View Full Version : Milk and a partly non-steppe model for IE origins

08-04-2013, 05:06 PM
Dienekes has posted this paper


Its not new news really as we have known about the work of Evershed etc on this for many years. However, its idea that lactose persistence spread c. 5500BC from somewhere in SE Europe is interesting. It fits evershed's dating for the spread of milk traces on pots in Europe from around 5200BC in Bulgaria to the British Isles and Scandinavia by 4000BC. Again the basic idea of the importance of developed dairy pastoralism and the ability to digest picking up some time after the initial farming waves in Europe is not new. However, it did strike me that the genetic ability to digest this not only spread through Europe but it surely also must have spread into the steppe too in order to give rise to a pastoralist culture there.

Its not the first evidence of inputs from the farming world into the steppes. I have posted papers before about this. However, from the point of view of the ability to digest milk, it is the post-5500BC inputs from the west that seem to be important. I have mentioned before and posted links regarding craniological evidence for inputs of people from the west into Stredy Stog and a significant, particularly male, element in the Skelya/Novodanilovka elite groups of that culture. The spread of the ability to drink milk as adults in the steppe would seem very likely to have come from this input by neighbouring farming groups west of the Dnieper of which there were several groups at different times.

I just wonder if the importance of this input crossing the Dnieper into the steppe has been underestimated. It ties into my recent feeling that Stredny Stog/Skerla groups who controlled the steppe part of the Carpatho-Balkan links in metal, lithics and other influences might have had a large amount of farmer genes, apparently especially on the male side. This is the only group who had settlements from the Dnieper to the Don (late extending into the Danube and Balkans) and who controlled the steppe part of a network that ran from the Balkans to the Urals. They IMO should on the face of it have been the most linguistically influential group on the steppes between 4500BC-well after 4000BC.

Anthony links them as spreading pre-Anatolian west c. 4200BC but can we really rule out the possibility that the Neolithic element within Sredny Stog didnt introduce pre-Anatolian dialects INTO the steppes c. 4500BC or earlier and that they were earlier located within the Balkans-Anatolian Neolithic groups. There clearly was input from the Neolithic world seen in craniology and the mutatations that allowed adults to drink milk must have somehow also spread from the farming world. So, I do not see any real problem with seeing pre-Anatolian archaic IE passing INTO the steppes.

There is evidence of an Anatolian substratum in SE Europe including Greece. If that was the case then the move of early steppe groups of Sredny Stog origins may have really been an internal change with at least an element of it effectively being back migration. That would create a situation where some sort of archaic IE/Anatolian type languages could have been spoken across both the steppes and in the Balkans-Anatolian farming groups (who do appear to have retained contact) around 4500BC. Presumably this would have created a complex pattern of dialects but movements from the steppe into the Balkans c. 4200BC could have effectively converged the dialects of the steppe and the neighbouring part of the farming world again maybe into something between Anatolian and PIE. Well at least those in the contact zone around the Danube and SE Europe but perhaps leaving areas like Anatolia and Greece to remain in their own forms.

Clearly that wouldnt be the end of the story though and later waves from the steppes are significantly later enough for divergence of dialects and for the Yamnaya waves to have changed or influenced the dialects in some areas. Its hard to work out the details. If the steppe wave c. 4200BC from the Dnieper had influenced dialects in some areas of Old Europe into one that was closer to PIE in form, can we really make too much emphasis on the wheel vocab? The wheel seems to appear all over around 3500BC, including Old Europe and SW Asia, so clearly pre-Yamnaya groups would have had knowledge of the wheel for a couple of centuries. In fact, Heyd shows that Yamnaya west of the Black Sea dates to after 3000BC, so the wheel would have been known in the Balkans and east central Europe for half a millenium before Yamnaya showed up. I would add that it makes little sense to me if the ancestors of Anatolian really were based in the east Balkans and related to a steppe group who arrived c. 4200BC and remained there until 3000BC but did not pick up knowledge of the wheel. It makes more sense to me if the most archaic IE dialects simply had always been somewhere like NW Anatolia and that the break up of their old links with the Balkans simply meant that they were in an isolated position when then the wheel was invented and made up their own words for it or recieved the knowledge from Mesopotamia rather than Europe.

So, I dont think its impossible at all to rule out that the roots of IE actually started in Anatolia and entered Europe c. 5200-5500BC in the east Balkans and from there passed into the Sredny Stog groups in the Dnieper-Don area who controlled a large network linked to old Europe that stretched through the steppes from the Dnieper to the Urals around 4500BC. The steppes then would have had an important secondary role in dialect patterns when they entered Old Europe c. 4200BC and again with Yamnaya around 3000BC. A model like that allows for the possibility that those parts of farming Europe settled after 5500BC could have actually recieved para-Anatolian type dialects spread from that point until 4000BC or so as farming did its final spread. While you would expect divergence, reconnection with later groups and networks like beaker and corded ware may have converged them again, something that would have been easier to do than to totally change the language.

Jean M
08-04-2013, 05:42 PM
I made that paper available to you a few days ago, but it is open access anyway.

08-04-2013, 07:20 PM
Cheers anyway Jean. Like I said we have both talked about this a lot before on other forums but probably mainly on dna forums which is now gone. So, I think its worth having a thread on this site too. .

I made that paper available to you a few days ago, but it is open access anyway.

08-04-2013, 10:36 PM
I like that their map actually acknowledges lactose persistence actually does exist at high levels in South Asia. However, both lactose persistence SNP's 13910T and 22018A exist there as well in addition to the likely local mutations. I am guessing it didn't originate in Central Europe taking this into account.