search this blog

Tuesday, September 15, 2015

Early Bronze Age migrants and ethnicity in the Middle Eastern mountain zone

Open access at PNAS:

Abstract: The Kura-Araxes cultural tradition existed in the highlands of the South Caucasus from 3500 to 2450 BCE (before the Christian era). This tradition represented an adaptive regime and a symbolically encoded common identity spread over a broad area of patchy mountain environments. By 3000 BCE, groups bearing this identity had migrated southwest across a wide area from the Taurus Mountains down into the southern Levant, southeast along the Zagros Mountains, and north across the Caucasus Mountains. In these new places, they became effectively ethnic groups amid already heterogeneous societies. This paper addresses the place of migrants among local populations as ethnicities and the reasons for their disappearance in the diaspora after 2450 BCE.

Mitchell S. Rothman, Early Bronze Age migrants and ethnicity in the Middle Eastern mountain zone, PNAS, July 28, 2015 vol. 112 no. 30, doi: 10.1073/pnas.1502220112


Davidski said...

Anyone know how the BA Armenians from Allentoft et al. fit into this picture?

Mike Thomas said...

The K-A culture originated in the south Caucasian piedmont. Its beginning was virtually the same time as Majkop (C. 3800 - 3500BC), and both likely began due to influences and impulses from the Mesopotamia and/or Central Asia, whether trade or colonization, etc. From the south Caucasus it expanded down to eastern Anatolia and even to Syria, and to Iran on the direction, later. So the expansion from the southern Caucasus to southwest and southeast mirrors that from Majkop to the NorthEast and Northwest.

The latter part of the K-A culture has not been defined definitely. There is some evidence for possible movements from north of the caucasus to the south c. 2500 BC, and the K-A culture fragments by c. 2000 BC and each region goes its own way.

IM not sure exactly it related to Armenians, possibly nothing directly because the latter are Iron Age peoples.

jackson_montgomery_devoni said...

I have a strong feeling that the the Kura-Araxes population will be dominated by Y-DNA haplogroups J1 and J2.

Davidski said...

Interesting quote from the above paper:

"The beginning of mobile populations, marked by kurgans and the contemporaneous building of walls at Shengavit, Mokhra Blur (23, 28, 29), and Ravaz (48), is already evident in the early third millennium BCE. Ultimately, in the homeland, the Kura-Araxes adaptation would be displaced by a more mobile and militaristic one associated with the so-called Kurgan Cultures (15, 49)."

Nirjhar007 said...

It seems that the Kura-Araxes was a culture complex,as observed also on German wiki (, but the core area in historical times was occupied by Urartu and Caucasian Albania, two areas related to NE Caucasian languages (see this: The western Araxes was also occupied by Armenians, and also Kartvelian Georgia was part of it. So, the Hurrians and related people are probably involved, carriers of J1 and maybe also R1b, but also Kartvelians and probably some Proto-Armenians.

Aram said...


Thanks to Your Afanasievo K12 calculator I made a graph where BA Armenia samples are in chronological order.

It seems there is a trend of increasing 'Afanasievo' component that reaches it peak somewhere at 1200-900 BC.
The oldest sample RISE 413 is aged 3800 years, and it clusters with modern Armenians on the PCA. The RISE 413 is definitely after (700 years after) Kura-Arax and it is from Trialeti culture period.

So I expect Kura-Arax to be more Bedouin and less 'Afanasievo' than BA_Armenians. Of course it is a just speculation based on a small sample.

Mike Thomas said...

The article is very impressive.

I found this interesting.
"On a continuum, they {ie K-A} are much more similar to the styles of other peoples of the mountains than those of lowland Mesopotamia" (p 9192)

Aram said...

I link the apparition of that South Caucasian kurgans with the Z2103 coming from Daghestan at 2700-2300BC. It is possible that Z2103 started to infiltrate even in earlier periods.

In places where Kur-Arax culture survived and continued it's tradition in modern Syrian Kurdistan and Taurus mountain (Malatya) region the Hurrian cities are attested, that is why some scholars suggest a Hurro-Urartean identity. But some others think a IE presence also. For example Proto-Anatolian. If Proto-Anatolians where already there, they must occupy a more Northern zone because the Southern zone is definitely Hurrian. I expect Hurrians to be a predominantly J2, with some J1, G2, R1b and T. Mitanni starts at 1500 BC so some R1a-Z93 is expected there also.

Aram said...

Yes Mike

They were definitely highlanders. They adapted very well to mountainous relief. That's why I am sceptic about Uruk migration theory. Of course there could be some migration from Mesopotamia but I have hard time to imagine how people from Mesopotamia can become the dominant people in highlands.

Nirjhar007 said...

This is also significant-
''There is little indication that the interactions were at all violent. Ethnic communities over time integrated into these populations while keeping their identity.''.

Davidski said...

Isn't Kura-Araxes supposed to be Proto-Indo-European according to the Armenian Plateau hypothesis?

If so, then why aren't Indo-European languages attested from most of the Near Eastern territory into which the Kura-Araxes people spread?

Also, the part about the militaristic Kurgan groups putting pressure on the Kura-Araxes people in their homeland, thus forcing them to emigrate south, doesn't gel too well with the Armenian Plateau hypothesis.

Aram said...

Yes Nirjhar

That is important. Despite the fact that they had some weaponry, and they knew how to build fortification they were not militaristic. But when they wanted to enter in to Northern Mesopotamia they had some conflicts.

For the Proto-Armenian, in Gamkrelidze Ivanov theory the Kur-Arax is a major source of IE people, then they move to North, Maykop, then to Steppe. Proto-Armenians remain in South. Indo-Iranians spread by Southern route in Iran and India.
But there is no genetic evidence for that theory yet, that's why I am very cautious on that question.

Mike Thomas said...


I don't think G& I linked their armenian hypothesis to any particular culture . Theirs was mostly a linguistic arguement , although they mention something about chariots ..

Aram said...

Yes Davidski

According to Armenian Plateau theory the Kur-Arax is the source of IE.

And of course the fact that some Kurgan people from North pushed them to South is a problem for that theory.

p.s. In any cases Kur-Arax had regular contacts with Steppe. Their pottery is found as North as around Volga river.

Nirjhar007 said...

Well David the Caucasian region does not seems to have any significant gene flow from S Russia , The K-A seems to be a cultural complex (had different language speakers) but IE were there from very ancient times, the connections Kartvelian etc is a good indication of that.
Yes i know but don't worry i'm preparing something which will blow your minds ;).

Nirjhar007 said...

^ the connections *with* Kartvelian etc.

Aram said...

Kur-Arax is viewed as one source of IE, the other is Maykop. And some others in North Iran.

Mike is right G&I themselves didn't suggest any culture. The most important criticism of Gamkrelidze Ivanov theory was that they didn't try to explicitly link their theory to any archaeological culture. Later Stanislav Grigoriev worked on that linking.

The other criticism was the presence of attested non-IE languages in that proposed 'homeland'.

Nirjhar007 said...

Actually IMO ''Teal'' comes from the Iranian-Caucasian area, there developed IE but also other languages like Hurrian and Kartvelian, which reveal some connection with IE!

Chad Rohlfsen said...

KA and Maikop can be from the same source population but adapt to geography and mixing with different folks. The Caucasus, through Iran, is pretty similar as advancements and pre-kurgan burial.

Mike Thomas said...


To answer your question about the armenian hypothesis being now disproven- the answer is not really. (Not that it's what I've espoused)

First of all, this movement of a "kurgan like" culture could have local origins because I've already mentioned that the k-A had a pastoral, mobile lifestyle to begin with, and kurgans from even pre-KA era, although never as common as it became on the steppe. And we do have evidence of movement from south to north of the Caucasus.

Secondly, the KA culture ending was more of a complex fragmentation than a northern invasion, and the middle Bronze Age in the South caucasus and eastern anatolia is poorly studied at present. Certainly if one read Katherine marro's work, one isn't left with an impression of clear cut trajectories and invasions. In fact some areas show movements from the East, some have continuity, etc ..So just like during the heyday of K-A, and other cultures like BB and Yamnaya , these 'cultures' were riss-crossing networks of connectivity, and not simple sweeping movements north to south, or east to west.

So the evidence can be interpreted in any way fitting the eye of the beholder.

andrew said...

Either K-A or Maikop or both were very likely the source of the technological packages that gave rise to the Copper Age and Bronze Age in West Eurasia (including South Asia) and probably to the subsequent Iron Age as well. But, it is hard to say how that translated into language and culture.

Personally, my inclination looking at the data is to see Maikop as the immediate source for the technological package that PIE people adapted and used to drive their expansion, while K-A looks attractive as the ultimate technological package source for Bell Beaker spread of those technologies in Western Europe (the exact route is hard to figure).

My guess is that K-A was heavier in Y-DNA R1b (and lighter in Y-DNA J1, J2 and G2) than some of the other posters have suggested.

But, I don't claim that these are much more than educated guesses based mostly on the modern genetics of the region and history.

Mike Thomas said...

Fair suggestions. But I'm curious as to how you Link KA with BB ?

andrew said...

The BB culture was driven by skilled metal workers who migrated to areas where there were metal deposits in Europe, but for a variety of reasons too involved to discuss in this comment, probably weren't linguistically Indo-European. And, during the existence of this culture and its pre-IE descendants in Western Europe, the proportion of Y-DNA R1b in that region surged a lot.

In a nutshell, both the linguistic angle and the genetics (and other evidence) make a Southern Caucasus source for the technological package seems a much better fit than a Northern Caucasus source at the time of Maikop/KA and there aren't really any other very good sources for the technological package. KA is also a pretty good candidate for the source technology for pre-IE Hattic people in Anatolia and the Minoan culture in Crete (the later of which is very similar to modern Western Europe). The Hattic culture in Anatolia may not have been particularly old (e.g. it may have only been 500-2000 years old) when the Hittites, in turn, conquered Anatolia from them. The Hattic people, in turn, may have conquered the descendants of the EEF people in Anatolia using their superior metal working technology which they had long before it was widespread elsewhere.

Mike Thomas said...

I certainly agree it's possible. Although this R1b issue needs resolving . It might explain why so many R1b clades arent IE, as well as the 'oddness' of Iberian copper age.

But I thought you'd stumbled across a good archaemetallurgy paper proving such connections

Va_Highlander said...


The data I have suggests that annealing of native copper was practiced in eastern Anatolia, ca 8000 BCE. Copper appears outside of its original core area of eastern Anatolia and northern Iraq by 6000 BCE. The earliest and best-documented evidence for the smelting of copper has been found at sites such as Tal-i Iblis, in southeast Iran, and Belovode, in eastern Serbia, dating from the late-sixth/early-fifth millennium BCE. Maykop obviously post-dates these developments by millennia and cannot, therefore, have given rise to the Copper Age.

Copper alloyed with arsenic or lead, whether accidental or deliberate, appears throughout Southwest and Central Asia and Southeast Europe by the late-fifth/early-fourth millennium BCE and arsenical and antimonial bronze were deliberately produced by the middle of the fourth. Tin bronze first appears in Southwest Asia by the end of the fourth millennium BCE and in Central Europe and Central Asia by the early-third. Given this chronology and areal distribution, it's a bit difficult to see how these developments might be related to Maykop either, or to Kura-Araxes.