More on Corded Ware vs Yamnaya

More on this topic, which I find fascinating.  From the commentary at Davidski's blog, let me quote just a bit of the discussion:

Michal said, first:

There is simply no way neither for Corded Ware alone nor for Yamnaya alone to have been PIE-speaking. All data indicate that these were closely related populations speaking different Post-Late PIE dialects, so when looking for a hypothetical Late PIE source, we should direct our attention towards the pre-Yamnaya populations on the steppe, with Khvalynsk and Sredny Stog being the two best candidates for the moment. 

Both Anatolian and Tocharian are lacking the wheeled wagon vocabulary. The Tocharian and Anatolian words for "wheel" derive from a PIE lexeme (*h2werg-), which is different from the corresponding words attested in languages descending from Late PIE (where the word for wheel is derived from either *kwel- or *Hrotós). Also, Tocharian does not share the Late PIE-derived words for wagon, axle and thill, so the only Tocharian word indicating the shared vocabulary for wheeled vehicles is kokale/kukäl (chariot), showing analogy to a Greek word for wheel (kúklos), which, however, doesn't seem to be enough to claim that Tocharian and Late PIE diverged only after 4000 BC. Thus, it seems almost certain that not only Archaic PIE (Proto-Indo-Hittite) but also Early PIE (ancestral to Tocharian and Late PIE) were spoken before 4000 BC, and most likely before 4500 BC. 

Corded Ware was associated mostly with R1a-M417 and likely ancestral to both the Balto-Slavic and Indo-Iranian speaking populations, while the Post-Late PIE dialect spoken in Yamnaya was associated with R1b-Z2103 and most likely ancestral to populations speaking Greek, Armenian and Albanian (plus Daco-Thracian, Mycenaean and Phrygian, when counting the extinct languages that are currently known to us). 

Another group of Post-Late PIE languages, including Italo-Celtic and possibly Germanic, seems to descend from a Post-Late PIE dialect spoken by the R1b-L51 people. As for now, it is hard to say how this group was related to Corded Ware and Yamnaya and when exactly their ancestors moved from Eastern to Central Europe. Also, it is not clear how and where that ancestral R1b-L51-rich population contributed to the emergence of the Bell Beaker culture and whether the R1b-U106 folks were still part of that group at that very time.

Followed by this response from Davidski:

I appreciate your insights, but things move quickly in this area, and you're saying things that I would've said a year or two ago, but no longer. 

One of the problems with your assumptions is that Z2103 is found in steppe populations other than Yamnaya, and you'll be seeing it pop up in many more pre- and post-Yamnaya samples. So how do you know that the language spoken by Yamnaya was ancestral to Greek or Albanian, when the Z2103 in Greeks and Albanians might be from other steppe or steppe-derived populations? 

And, of course, there's no reason why Z2103 should be tied to the same language family in Yamnaya, Catacomb, Dereivka, Babino, Vucedol etc. 

As for the question of the origins of L51, P312 and U106, I'm guessing you're aware of my view that they were somehow involved or caught up in the Corded Ware expansions, so if that pans out, then it will seriously leave Yamnaya out in the cold.

And Archi's response to the last paragraph by Michal:

These peoples are associated with the Babino culture (and its derivatives) which took the place of Yamnaya-Catacomb cultures. The Babino (Multi-Cordoned Ware) culture comes from Corded Ware which captured the substrate Catacomb culture.

Synome's response to that idea:

Working just from the Chang linguistic tree and our current genomic evidence, I would agree with Davidski that there are very few existing languages that are likely to be derived from Yamnaya itself, maybe Tocharian and Albanian. 

Greco-Armeno-Phrygian would be more likely to descend from a post Yamnaya culture on the PC steppe, like KMK.

Note that KMK, Babino and Multi-cordoned Ware Culture are all alternative names for the same archaeological material culture.

Michal's response to all of this:

@Synome
When knowing the commonly accepted positions of both Albanian and Tocharian in the IE tree, it is simply impossible to believe that these two langauges descend from one ancestral Proto-Albanian-Tocharian language spoken in Yamnaya. 

@Archi
By 2000-1500 BC every population in the Near East knew the wheel, yet it doesn't mean that they all spoke a language descending from a Proto-Indo-Hettite dialect. Also, the Anatolians and Tocharians used a very different word for wheel when comparing them with populations speaking the remaining IE langauges that all descended from Late PIE, thus from a language spoken only after the Anatalion and Tocharian languages diverged from the branch ancestral to all remaining IE languages. 

You mean the Armenians descend from the Babino culture? I strongly doubt this. There is no doubt that the Late Yamnaya people who stayed on the steppe were very strongly influenced by the Corded Ware-derived populations moving in the NPC steppe, and this could have been the major reason why ancient Greek and Sanskrit share so many similarities, so some linguists were confused by this and classified Indo-Iranian and Graeco-Phrygian as more closely related to each other than to Balto-Slavic, while in fact this only seems to reflect the relatively late interactions between the not-so-distantly related Post-Late PIE dialects spoken on the steppe in the Post-Yamnaya period. A very similar mechanism was likely engaged in modifying the Yamnaya-derived languages (Dacian and Thracian?) by the Iranian-speaking people from the North Pontic steppe. 

@Davidski
I must admit I don't understand your objections. First, if Z2103 is found in a Pre-Yamnaya population on the steppe, how will it discredit my assumption that the Yamnaya folks spoke a Late IE dialect (let's call it Greaco-Armeno-Daco-Thracian)? Also, if Z2103 is found in some neighbors of Yamnaya after this culture expanded west and east, how will it make my scenario implausible? There were of course some Yamnaya-descending populations rather than Yamnaya itself that were repsonsible for spreading both Z2103 and IE languages to such destinations like Albania or Greece, yet this doesn't make their Yamnaya origin less important. 

Additionally, I find it almost impossible to believe that it was exclusively some Corded Ware-derived waves of people that were initially responsible for spreading very different IE branches (including Tocharian, Anatolian and some Late PIE-derived branches, like Albanian, Greek and Armenian) in different directions, as this does not fit neither the IE tree nor the Y-DNA data we have. In other words, there is absolutely no correlation between the genetics and linguistics when assuming that only Corded Ware was associated with PIE. 

What are the Corded Ware-derived Y-DNA lineages that are supposed to be associated with the Anatolians, Tocharians, Greeks, Albanians and Armenians in your scenario? Do your choices make any sense from the phylogenetic and linguistic point of view? When exactly those ancestral populations are supposed to have left your hypothetical Corded Ware-relatd PIE homeland? 

Finally, I am not sure if this is what you propose, but are you suggesting that there were some non-Yamnaya Corded Ware-derived Z2103 lineages that were responsible for spreading the IE languages south of the Corded Ware territory? If so, what specific Z2103 lineages would you have in mind?

And Davidski's response to that response:

My point was that you can't link Yamnaya linguistically to any Indo-European speaking groups via Z2103, because Z2103 wasn't exclusively a Yamnaya lineage. And even if the Z2103 in some Indo-European groups ultimately came from Yamnaya, this doesn't mean that their languages did also. 

Indeed, it's generally agreed that Armenian spread into the south Caucasus from the Balkans well after Yamnaya was gone. So even if many of these hypothetical Proto-Armenians carried Z2103 lineages ultimately derived from Yamnaya, they may have got their language from somewhere entirely different. But anyway, ancient DNA tells us that they may have carried I2c. 

And I never claimed that Corded Ware spoke archaic PIE, nor that it spread Anatolian languages to Anatolia. 

But I suspect that a kurgan group from the western edge of the steppe closely related to Corded Ware was somehow involved in this process.

So there you have it.  The notion is out there that none of the branches of Indo-European that we know of are descended from the Yamnaya culture, even Anatolian and Tocharian.

Of course, even if that were true, that doesn't mean that they didn't speak a closely related language, merely that it left no descendants.  At best, you could call it a para-PIE language, most likely it's merely a dead-end dialect.  There is still a lot of confusion about the following, for example:

• where exactly did the Anatolian branch, the first to split off, come from and who spoke it?  What was their route into Anatolia, and did they originate on the boundary between the western steppe and Old Europe, say, in the Usatovo culture or some similar horizon, as Anthony proposed in the past?   (This is probably still assumed; either way, Anatolian probably breaks off from the western steppe prior to the formation of Yamnaya proper, which is an eastern development which spread westward to encompass the entire steppe.  Anatolian is not presumed to be a Yamnaya descendant even in the prior model.)
• For that matter, what exactly is the genesis of other paleo Balkan Indo-European languages, including the ancestors of Armenian, Greek, Phrygian, Albananian, and of course, languages that are poorly known but only from the Balkans such as Thracian, Dacian and Illyrian?  We see both Yamnaya derived and later post-Corded Ware derived proposals in the discussion above.
• What was the genesis of the Tocharian languages and does the Afanasievo culture have anything to do with them or not?  The recent spate of DNA from the Tarim basin was not helpful in resolving this, because of a lot of issues: Tocharian isn't attested until quite late, for instance, but it is presumed by linguistic means to have split off from PIE very early; the first, in fact, other than Anatolian, to do so.  The Anatolian area was also swept over by Andronovo variants (as well as other populations from the east) prior to the attestation of the languages, so what would be needed is some kind of direct link discovered between Afanasievo and later Tarim populations, but not so late that they are confused with Andronovo variants.  And as Davidski himself point out a month or two ago, the so-called Tocharian samples that were published are almost certainly actually Huns, not Tocharians anyway.
• There's a six hundred year gap between the end of the Sredni Stog and the beginning of the Corded Ware horizon.  Much of the territory where the transition from Sredni Stog to Corded Ware could putatively have happened is occupied by other culture that seem to be either non-steppe, or at best, mixed steppe and Old Europe, like Globular Amphora, Baden, Ezero, etc.  The actual origins of the Corded Ware, with 75%+ steppe DNA but not Yamnaya DNA are a major missing link still.  Granted, the state of Ukrainian archaeology is pretty bad, which is a major source of the uncertainty.  Better field work should probably help develop some clues to the direction of Sredni Stog expansion into becoming Corded Ware, if that is indeed what happened.

Another interesting aside is that my objections to the parsimony of the physical appearance of the Yamnaya people, as reported, vs their alleged descendants across the board is not really an issue anymore if they aren't actually descendants of Yamnaya, but rather of Corded Ware.  Not that I don't still think the Yamnaya descriptions are probably subject to sampling bias (and other bias, for that matter) that makes the reports somewhat suspect when applied to the population overall, but it actually doesn't really matter either.

I still suspect that the Yamnaya people are the genesis of the Tocharians through Afanasievo, and I suspect that they are a substrate of later appearing steppe cultures, like Srubnaya, Catacomb, etc., if not the direct ancestors of some of them, although given the inherent violence of the time and the rabid extinctions and blossoming of y-DNA lineages, they may have contributed many more women than men to the populations that followed them.  Certainly by the time Sintashta and Andronovo appeared, the directly Yamnaya descended cultures were gone, replaced by some who's primary ancestry (on the male lines, at least) were Corded Ware.