Wednesday, October 17, 2018

Lining up Anthony and genetics

OK, I guess yesterday's post wasn't quite my last word (for now) on genetics after all.  Sigh.  Here I am again trying to lay out David Anthony's model with some genetics.  Most of this stuff is in 1) his book, The Horse, the Wheel, and Language and 2) the Wikipedia summary of his book, which is where "Revised Kurgan Theory" redirects you.  But I'm interested in seeing how well (if at all) it lays out with genetics.  Granted, we don't have genetic data for a lot of these archaeological material cultures.  But let's see if I can lay this out in a way that's less text dense and makes sense to me personally.
  • Any material cultures that are earlier than 4,000-3,500 BC can't be Proto-Indo-European, or even proto-Proto-I-E, because there wasn't any animal husbandry, which is a necessary component of the Indo-European cultural package.  The two cultures that precede this, and which first start to pick up some animal husbandry late in their development (although they are initially hunter, gatherer and fisher cultures) are Dnieper-Donets I (5,800-5,200 BC) in the West and Samara in the East on the Volga bend area.
  • In the Balkans and Danubian region, the Criş (5,800-5,300 BC) is an EEF culture that first brings cattle to the region.  The Bug-Dniester culture (6,300-5,500 BC), local hunters and foragers and fishers from the eastern Balkans (using the term rather broadly) was heavily influenced by the arrival of the Criş and was the "membrane" through which the Neolithic Revolution came to the steppes, specifically to the Dnieper Rapids area, and the Dnieper-Donets culture.  These steppes are still too early to be called anything like PIE and have it make sense linguistically, although no doubt either the Dnieper-Donets and/or the Samara cultures (at least) spoke languages that later evolved into PIE.
  • Several Balkan EEF and mixed cultures (that are presumably not proto-Proto-Indo-European) cultures; Criş, Vinča, Bug-Dniester, etc. developed into the Cucuteni-Trypillian culture (really more of a horizon than a culture per se) (5,200-3,500 BC), which actually spread eastward and represents the greatest expanse of paleo-Balkan Old Europe; it was densely populated and had what were probably the largest cities on Earth at the time; bigger than anything in China, or Egypt, or the Middle East, etc.  Dnieper-Donets I evolved into Dnieper-Donets II (5,200/5,000-4,400/4,200 BC) which was now a cattle herding economy instead of hunting/fishing, and cattle, sheep and goats migrated as far eastward (at least) as Khvalynsk in the Middle Volga region (4,700-3,800 BC), which was the successor to and evolved from of the Samara culture.  At this point, Dnieper-Donets II and Khvalynsk probably correspond to the R1a and R1b populations (respectively) that later emerge in the Corded Ware and Yamnaya horizons, although we don't have enough samples to say that that's true, only that we would expect that.  Somewhere in that steppe spectrum was probably spoken a language that was getting close to a very early Indo-Hittite.
  • Somewhere around 4,200 BC, the horse was domesticated across the steppe, although it didn't appear in large numbers until 3,500 BC.  It's not clear exactly how and by whom this happened.  A lot has been made of the almost certainly non-Indo-European Botai culture from Central Asia and their horse domestication (or husbandry, or whatever exactly they did with horses) but curiously, the Botai seem to have contributed little if any genetic or cultural input into the more westerly steppe cultures, and the horses themselves seem to be the forerunners of the Mongolian wild horses, and contributed no genetic material to today's domestic horses.  Most likely, somebody out in the Khvalysnk culture can be credited with independently getting the ball rolling in domesticating the horse that led to today's modern horse.
  • At some point here, probably under an elite dominance or influence, at least, some Khvalynsk "chiefs" or cult figures, or other high status individuals of some kind turned the Dnieper-Donets II into the Sredny Stog culture on the western steppes (4,400-3,300 BC).  Marija Gimbutas saw the Sredny Stog as the core of early PIE, and based on the very few samples we have genetically, it is probably R1a in terms of Y-DNA and probably the direct ancestor of Corded Ware (as well as an important input into other cultures as well).  Exactly what the relationship between Sredni Stog and Khvalynsk is is still debatable.  Some more genetic testing would be welcome. Personally, I expect that Khvalynsk raiders or an elite warrior caste, or something like that, came to dominate the Dnieper-Donets II peoples and spurred the evolution into Sredni Stog.  Whether they brought an early version of Indo-Hittite with them, or if the Dnieper-Donets people already spoke a different dialect of it themselves is TBD.
  • Somewhere around 4,200-4,100 BC the climate deteriorated, as something not unlike the Little Ice Age struck Europe (previously, the Neolithic climate had been similar to the Medieval Warm Period.)  This made the steppes cooler and drier, and had a major impact on the arability of the farmland of the Old European Cucteni-Trypillian culture.  Settlements were burned, abandoned, fortified, etc. and evidence of migration from Sredni Stog into the Balkans starts around 4,200-4,000 BC with the formation of the Suvorovo-Novodanilovka culture.  Anthony sees this culture as not a mass migration and population replacement, but more of an example of elite dominance.  The S-N culture probably spoke an archaic Indo-European language, or Indo-Hittite language, and the Anatolian languages which later emerge like Hittite, Luwian, Palaic, etc. are assumed to come from this very early separation from the PIE steppe homeland.  I'd really like to see some archaeogenetic data that backs up this connection of Sredni Stog => Suvorovo-Novodanilovka => early Anatolian speakers in Anatolia, because as far as I know, there isn't anything much to speak of yet.
  • The next phase also really needs some genetic confirmation.  According to Anthony, this is where "classic" PIE develops over the course of several broadly related material cultures that were all in contact with each other, but which also retained a fair bit of regional variation (and of course, we know that there was still Y-DNA differentiation here too.)  There is substantial influence and trading (and probably at least some intermarriage and cultural diffusion too) with the Caucasian Maykop culture (3,700-3,000 BC), which was almost certainly not Indo-European, and may have been an early proto-Hurrian/Hattian speaking people.  But they were a major source of metallurgy, and and class differentiation, as well as the source of the kurgan.  They were also a vector for diffusion of some Middle Eastern cultural influence.
    • The Mikhailovka culture (3,600-3,000 BC) was the westernmost of these, and seems to have had substantial cross-pollination with non-Indo-European Trypillian cultures.  Not only do steppe-type potteries appear more frequently to the west, but Balkan/Mediterranean physical types appear more frequently among this culture.  More genetic evidence is desired.  It does maintain cross-steppe contact, however—late in this phase, Repin style pottery from far to the Volga/Don east is found here still.  Kemi-Oba is a Crimean variant of this culture.
    • Post Mariupol Early (3,800-3,300 BC) and late (3,300-2,800 BC) with a closer resemblance to the Suvororo-Novodanilovka peoples on the Dnieper Rapids area.
    • Late, or Phase II Sredny Stog, 4,000-3,500 BC on the Dneiper, Donets and western Don valleys.
    • On the northern Don region, Maykop contacts helped spur the regional Repin variant (3,950-3,300 BC) from a late Khvalynsk variant, although Khvalynsk in its more "pure" form continued on the Volga area until at least 3,800 BC, and probably longer.  At some point, the Afansievo culture splits off from the Repin culture and heads further east to the Altai area (3,700-3,300 BC.)  It's worth noting that the Afanasievo culture has been determined to be indistinguishable from the Yamnaya genetically, although we're getting to Yamnaya in the next bullet point.  Apparently, it made this rather epic journey across the steppes without any genetic admixture to speak of, and presumably took with it the fledgling proto-Tocharian languages, usually seen as the second most conservative and archaic, after Anatolian.
  • Between 3,500-3,000 BC there was further climatic deterioration, and this was especially hard on the eastern (and therefore drier and cooler already) steppes.  Late Repin and Khvalynsk turned into early Yamnaya on the Don and Volga regions, but it is really more of an economic change than a pottery change—a move to full-blown pastoral nomadism, wheeled wagons as houses, and only a few heavily fortified permanent settlements.  Stock raiding, warfare, elite dominance, The Patriarchy™, etc. all presumably evolved at this point as a reaction to the climatic changes, but also created a culture capable of sweeping into new territory, either because it had been depopulated due to climate change, or just because they were better at dominating culturally and providing economically than the EEF type folks that were there before.  The Yamnaya spread quickly across the steppe, from 3,300 BC or so when it was "born" out of Repin and Khvalynsk to 3,100 BC or so when it dominated the entirety of the steppes.  What happened to the peoples already living on the western steppe?  Were they culturally dominated and assimilated from (admittedly, already similar) cultures into the Yamnaya sphere, or displaced?  Anthony believed the former, but the genetics which have come out since then suggest the latter.  In any case, the Yamnaya horizon was long-lived, and lasted until 2,500 BC; a good 6-7 centuries of regional dominance over the classic Pontic-Caspian steppe territory.  What else was going on other than Yamnaya, though?
    • The Usatovo (3,500-2,500 BC) culture appears to be a Trypillian culture that was dominated by steppe elites and semi-transformed into a pseudo-steppe culture between the Dniester and the Vistula.  Anthony makes some vague references to pre-Germanic coming from this milieu, but that doesn't make sense to me.
    • A massive migration event seems to have happened between 3,100 BC and 2,800-2,600 BC or so into the Danube valley and further west.  A number of possibly Indo-Europeanized cultures appear during this period; Baden (3,600-2,800 BC), Globular Amphora (3,400-2,800 BC), Vučedol (3,000-2,200 BC).  Anthony suggests pre-Celtic dialects spread with the Beaker culture (2,800-1,800 BC) which spread from this area into Austria and up to the Rhine and eventually to the British Isles and the Iberian peninsula (its relationship with parallel developments going on on the Atlantic coast are unclear.)  Anthony suggests much later Urnfield and even Iron Age Villanovan cultures as the spread of the Italic languages (assuming that there is in fact a genetic Italic branch to begin with.)  It's worth noting that most accept the Urnfield as pre-Celtic, on its way to becoming the Iron Age Hallstatt culture which was definitely proto-Celtic, and the La Tene which was historically Celtic.  It's also worth noting a few things; this doesn't explain the arrival of Celtic to the British Isles exactly, nor does it account for the possibility of other branches of Indo-European, such as possibly Nordwestblock.
    • Anthony also, of course, credits the Corded Ware (2,900-2,350 BC) with spreading Germanic, Baltic and Slavic (or the languages that would eventually emerge as those, at least) to Northern Europe.  However, he glosses over the connections between the Corded Ware and further eastern cultures, like Sintashta, with its much more clear connection to Indo-Iranian.
  • The Beaker complex (and Urnfield, and other horizons that grew out of it) appear to have been the genesis of much of the R1b in Western Europe.  Most likely much of the western languages came from them; Celtic, Italic, maybe Germanic, to some degree.  The Corded Ware was mostly R1a, and is likely the source of much of that paternal lineage in today's eastern Europe, as well as in the Central Asia and Iranian and Indian regions today.  However, putting specific linguistic identities to many of the early post Corded Ware and later steppe cultures that eventually replaced Yamnaya is a little tricky.  It's in fact quite likely that descendants of Corded Ware cultures, displaced by the spread of Yamnaya earlier, in turn displaced the remnants of late Yamnaya and post-Yamnaya steppe cultures later on when further climate change made the steppes again more difficult to live on, an event which started in 2,500 BC and peaked in 2,000 BC.  Some of the cultures and movements are as follows:
    • Post Yamnaya cultures include Catacomb (2,800-2,200 BC) and Poltavka (2,700-2,100 BC)
    • Post Corded Ware, or eastern Corded Ware variants, perhaps, include Middle Dnieper (3,200-2,300 BC), Fatyanovo-Balanovo (3,200-2,300 BC) and Abashevo (2,500-1,900 BC.)  Anthony proposes that Poltavka and Abashevo peoples moved, following the further drying of the steppes, to where they had access to marshlands to water their cattle, developing fortified strongholds in the southern Urals and becoming the Sintashta culture (2,100-1,800 BC).  Meanwhile, to the immediate west of Sintashta, Catacomb, Poltavka and Potapovka cultures evolved into the Srubna culture (1,900-1,200 BC)
    • The Srubna and Sintashta cultures both seem to have been major vectors of R1a.  Sintashta in particular later became Andronovo, and is seen as a logical precursor to Indo-Iranian, although the same is often true of all of these cultures in general.  It becomes extremely difficult to propose anything that shows archaeologically a split between Balto-Slavic and Indo-Iranian, to say nothing of the later splits between Baltic and Slavic and Indic and Iranian specifically.  In fact, as Mallory pointed out 35 or so years ago, the fact that Indic and Iranian seem to be right on top of each other at a point in time where they need to be separating is a major unresolved piece of tying linguistics and archaeology together.  By the time Srubna goes away, it is replaced by the historical Scythians, but that doesn't mean that the Srubna are the ancestors of the Scythians.  In fact, it's been proposed that they were proto-Cimmerians, displaced by pressure from Scythians, which lead to their own historical raids on Anatolia.  This would be congruent with Herodotus, and would possibly give them a Thracian rather than Iranian linguistic identity.
So, this model isn't bad for the north and western languages, as well as Anatolian and Tocharian.  It becomes way too handwavey for my taste when it comes to explaining the various Balkan languages, although given that we know little about them, maybe that's the best we can hope for (and by Balkan I mean also Greek, Phrygian, Armenian, Thracian, Dacian, Illyrian, etc. which all presumably migrated into their later attested historical seats in Greece, Anatolia and Armenia from the Balkans; the Thracians (and Dacians?) might actually be a later arriving group from the steppes.  It also handwaves the origins of Indo-Iranians and Balto-Slavic peoples (and Germanic) from the same Corded Ware horizon (combined with post Yamnaya, I presume, although maybe the post-Yamnaya do a better job of describing some of the stuff that we later associate with the Balkans, like Thracian, for instance.)  Major dialectical "families" are only weakly correlated with archaeological complexes; we don't see a specific centum/satem split, for instance, and the explanation for the southern late Indo-European commonalities are kind of handwavey; assumed to possibly be areal features that spread from the Balkans all the way to east of the Caspian.  This (presumably) led to a number of correspondences that Indo-Iranian appears to have with Greek and Armenian, for instance, whereas you'd otherwise expect it to more closely resemble Balto-Slavic (which it also does, in some ways.)

Of course, some of this stuff will likely never be resolved, because we're unlikely to find texts in some of the languages that we don't know enough about to resolve their relationships, like various Italic languages, Thracian, Dacian, Illyrian, etc.  Holy smokes if we could, though, right?  On the other hand, further archaeological and especially archaeogenetic surveys can possibly tell us quite a bit more still

No comments: