Update on J1-Z1842 and Kura-Araxes culture.

 Update on J1-Z1842 and Kura-Araxes culture.

The recent Agdzhoyan et al. paper about Dagestani people added some long waited details to the structure of the J1-Z1842 and its distribution.

We learned that beside ZS3114 and CTS1460 there is a third important branch in Z1842, which is very specific to Dagestan. Its ZS3042 which is absent in Armenians and apparently very rare in South Caucasus. In Yfull many Dargins and Kaitags are present in that branch.

Another important information is the branch ZS3114 which is more frequent in Lezgins than other Dagestani people. It has a more southern distribution. In Armenians its found almost exclusively among eastern Armenians. If the language of Lezgi speakers is related to this ZS3114 branch then this can support some linguistic theories making Lezgi-Udi group as the first split in Nakh-Dagestani family.

All this data forces us to rethink the origin of J1-Z1842 and place its locus of differentiation close to Eastern Caucasus range.

However I still dont think that the expansion of Z1842 started from the northeast of Caucasus. There are many arguments against such a theory. But one the most strongest is the autosomes of oldest Z1842 ancient DNA. They all have significantly lower levels of steppe ancestry than Areni cave preceding them. Its very hard to imagine that Z1842 lived for thousand years in the north and did not acquired substantial steppe ancestry.

What this data tell us is another possibility. The branch specific to Dagestan the ZS3042 could have diverged earlier with Chalcolithic culture Sioni - Ginchi (5000-4000bc). Ginchi is a site on north Caucasus with similar pottery as in Sioni (Georgia). As for the parallel CTS1460 which has a great diversity in South Caucasus/Near East and was found in multiple ancient Kura֊Araxes sites is a genuine Kura-Araxes lineage. Its TMRCA in Yfull also more realistically fits to the spread of Kura-Araxes culture.

And finally about the origins of KA culture itself. Recent claims of KA pottery in Chalcolithic levels of Areni, Ovčular tepesi in Nakhichevan and in some sites in Turkey combined with the fact the extra CHG/J1 ancestry in KA samples was from more eastern regions means that the formation of KA culture and its later spread are not necessarily done by the same population. So there must be two distinct KA with probably two distinct autosome profiles.

Alternatively the extra CHG could still be from unsampled forest mountain regions of Lori and Tavush. They could harbor a hunter gatherer (Aknashen like) enclave different from Areni and Shulaveri culture. But even with this scenario the contradiction is not solved.

According Passerini et al. the oldest KA was in lowlands with collective kurgan burials. Only starting from 3300BC highland KA with permanent settlements emerged. This later KA that we can call KA 2 can have a different locus (left side of Kur river?) and population (CHG rich) than the first one. KA 1 in the other side can be more related to Areni type population. This hypothesis obviously require confirmation by ancient DNA

The patriarch Phawos (Փաւոս)

Khorenatsi lists multiple kings/patriarchs for a short period that fits into the end of Urartu and the rise of Orontid (Eruanduni) Armenia. This can be interpreted as an argument for the complexity of that period.

We have already seen one of patriarchs of that period. The Pharnawaz who seems to be a real person related to Parnaki tribe. Another patriarch of that period is Phawos pronounced as P'awos. His name is written in various forms in different texts but the canonic form is Phawos (Փաւոս).

Finding an ancient tribe or person related to him seems impossible. But the more close look on ancient provinces of Armenia gives a remarkable result.

Strabo in his Geography mentions a certain province Phauene in Armenia. Few paragraphs later he mentions Phaunitis between Basoropeda (Vaspurakan) and Caspiane (Paytakaran). 

There is also Phauenê,⁠ a province of Armenia, and Comisenê, and Orchistenê, which last furnishes the most cavalry. Chorzenê and Cambysenê are the most northerly ...

....

(Artaxias and Zariadres) enlarged their kingdoms by cutting off for themselves parts of the surrounding nations, — I mean by cutting off Caspianê and Phaunitis and Basoropeda from the country of the Medes;

If we remove the Greek suffixes usually added to foreign toponyms we get the root Phau/Phaun. This root is virtually identic to the possible root Phaw- in the name of patriarch Phawos. We can deduce that the name of Phawos is derived from an old tribe that lived probably somewhere in Syunik/Artsakh and/or south of Araxes also in the adjacent northern regions of Iran. 

This hypothetic Phaw-os tribe gived his name to that region but overtime the name was lost.  Some scholars has proposed that the name Phauene is preserved in Pharisos of Artsakh (north of Mrav range) while Hewsen suggested that Phaun-itis is a corrupted form Siwnik/Syunik. Both proposals are problematic from linguistic perspective, even though in one manuscript the name of Phawos was written as Pharos.

Keeping aside whether the name survived to medieval period or not, there is another interesting name in ancient texts. The tribe Pausikae was living next of Caspians (Paytakaran) in 11th satrapy according Herodotus.  If we consider -ikae as a Greek suffix or a particle from the language of Pausikae we get the root Paus- in a region where later Strabo mentions Phaun-itis. This is a strong argument that such a tribe lived in that region.

Finally the ultimate origin of Pausikae/Phawos is obviously unknown. Some scholars suggested an Iranian origin for Pausikae based on a vague similarity of name with some Central Asian tribes. If -ik is a suffix from their language then they could be both Armenic and Iranian. And finally if -os was a suffix from their language then this links them to the wider Gutian group where such a suffix is found in names. My own opinion is that Phawos more likely is descending from the R1b tribes present in that region. They became part of wider Armenian identity while some others were probably absorbed in Iranians.

Essay on the language of Gutians

The R1b from Bakr Awa (Mesopotamia) is now published in FTDNA. Its confirmed as Z2106 and is positive to Y100087 snp. This branch has strong presence in ancient remains from Ukraine and Russia. Thus the steppe origin of this Bakr Awa R1b is beyond the doubt.

This is the first case of Y100087 from West Asia, while other cases of Z2106 were from the parallel CTS8966 branch found from Armenia and possibly Levant.

As previously mentioned the best candidate for its affiliation is the Gutian connection. Even though its quite possible that over time he and similar people were assimilated in to locals, but their initial identity might be IE.

Its remarkable that Henning and after him Gamkrelidze and Ivanov proposed the IE origins of Gutians. Even though they linked it to Tocharian which is untenable in current stage of our knowledge.

Now with genetic data on hands lets look closer at Gutian names. 

At last three kings have the root Yarla in their name.

• Yarlagab
• Yarla
• Yarlaganda

In most likelihood the 'Yarla' is a Gutian deity name or a dynastic name while the rest are particles or other roots.

Yarla can be compared to Slavic deity Yarilo which has IE origin from *yeh₁r-. Its a Spring deity. The root is present in English in the form of word "year".

Another interesting root is the -gab. Its present in two names

• Sarlagab
• Yarlagab

It fits well with the PIE root ghabh- (also reconstructed as *gʰebʰ-). Which means to give, to hold. A frequent root in names. The English word to give is derived from that root.

Incidently the name Sarlak supports the fact that "Sarl" in Sarlagab is a root. Most probably a deity name. However this name is not from Sumerian king list so we are not sure anout its identity. He could be the same as Sarlagab ( or Zarlagab) 

Another possible root is the -gan present in 

• Tirigan
• Yarlaganda

Can be compared to PIE root *ǵenh₁. To give birth. Multiple names ending with -es or -us also supports the IE classification. The potential root Tir is also remarkable. It can be compared to Armeno-Iranian deity Tir.

Some other names are obvious Mesopotamian influence. Like Puzur-Suen. While the particle -um in some names is probably an Akkadian suffix.

Summing up the scant Gutian onomastic data we see evidence that can support their IE classification. However the potential IE language behind Gutian names seems to be somewhat different from the Proto-Armenian. It seems to be a centum, while Armenian is satem. It also preserved the phoneme [s] while Armenian converted them to [h] and lost them. What is also remarkable is that their Y DNA was widely present deep in steppe which once more could mean that the Gutian language was a (close?) parallel language to Armenian rather than a simple dialect of Proto-Armenian. 

With more ancient DNA we will have more clarity on all aspects of Middle Bronze Age migrations from steppe to West Asia.

A multidisciplinary approach for elucidating Mushki origin

A scientific article by Aram Palyan and Seb Peltkyan about the ancient people known as the Mushki, who probably had a Proto-/Para-Armenian origin.

In this article, the authors use a multidisciplinary approach to the question of the origin of the Mushki, relying, among other things, on paleogenetic data. Soviet researchers had already associated the Mushki with the Proto-Armenians. Thus, Igor Mikhailovich Diakonoff considered them Proto-Armenians who had migrated from the Balkans. This was based largely on the fact that later Assyrian sources used the name Mushki for the Phrygians as well, which gave rise to the theory that they had migrated from the Balkans through Anatolia toward the Armenian Highlands. However, the Mushki are first mentioned by the Assyrians farther east, precisely in the Armenian Highlands. Modern evidence instead supports a migration of the Mushki from the east — that is, from the Armenian Highlands toward Anatolia — where they apparently carried their ethnonym with them, which was later used by the Assyrians to designate the Phrygians.

The main arguments of the paper:

• The homeland of the Mushki was at the headwaters of the Euphrates during the Bronze Age. Their actual endonym was most likely *mosk'.

• Beginning around 1200 BCE, they migrated southward and southwestward, where they came into contact with the Assyrians and possibly with the post-Hittite kingdoms.

• Since the Phrygians are closest to the Greeks and unquestionably originated in the Balkans, the Mushki could not originally have been Phrygians, although by the 8th century BCE the term Mushki referred to the country of Phrygia.

• The reasons why the term Mushki became a name for the Phrygians are unknown, but there are many examples in history of such transfers of names, including France, Bulgaria, Palestine, etc.

• The Paleo-Balkan term musk, meaning “mule,” was most likely derived from the name of the Mushki. Both archaeological and paleogenetic evidence suggest that mules began to be bred in Anatolia and from there spread into Europe.

• The term Mushki has no clear etymology, but a number of strong arguments suggest that -k' is a suffix. Tribes and toponyms/onomastic forms with such a suffix are attested around the headwaters of the Euphrates and in northern Anatolia.

• The ethnic identity of the Mushki remains unclear. However, the presence of this suffix, together with the fact that during the Iron Age the genetic marker of the Trialeti-Vanadzor culture (I2a2b) migrated westward, supports the possibility that the Mushki may have been an Armenian tribe, although not Proto-Armenian in the sense previously proposed. A connection with the Kaskians, who inhabited northern Anatolia, is also possible.

• The migration of the Mushki and other tribes also left an archaeological trace in the form of grooved pottery (Grooved Ware, c. 1200 BCE).

• A Kartvelian origin is unlikely. However, the Georgian terms Mtskhe and Mtskheta (Samtskhe) may derive from the name of the Mushki. Genetic traces of a migration of the Mushki from the headwaters of the Euphrates northward into present-day Samtskhe in Georgia also do not support the Georgian hypothesis.

• The transition from mosk > mc'xe occurred in two stages and was probably completed by the 5th century CE. First there was the transition mosk' > moc', followed by the addition of the suffix -x. The Byzantine term Mesch is the Greek designation for the Georgian mc'xe.

• Ultimately, the Mushki (Greek Moschoi) ceased to exist as a distinct ethnic group with its own identity and were assimilated among Armenians, Georgians, and Cappadocians, where they left their mark in the form of mythical founding fathers.

Text from the Armenian DNA Telegram channel.

The original paper: An Examination of the Muški Tribe: A Multidisciplinary Approach Including Paleogenomics.

Common ancestor of Greek, Armenian and Albanian.

 Common ancestor of Greek, Armenian and Albanian.

In Petrosyan and Palyan 2023 we proposed a theory that the common ancestor of Greek and Armenian was the Catacomb culture. The rationale for this proposal was the date when steppe ancestry appears in respective regions. Around 2500BC in South Caucasus and 2300BC in southern Balkans Greece.

However the increasing number of ancient DNA from Balkans and Greece do not support this theory. Instead in Yediay et al. 2024 preprint the common ancestor of Graeco-Armenian was named the late Yamnaya period.

Yamnaya indeed had a massive expansion in Balkans after 3000BC. While Catacomb did not have apparently. See the maps.

Neverthless the Yamnaya culture did not reached central and southern Greece during the Yamnaya wave. The old Minoan and Anatolian farmer related ancestry was still present in EBA south Greece.

Proto-Greek alongside Phrygian and Paionian apparently formed in Bulgaria - North Macedonia region in Early Bronze Age as a result of mixture between Yamnaya and local Balkanian farmers. Later after 2300BC a new wave of migration moved to south Greece where the Mycenean culture formed after 1700BC.

This new version explains why Mycenean Greeks had a such diluted steppe ancestry. Because it was diluted twice. It also explains the strong presence of J2-Z6064 in Mycenean culture. This J2 was Euro-Anatolian farner lineage that integrated into the the emerging Proto-Greek community and expanded with them to central and south Greece. Albanian is another language that almost certainly descend directly from Yamnaya migration to Balkans. Even though they apparently had additional linguistic influences from later Thracian groups (E-V13).

This new version does not change much overall the picture. Except that the separation of Armenian and Greek is pushed back by some 200-250 years.

It also do not contradict to available linguistic data (Martirosyan, Kim etc) which will be discussed in a sepaeate topic.

Great news for understanding the Kura-Araxes culture legacy.

Great news for understanding the Kura-Araxes culture legacy.

Now that ancient DNA files from Bakr Awa are uploaded to FTDNA, we can make more precise conclusions.

The most important news is about G2b from MBA Bakr Awa (~1800BC). It is from the same branch as the G2b from Kaps in Armenia EBA. This G2b from Kaps was from a Kura-Araxes layer. Another Kura-Araxes G2b sample (I22907) is upcoming from Akbari et al. dataset. And multiple other G2b samples from later Lchashen period with typical Etiuni profile. You can see them on the first picture.

All this support the idea that G2b-FGC3032 was an important Kura-Araxes marker that moved to Zagros and reached Bakr Awa, which is marked by the blue dot on the second map showing sites with KAC pottery.

Given that Bakr Awa was an important Hurrian site in MBA this raises sharply the probability that Hurro-Urartian languages descend from Kura-Araxes culture. Or at last from southern parts of it. Other haplotypes supporting such a possibility are three cases of J1-PH128 from Bakr Awa connected to South Caucasus by the parallel branch. We don't have yet the proof that it was a KAC lineage but chances are quite high that it was. Another interesting haplotype is the J2-M92 in Achaemenid era Bakr Awa. The M92 was found in Kura-Araxes site from Georgia and in Levantine site Meggido in MBA also.

Combined with the possibility that Nakh-Dagestani people also descend from Kura-Araxes a legitimate question arise. If Kura-Araxes origin of Hurrians is confirmed, do this proves that Diakonoff and Starostin linguistic theory was correct. Well Kura-Araxes was roughly contemporary with Yamnaya. If those two linguistic groups descend from the same Proto language living in Early Bronze Age then other linguists also would have easily noticed their connection, like in the IE case. However that theory didn't get acceptance and recently Alexei Kassian completely rejected such a link.

Another possibility is that Hurro֊Urartian was language of farmers, which by the way is supported by their lexical parallels with Sumerian. While Proto-Nakh-Dagestani was a language of east Caucasian hunter gatherers that were shifting to food producing way of subsistence. In that case PHU and PND will not have recent common origin but will have occasions to contact each other and exchange words. Additionally the presence of Proto-Anatolians in Late Chalcolithic historic Armenia can explain the findings of Fournet and Bomhard about an archaic IE influence on Hurrian.

Ethnic Diversity in the Zagros Mountains: Part 1

 Ethnic Diversity in the Zagros Mountains: Part 1

Bronze Age Mesopotamian sources mention many ethnic groups and polities in the Zagros region, including the Hamazi, Gutians, Turukki (possibly the same as the Tukris), Lullubi, Kassites, Elamites, and others. Beside those listed, early Indo-Iranians must have also been present there.

In this group, I have already mentioned—based on a single sample—that the Elamites apparently had a high level of Iran/Zagros Neolithic ancestry. Therefore, their language could have originated from the Iran Neolithic. We have also discussed the possibility that the spread of Grey Ware (Dinkha Tepe II, after 1800 BC) in the Middle Bronze Age of northern Iran could be related to the Kassites and those unnamed Aryans.

Figuring out the genetic profile of the other mentioned groups is more problematic due to the absence of data related to their archaeology. However, in one case, a connection seems very likely—practically close to certitude. Some scholars have proposed theories that the Gutians were an Indo-European tribe ancestral to the Tocharians. Current data do not support the Tocharian link, given that the Tocharian origin is better explained by the Afanasievo culture.

Despite this, the idea of the Gutians as an Indo-European tribe is supported by the presence of R1b-Z2103 (Z2106?) in Middle Bronze Age Bakr Awa. R1b-Z2103 is a lineage that expanded in the Yamnaya culture and moved to the South Caucasus around the 2550–2500 BC period. From the South Caucasus, it moved further south toward the Zagros Mountains, and from there, they entered Mesopotamia. Cuneiform sources describe them as barbaric, destructive people who did not respect Mesopotamian gods and rituals. Some have proposed that they probably had light skin (namrum). However, modern scholars translate the meaning of namrum differently: as bright, shining, or healthy, of good quality for people or slaves. There are also reasons to believe that the Gutians elected their kings for specific periods. If true, this sets them apart from other West Asian societies where power was transferred hereditarily, and brings them closer to the ancient Greeks.

Only two dozen names are known from them, and not a single word. Thus, their linguistic classification remains obscure. In addition to these facts, I want to add that the term "Guti" was apparently a frequent endonym among various Indo-European peoples. The Tocharians called themselves Kuči, which could be derived from Guti based on regular phonetic shifts. The names of the Thracian tribe Getae, the Germanic Goths, and the Jats / Jutts in India could be related to this term. All of these parallels further reinforce the possibility that the Gutians could have originally come from the Yamnaya culture. Obviously, more ancient DNA is needed to further reinforce this theory, but at the current stage, the Gutians are clearly the best candidate for explaining the appearance of R1b in Bakr Awa. Note that this does not mean Hurrians completely lack any R1b; they could have had it, but they must have acquired it from neighboring IE tribes, meaning it was not native to them.

Later the term Guti/Quti has transformed in to catch all term denoting various people in North Zagros and Urartu region. The last mention of Guti is from the 6th century BC in Babylonian sources in the context of Persian expansion.

My next post will be about their possible language classification and the question of how it relates to Proto-Armenian, which also formed as a result of the same migration.