The genetic prehistory of the Greater Caucasus[preprint Harvard/Jena]

[Agamemnon]

There was a proposed so called Alarodian languages family but I don't know if this theory has the contemporary support after Sergei Starostin's death in 2005.

Yes, I am well-aware of the Alarodian theory. The main weakness with this theory is that it was put forth under the framework of the "Dené-Caucasian" and "Nostratic" macrofamilies, much of the evidence Starostin brought up in its favour was spurious to begin with. A genetic relationship between NEC and Hurro-Urartian deserves to be explored more seriously without the intention of validating higher-order macrofamillies such as Dené-Caucasian, going off the ancient genomic data odds are such a genetic relationship might be validated in much the same way Dené-Yeniseian was.

[Yaroslav]

It's doubtful P58 spread southwards with the KA culture. If we look at the phylogeny of the most successful branches under P58 (namely L862), KA does nothing to explain their spread. Unless P58 is basically absent throughout the Chalcolithic in the Levant and shows up in Khirbet Kerak, then a case might be made in favour of such an association. If anything, Z1842 in KA does a disfavour to any association between P58 and this culture.

I do not insist and quite admit that my thoughts could be wrong. I remember that Chiaroni et al. (2010) found P58's greatest diversity in Zagros and Tauros mountains of the Eastern Anatolia:

For J1e, the Zagros/Taurus mountain region displays the highest haplotype diversity, although the J1e frequency increases toward the peripheral Arabian Peninsula.

Interpolated J1e mean haplotype variance spatial distribution:


That was my thoughts' starting point. In this case I thought P58 had to expand southward from Taurus/Zagros. So to determine when that southward movement happened we should have a look when the Armenian, Iranian, Kurdish and Turkish P58 lines diverged with their southern Levantine and Mesapotamian P58 neighbors. Unfortunately I don't know well the P58 subclades but the ages of the above mentioned divergencies seemed to be of the Bronze Age period.

That was my logic.

Regarding the Bronze Age in the Arabian peninsula, you might want to read this post.

Thank you so much! I will read it with great interest.

[Yaroslav]

By the way VEK007 is on Gedmatch Genesis now, Kit Num: XD6258082

For example here are his MLDP K16 Modern 4-Ancestors Oracle:

Admix Results (sorted):

# Population Percent
1 Caucasian 52.90
2 Steppe 25.89
3 Neolithic 9.17
4 Indian 8.89
5 NearEast 3.15


Finished reading population data. 517 populations found.
16 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Chechen_Chechnya @ 9.057971
2 Abkhazian_Lykhny_Lykhny @ 10.258644
3 Adygei_Adygea @ 10.888059
4 Balkar_Kabardino-Balkaria @ 10.908051
5 Kumyk_Dagestan @ 11.142260
6 Georgian_Kakheti @ 11.926273
7 Ossetian_North_Ossetia @ 11.927305
8 Georgian_Svanetia @ 11.928255
9 Abkhazian_Gudauta @ 12.232993
10 Georgians_Zugdidi @ 12.978492
11 Georgian_Georgia @ 13.048143
12 Abhkasian_Abkhasia @ 13.108991
13 Georgian_Abkhazia_Abkhazia @ 13.300290
14 Adjar_Adjaria @ 13.451722
15 Lezgin_Dagestan @ 13.669104
16 Lak_Dagestan @ 13.987283
17 Azerbaijani_Dagestan_Azerbaijan_Dagestan @ 14.000269
18 Georgian_Megrelia @ 14.090484
19 Tabasaran_Dagestan @ 14.535103
20 Circassian_Circassia @ 14.718671

Using 2 populations approximation:
1 50% Chechen_Chechnya +50% Georgian_Kakheti @ 8.197195


Using 3 populations approximation:
1 50% Chechen_Chechnya +25% Chechen_Chechnya +25% Georgian_Abkhazia_Abkhazia @ 7.786840


Using 4 populations approximation:
+++++++++++++++++++++++++
1 Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya + Georgian_Abkhazia_Abkhazia @ 7.786840
2 Ava_Dagestan_Khunzakh + Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti @ 7.907899
3 Ava_Dagestan_Khunzakh + Chechen_Chechnya + Chechen_Chechnya + Georgian_Abkhazia_Abkhazia @ 7.932926
4 Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti @ 7.969103
5 Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti + Lak_Dagestan @ 7.993087
6 Chechen_Chechnya + Chechen_Chechnya + Georgian_Abkhazia_Abkhazia + Lak_Dagestan @ 8.020758
7 Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti + Tabasaran_Dagestan @ 8.021328
8 Chechen_Chechnya + Chechen_Chechnya + Georgian_Abkhazia_Abkhazia + Tabasaran_Dagestan @ 8.049325
9 Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti + Lezgin_Dagestan @ 8.117599
10 Ava_Dagestan_Khunzakh + Chechen_Chechnya + Georgian_Kakheti + Georgian_Kakheti @ 8.127557
11 Chechen_Chechnya + Chechen_Chechnya + Georgian_Abkhazia_Abkhazia + Lezgin_Dagestan @ 8.129585
12 Armenian_Gavar + Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya @ 8.157759
13 Armenian_Yerevan + Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya @ 8.163782
14 Assyrian_Turkey + Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya @ 8.184855
15 Armenian_Martuni + Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya @ 8.185712
16 Chechen_Chechnya + Chechen_Chechnya + Georgian_Abkhazia_Abkhazia + Georgian_Kakheti @ 8.194045
17 Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti + Georgian_Kakheti @ 8.197195
18 Azerbaijani_Dagestan_Azerbaijan_Dagestan + Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti @ 8.199531
19 Chechen_Chechnya + Chechen_Chechnya + Georgian_Kakheti + Kumyk_Stalskoe @ 8.200113
20 Jew_Azerbaijan + Chechen_Chechnya + Chechen_Chechnya + Chechen_Chechnya @ 8.207031

[royking]

Ted Kandell has found two M67-CTS900 samples in Maykop---even the Eneolithic sample. This suggests that M67 and its sister M319 could convey the NW Caucasian languages to Anatolia and even eastern Crete during Minoan era. The Z1842 fits well with my theories about Kura-Araxes cultures bringing this lineage to the Levant. P58 is a bit of a mystery--maybe residing in North Mesopotamia and moving into the Levant during the EBA to MBA.

[Generalissimo]

Both of them say the same thing though.

"The northern Caucasus groups are genetically similar to contemporaneous populations south of it, suggesting human movement across the mountain range during the Bronze Age."

But not into the steppe, which is what the first abstract was wrongly suggesting.

[Principe]

Ted Kandell has found two M67-CTS900 samples in Maykop---even the Eneolithic sample. This suggests that M67 and its sister M319 could convey the NW Caucasian languages to Anatolia and even eastern Crete during Minoan era. The Z1842 fits well with my theories about Kura-Araxes cultures bringing this lineage to the Levant. P58 is a bit of a mystery--maybe residing in North Mesopotamia and moving into the Levant during the EBA to MBA.

Roy is it possible to see which branches of CTS900 that were found and we can see the results for all the J2a?

[smal]

PG2001 Progress 2 Eneolithic steppe R-V1636
PG2004 Progress 2 Eneolithic steppe R-V1636

[smal]

RK4001.A0101 Rasshevatskiy 4 Catacomb R-Z2108

[Trojet]

I just checked I6266 Klady 5200.0 Maykop J2a

He is positive for two SNPs at the J2a-Y11200 level, namely Y11861+ and Y18325+. Downstream he has no calls for any SNPs at the J-Z30677 level, and one negative SNP at the J-Y30811 level which is Y30811- itself. Also one negative SNP at the J-Y17946 level which is Z30682-

[Agamemnon]

If the "Alarodian" language family is now on solid ground due in part to this ancient DNA, does this (by process of elimination) aid in placing Caucasian and/or Kartvelian into broader language families? NW Caucasian w/Hattic? Kartvelian with some of the Neolithic Aegean languages?

Not exactly. For all we know, K-A might well've been multilingual for example. We should expect at least one of the G2b samples in this study to be on the same branch as the G2b sample from Wezmeh cave, while the G2a sample from Maykop is only remotely related to the branch of G2a most common in NW Caucasian and Kartvelian-speaking groups, that being said it might have something to do with NE Caucasian-speakers to some extent. Kartvelian might have more to do with Hattic than NW Caucasian, for what it's worth (not much, admittedly) both of these are typologically more similar to each other than they are to NWC. The Pre-Greek substratum could be related to Tyrsenian, there were at least two unrelated non-IE languages in the Aegean and it isn't at all certain the Pre-Greek substratum was related to Minoan or Eteocretan.

The biggest obstacle in the Caucasus is that the language families there, despite sharing some common features, are extremely divergent. It isn't at all obvious that they are related to other languages, let alone to each other. For example, NWC is characterised by an extremely complex verbal morphology coupled with agglutination, an extremely large consonant inventory (the largest out of all non-African languages) and an extremely small vowel inventory (as low as one); NEC on the other hand is characterised by an extremely rich and complex case system (some of these languages have dozens of cases, Tsez has more than 60, Tabasaran more than 50, etc), a complex consonant inventory and an equally complex vowel inventory (even the simpler Nakh languages, like Chechen, have so many vowels that they defy linguistic transliteration).

And of course, the main issue with languages like Hattic or Hurrian is our less-than-perfect comprehension of these languages, it honestly isn't a stretch to say we know more about PIE than many of these languages. So proving a genetic relationship between these languages is a pretty daunting task.

I do not insist and quite admit that my thoughts could be wrong. I remember that Chiaroni et al. (2010) found P58's greatest diversity in Zagros and Tauros mountains of the Eastern Anatolia:



Interpolated J1e mean haplotype variance spatial distribution:


That was my thoughts' starting point. In this case I thought P58 had to expand southward from Taurus/Zagros. So to determine when that southward movement happened we should have a look when the Armenian, Iranian, Kurdish and Turkish P58 lines diverged with their southern Levantine and Mesapotamian P58 neighbors. Unfortunately I don't know well the P58 subclades but the ages of the above mentioned divergencies seemed to be of the Bronze Age period.

That was my logic.

I would not place too much stock in present-day diversity, ancient DNA has too often refuted such approaches, R1b and R1a arguably being some of the best examples. That's not to say P58 did not come from the Zagros or Upper Mesopotamia, this remains an open question and what we do know so far does indicate a north-to-south spread. The real problem with an arrival with Kura-Araxes is that the entire phylogeny of J1-L862 (and other branches with a similar distribution) points to major founder effects (what this study calls "star burst expansion") during the Chalcolithic period and the EBA, the distribution of J1-L862 both in ancient and present-day contexts strongly suggests some sort of correlation with the spread of Semitic speakers, Common Semitic unity came to an end during the first half of the 4th millennium BCE... All of this is unlikely to be coincidental, and K-A does nothing to explain the above, especially since the earliest K-A material in the Levant dates back to the EB III period (~2700 BCE), in other words roughly a thousand years after the major founder effects.

And then there is also the fact that according to Harney et al., Levant_BA_South may "represent a remnant of a population that formed after an initial spread of Iran_ChL-related ancestry into the Levant that was not affected by the spread of an Anatolia_N-related population, or perhaps a reintroduction of a population without Anatolia_N-related ancestry to the region." If true (and there is good reason to expect precursor to Levant_BA_South in the Chalcolithic Southern Levant), J1-L862 is one of the most obvious candidates for the introduction of Iran_ChL-related ancestry.

Ted Kandell has found two M67-CTS900 samples in Maykop---even the Eneolithic sample. This suggests that M67 and its sister M319 could convey the NW Caucasian languages to Anatolia and even eastern Crete during Minoan era. The Z1842 fits well with my theories about Kura-Araxes cultures bringing this lineage to the Levant. P58 is a bit of a mystery--maybe residing in North Mesopotamia and moving into the Levant during the EBA to MBA.

Fascinating, Y3612, which is by far the most common lineage among Nakh-speaking populations, is a branch of CTS900. This does reinforce the perception that J2a-Y3612 men adopted Northeast Caucasian speech and that J1-Z1842 is the diagnostic Proto-Northeast Caucasian marker (already obvious judging from the paucity of J2a-Y3612 in non-Nakh groups).
I definitely agree regarding Z1842, it probably arrived in the Levant with KKW. I'm much more skeptical of an arrival of P58 during the EBA IV/MBA I, admittedly this is even less likely than an arrival with Kura-Araxes (which is unlikely for the reasons I spoke of above).

PG2001 Progress 2 Eneolithic steppe R-V1636
PG2004 Progress 2 Eneolithic steppe R-V1636

I1635, the K-A sample from Armenia, also was R1b-V1636.