Corded Ware and Yamnaya

Well, Davidski at Eurogenes has made a bold prediction, which I'm curious to see falsified or bolstered by more sampling.  His suggestion is that the model for the spread of Indo-European languages which I've espoused here is unnecessary; most of the extant Indo-European languages (and by extension, the Indo-European peoples) can actually be derived directly from the Corded Ware horizon (plus whatever locals they superimposed themselves over while doing so) and the specifically Single Grave variant is the source for Western Europe, not Yamnaya derived Bell Beaker cultures.  This is based on two key assumptions:

1. At a genome wide-level, the Corded Ware and Yamnaya are extremely closely related.  His contention here is that that Y-DNA sorting between an R1a and an R1b patrilineal group is either an artifact of insufficient sampling, or simply a much less important data point than the genome wide statistical comparison.  Utilizing this genome-wide comparison, even the eastern and central European Bell Beakers, who are credited with being the source of R1b lineages in the North Sea region (famously, those who settled Britain in the EMBA period) can be derived from the Corded Ware and don't require an external stimulus.
2. And even if it did, the usual suspects are suggesting that they provide the wrong specific clade of R1b.  The Carpathian Yamnaya extension is most likely the source, not of an overlay over Corded Ware that creates the the northern European I-E languages, but rather, of the southern Balkan languages and spread even further south and often eastward, such as Thracian, Illyrian, Phrygian, Greek, Armenian, etc.

Now, to be clear, what he's saying is that it looks like the R1b haplogroups that are today common in western and northern Europe cannot be derived from the Yamnaya expansion into the Carpathian basin.  Let's be careful that we don't load that claim with too much baggage.  And he's also admitting that it's a somewhat speculative assertion at this point, because there isn't enough data yet to confirm or refute it.  But he is making a prediction, based on what little data we do have, about what he expects more data to reveal.  And that prediction is that the Dutch Bell Beakers, who are the source of the later British Bell Beakers and probably the Nordic Bronze Age a little bit later, don't necessarily require an overlay of Yamnaya over Corded Ware, but rather Corded Ware over locals is sufficient to explain them.  

Of course, population movements are not so binary, so when we talk in generalities, it doesn't require that every individual conform to it, merely the preponderance.

Anyway, here's a few interesting comments from the thread where this prediction was made.  I leave these here without further comment from me, other than to note that this proposes a very different model than the one that I've seen online from most folks who are doing up-to-the-minute discussions, which consider Corded Ware to have been a satem language descended from the western half of the steppe (Sredni Stog II) with R1a Y-DNA haplogroups, and eastern Bell Beakers to be the classic example (among many other cultures of Europe that were related to it) to have been the source of centum languages, descended from the eastern half of the steppe (Repin and Khvalynsk > Yamnaya) and heavily R1b.  Although, that's the gist of why it's a bold prediction, isn't it?

This new set of proposed movements is extremely difficult to reconcile with linguistic isoglosses and the various trees for PIE. In general, the satem languages and Greek form an innovative core compared to Italic, Celtic, Tocharian, Anatolian, all of which left earlier. A derivation of those which left earlier directly from Yamnaya, through the Balkans and Bell Beaker in Western Europe for Italic-Celtic and other Western IE dialects (e.g. Ligurian, Venetic) and Afanasievo into Asia for Tocharian, plus a derivation of the innovative set of Indo-Iranian, Greek and Armenian from R1a and EEF containing Corded Ware, seem to make good sense of the genetics and the linguistics. If Western IE ancestry also derives from Corded Ware, we don't have much to explain the distribution of features among the IE languages.
Any proposed set of movements have to explain the core-periphery distinction described above, plus the following features:
1. Anatolian shares a few features with Western IE languages
2. Italic and Celtic must be close to each other until they split
3. Greek is partly in the innovative group and partly in the periphery
4. Germanic is like Greek, but is close to Celtic throughout its history as well
In particular, if Celtic, Italic, Germanic, all the way to Indo-Iranian all drive from Corded Ware peoples, how can it be that Bell Beaker is responsible for a single language family, Celtic, while Corded Ware--not much older than Bell Beaker--is responsible for a huge diversity of IE branches? The temporal chronology doesn't make much sense to me.

 ~~~

I recently believed this as well, but I changed my mind after reading Chang et al 2015.It looks like the innovative core was the centered on the steppe area. As waves of IE cultures spread from the steppe to surrounding environments, they lost contact with the core and thus didn't share in the innovations.In the example of the Satem isogloss, it looks like Corded Ware was originally Centum, but parts of it were Satemized later in a new wave emanating from the steppe, around the time of Sintashta. This wave of innovations spread to some areas adjacent to the steppe, like the Balkans and Eastern Europe, but cultures at the periphery of the steppe zone like proto-Germanic and proto-Greek were only partially influenced and more distant groups like Celtics, Italics, and Tocharians escaped the influence entirely. It does seem that the composition of steppe groups changed from Yamnaya-like Yamnaya, Poltavka and Catacomb groups to Corded-ware-like Sintashta, Srubna and Andronovo groups, and the later set is associated with Satem isogloss and other innovations.

~~~

Perhaps some parts of CW were centum, and that explains everything, but that basically means Corded Ware gave rise to all the languages that are part of late IE outside Anatolian and Tocharian, while BB gave rise to only Celtic, and maybe some para-celtic groups like Ligurian. How can that be when both groups are very little separated in age? Does this make sense to you?

 ~~~

If we accept what Davidski (and others) have postulated, the scenario looks something like the following if you try to correlate the Chang tree with archaeogenetic evidence:

Pre-Yamnaya>>Anatolian
Circa Yamnaya>>Tocharian
CW>>Italic, Celtic, Germanic, Balto-Slavic, Indo-Iranian, (Daco-Thracian?)
Western Yamnaya>>Greek, Armenian, Albanian, (Phrygian?)
Some of these branches influenced each other after initial divergence, which would explain why there are isoglosses shared across some of these branches that remained contiguous with each other.

 ~~~

It should have been obvious long ago that most Indo-European languages have their roots in the Corded Ware complex, considering how similar Sintashta was to Bell Beakers, and how different they both were from Yamnaya.
There's been way too much focus on Y-haplogroups and their phylogeny by most, resulting in some wayward assumptions being popularized.

 ~~~

The Balkans are the only region of IE Europe today not dominated by late clades of R1a or R1b, and, judging from the aDNA record, did not have extreme male-biased replacement right away after contact with Steppe cultures and reached its current levels of steppe ancestry much later compared to Western, Northern or Eastern Europe. The EEF elites interred in kurgans mentioned by David are pretty indicative. The models of acculturation and integration may have more traction here than elsewhere.