city Vimminacium (Olalde 2023) in what is now the eastern part of Serbia. A large military camp was present in the city.

• E-L791 is remarkable because it's very close to Napoleon's Y DNA. Another L791 from Roman era Croatia.

  Distance to: Croatia_Zadar_Roman_oLevant.SG:R3742.SG
  0.03003301 Greek_Cappadocia
  0.03236809 Greek_Central_Anatolia
  0.03434637 Greek_Crete_Heraklion
  0.03806465 Armenian_Gesaria
  0.03847296 Greek_Crete

• I2-Y16419 is ultimately from Trialeti-Vanadzor culture. Based on genetics he was probably from Kesaria/Kayseri.

• The presence of J2-L70 and J1-BY94 is without surprise. These two haplotypes expanded in Iron Age Anatolia most probably.

• And the T1a2 could be local but in most likelihood, he was also from Anatolia.

Also, a lot of local Balkanian specific haplotypes like E-V13 with some Germanic ones.

Another evidence that Roman Empire had a significant impact on the genetics of South Europe. Read in more detail in the link below.

 

•  In 313 AD, the Emperor Constantine issued the Edict of Milan, which accepted Christianity

 

 

Remarks on the Van Urmia ware culture. VUC (2200-1400/1300BC)

Previously I already had a review on VUC. Here I will add some remarks on the frequently asked questions.

+ VUC is a sister culture of Trialeti-Vanadzor culture, it does not descend from it. Both descend from the same parent culture, but despite this they have differences. VUC had painted pottery, while TVC had a different pottery. The grave was directed north-south in VUC while TVC had a east-west orientation. And most importantly TVC were practicing cremation while VUC not.

+ How exactly migrated Proto VUC people is not clear. One option is via Araratian plain but the presence in Aji Chay river basin, toward the Ardabil region suggests that they could have entered from the Mughan region toward the sources of Aji Chay then moved downstream toward the Urmia basin, and then to Van.

+ Karmir-berd culture descend from VUC. We don't have samples from this culture. The Dzori geghi outlier can be related to it. The J2-YP879 from Keti LBA can be another descendant of Karmir berd culture.

+ Unlike the TVC the VUC do not have an evident progeniture. Over time new people settled in Urmia basin from north Mesopotamia and Central Iran cropping the territory of VUC people. You can witness this change with the genetics. Hasanlu MBA (a hypothetical VUC sample) is close to Assyrian - Armenians while the Hasanlu IA is close the Kurds. In most likelihood the regions that best preserved the VUC ancestry are in the eastern Van and northern and western Urmia. South and east of lake Urmia had too much genetic shift. This means that most plausible descendants of VUC culture should be searched in those regions.

+ During Iron Age in forementioned regions we have Biai people from which the Biainili term is derived. We have Armarili (another name Aramali ) supposedly near modern Salmas. According to Sargon the ruling dinasty of Urartu was from Armarili. We have Ayadi near modern Urmia city and Uayis with disputed location. It is quite possible that this Uayis is another name for Biai, which was pronounced as Væy - Vay. According to Petrosyan later terms like Vayoc' dzor are related to this Vay people name. A term that he links with Hittites though.

+ Based on this logic the Manna country doesn't seem to be a good candidate for being a direct descendant from VUC. It has too much Dinkha tepe 2 type ancestry related to Grey ware. Nevertheless, the Hasanlu IA had too much of R1b to be completely unrelated to VUC. One possibility is that some R1b moved further south and east and became part of Grey ware. And reexpanded later with them. This would explain high level of R1b in some Iranian populations like Lurs. Another reason is that some of Hasanlu samples are different genome wide. They are closer to older VUC profile and plot close to Armenians and Assyrians. You can see an example in the fifth chart. Overall, it seems that the Manna country had a multi-ethnic composition and more ancient DNA is needed from south of Urmia to understand better it's structure

Linguistic is a mature science and we should not expect any groundbreaking findings from a new paper.

Linguistic is a mature science and we should not expect any groundbreaking findings from a new paper. What can do a modern linguist is to have a new interpretation of already known facts, or at best he can find a few new lexical parallels. Nielsen's paper is important because it shows that the linguistic data do not contradict to the available genetic data about the origins of Armenians. And more important it can reinforce it, giving new details.

The most important argument for Proto Armenians dwelling in the northern parts of historic Armenia is the phonetic system of Armenian which is very close to Georgian and Zan phonetic systems.

Creanza et al. 2015 analyzed more than 2000 languages and based that came to the conclusion that the phonetic systems of two neighboring languages correlates better with geographic distance than with their affiliated linguistic families. Another interesting conclusion was that an isolated language drifts. But unlike the genetics were the drift decrease the diversity, the drift in the language increases the number of phonemes.

This by the way can explain the high number of phonemes in the North Caucasian languages. Returning to the Armenian and Kartvelian we can say that similar phonetic systems mean relatively long period of coexistence.

On the other hand, this do not mean that the Kur-Araxian culture which became the substrate for the Armenian language was necessarily and predominantly Kartvelian. First the number of mutual loanwords is too small for that. Most loanwords in Armenian are from the Zan branch which can be connected to the Colchian archaeological culture. Moreover, we do not know well about the phonetic system of Urartian. Given that we know about their phonology only via the cuneiform it is possible that their phonetic system was also close to the Armenian.

Another important question is the migration of IE Anatolians. If further genetic data do not show evidence that they migrated via Balkans, then the only available option left will be their origins or migration via the historic Armenia. In this scenario Kura-Araxes becomes indispensable for their spread. So, there is a need to look at the possible IE Anatolian substrate in Armenian. Nielsen mentions this.

And finally based on the available genetic and archaeological data there is little doubt that the Nakh-Daghestani linguistic family descend from a subset of Kura-Araxes. We can't now deduce the exact boundary of this subset, but we can expect that Proto-Armenian would have a contact with them also. So, another analyze of Armenian and Nakh-Daghestani connections is needed.

Returning to the Kartvelian family. Based on the current genetic and archaeological data the most likely cultures related to the Kartvelian family were the Proto-Colchian (2700-1700/1500BCE) and the Colchian culture (1700/1500-700BCE) in the western Georgia. Some Kartvelian presence in Koban culture is also possible but it's not relevant for ancient Armenian connections. What is more important is that Colchian culture axes were found in northwestern regions of historic Armenia which can explain the stronger presence of Zan loanwords in the Armenian. More ancient DNA and archaeological research from that region, will help to better understand this question.

See also

• Ancient Kartvelians and the Formation of Iberia

The origins of Grey ware culture in Iran

We have a decent number of ancient DNA from Urmia basin. Based on that we can say that the Urmia basin was a dynamic place. One of interesting findings is the increase of Iran/Zagros Neolithic ancestry in LBA period. (after the 1600bc). This increase started probably at the end of MBA.

The best archaeologic event that matches this genetic shift is the spread of Grey ware. Initially Grey ware was considered as an Iron Age pottery, but new studies of Iranian archaeologists propose a Bronze Age spread of this pottery which is found in many sites in North Iran (see the map, Fahimi 2019)

Apparently, the current genetic data supports their theory. The origins of this pottery were debated. Northeastern Iranian origin was proposed, but a North Central Iranian origin is also possible. The closest populations to Dinkha tepe 2 are Lors and Mazandarani. If this has any predictive value, then a Central Iranian origin seems more plausible.

In any case the current genetic data supports a more eastern origin of this pottery than the lake Urmia.

As for the ethnicities related to this pottery, the best candidate are the Kassites. Kassite's language remains unclassified. But it seems that they had an Aryan adstrat which is quite plausible given that Mitanni Aryans appear roughly the same period. They were also horse worshippers which also was in most likelihood an Aryan influence.

It is quite possible that Mitanni Aryans were also making the Grey ware before they moved to Syria. Where different potteries were produced by locals.

See also

Hasanlu_LBA_A singleton is almost certainly an Indo-Iranian person affiliated to Mitanni Aryans

The genetic profile of eastern Jews.

Most of You have probably noticed that some subgroups of Jews usually plot close to modern Armenians and to ancient samples from historic Armenia. The reason of this was not clear until now. But thanks to ancient DNA from Bahrain, we can now propose an explanation of this genetic puzzle
Asian Jews are usually known as Mizrahi Jews. But given that the groups of Jews that we will discuss in this post are only a subset of Oriental (Mizrahi) Jews, I will use for them the term Eastern Jews. The groups that usually appear genetically close to Armenians are from Iraq, Iran, Georgia, Kurdish, Mountainous Jews from eastern Caucasus and Uzbekistan.
In the Central farmer calculator, their genetic profile shows common features. They have relatively high level of "Central farmer" and more or less equal number of Levant and Zagros Neolithic ancestry. We don't have ancient DNA from pre-Islamic Mesopotamia but using the samples from ancient Bahrain we can deduce that this type of genetic ancestry was widespread in ancient Mesopotamia. You can see that modern Iraqi people are quite different from Iraqi Jews. Modern Iraqi Arabs have much higher Levantine ancestry and even some African ancestry which Jews and ancient Bahraini lacks. This is due to medieval Arabic migration from south.
This genetic pattern fits well with the historic events that triggered the formation of Eastern Jews. In 536 BCE Persians conquered Babylon. And the Jews from Babylon settled in Persia. From there they dispersed to various corners of Iranian empires. I am not big expert of Jewish history but based on overall understanding of the history I assume that Mountain and Uzbekistan Jewish diasporas formed in Sassanian empire period. While the Georgian Jewish community could have formed little bit earlier in Arsacid period.
It seems that all this Jewish communities derive their ancestry from the same Babylonian Jewish community which had a Mesopotamian genetic profile.
As for the reason of the Mesopotamian genetic profile plotting close to Armenians, it is due to the fact that both have a Central position on the genetic map relative to eastern Iran-Caucasus and western Levant-Anatolia. It's also possible that ancient middle and lower Mesopotamia was settled by migrants from the north which created a genetic connection between Mesopotamia and historic Armenia. With more ancient samples this puzzle will be solved.
PS the charts below show the closest populations to ancient Bahraini samples.