Mapping the Origins and Expansion of the Indo-European Language Family

[Mikewww]

I have finished my critical review on the computational phylogenetic method recently used by Bouckaert et al. (2012) and claimed to "prove" the Neolithic Anatolian homeland for Indo-European:
http://www.mv.helsinki.fi/home/jphakkin/Problems_of_phylogenetics.pdf

Thanks, Jaakko. I like that you evaluated the pros and cons of the primary alternatives in a thoughtful discussion.

For those who haven't read the paper, this appears to be the primary conclusion (correct me if I'm wrong.)

Thus the results of computational phyloge-netics cannot give decisive support for the Neo-lithic Anatolian homeland. When we take into consideration all the linguistic evidence which can give support to either homeland option (linguistic paleontology, ancient dialect boundaries, and the Uralic contacts), we can see that...
the Copper Age steppe homeland is still the most credible option for the Indo-European homeland.

In think the Anatolian vicinity is still a very viable consideration. I just don't see how the expansion of IE was early Neolithic, which is what Gray/Atkins and Renfrew propose.

I think we have continue to investigate the Anatolian region, but must look at it from a more geographically granular level. It is apparent to me that the differences between the Med, the Black Sea, the plateaus, and the mountainous sub-regional cultures should be considered specifically. All of modern day Turkey did not get hit by the Neolithic in the same way at the same time, obviously.

[Jarman]

true but that doesnt tell us what the pool of male lineages were at the start.  The same male lineages may have mixed with different local groups and developed peculiar phenotypes prior to expansion. Also, as I have posted before, both Bell beaker and P312 exploded so fast in 100 years or so that the peculiar beaker types noted (which I understand are a minority among a less distinctive general group) may have basically been one extended family c. 2500-2400BC (when they expanded really dramtically) with a family resemblence.  The geographical patterning of the major downstream clades in Europe whose variance is very little different from L11, P312 etc (i.e. L21, U152 etc) indicates that this line was travelling fast at the time it was exploding into branches.  This matches too the sudden extension of beaker out of its smaller core and across a wide area in 100 years or so. Basically if Mr P312 came into being in SE France (or similar), his grandsons or great grandsons were settling in the north a century later.  That to me is a dramtic but initially very small scale expansion (which by implication cannot have been a military one - there was not enough time between P312 and L21 for example for there to be many of them.   

From P312 to L21 may have been short but was it as short as L11 to both P312 and U106? To me the idea a fastly spreading single family implies a reason for other cultures to welcome them (or trade for them if they were slaves) and give them wives.  Maybe Papa L11 was a good metal worker - some sort of skill that made him and his descendants welcome as they roamed (or were traded) across Europe.  Another thought is that maybe they were some sort of priestly family - were any new religions rapdily expanding about the same time?

[Jean M]

In think the Anatolian vicinity is still a very viable consideration.

What on earth makes you think that?  It is not just the date that is wrong with the idea that PIE spread from Anatolia. It is everything. PIE evolved in direct contact with Proto-Uralic. That did not happen in Anatolia.

All the evidence, linguistic, archaeological and genetic, points to the Pontic-Caspian steppe as the PIE homeland. There is no archaeological or linguistic evidence of movements from Anatolia in the Copper Age towards cultures which are clear antecedents of IE-speaking ones. The IE Anatolian languages were intruders into Anatolia. They picked up words from languages there earlier.

Genetically and archaeologically we have a clear trail from Andronovo to Scythian (known to be IE speaking). We now have both R1a and R1b in  ancient DNA from cultures both European and Asian which we can assume were IE-speaking (on cultural continuity from the Bronze Age) and which have clear cultural antecedents in Yamnaya. How clear do we need this to be?

[Arwunbee]

How clear do we need this to be?

We need this to be very clear.  I suggest someone goes out and starts digging up 5000 year old femurs that contain plenty of juicy R1b marrow inside of them.  Only dig up those femurs that are clearly inscribed with PIE words, like "wheel", "iPod", etc.  Also be very careful to make sure the inscriptions are also 5000 years old, and not a forgery.

[alan trowel hands.]

Its interesting in the new beaker book that at least a couple of the contributors see beaker as the only real candidate for the spread of a unified language or close language group across chunks of Europe.  They are not really coming from a strongly migratory position or an anti-migration position but clearly see that a major pan-European social change is hard to envisage without at least a modest degree of human movement to create a layer of uniformity within and linking diverse local cultures.  That seem to be the way they are pulling in this beaker book and I see that model of creation of a common element as very compatible with the early branching of R1b downstream from L51 and also very compatible with the Indo-Europeanisation.  I think they have got a balance between anti-migraton and old style 'invasions' just right.  The very structure of R1b below L51 in Europe shows that in the early days this was indeed a very fast spreading (but thin) strata in the beaker period and its actual domination of male lines was probably something that was only attained over many centuries, perhaps taking a couple of millennia to become very dominant in areas.  I am very happy with this model although the new book (as Jean pointed out) does dodge the origin issue which is more perplexing to people in this hobby than it is perhaps to the authors coming from a anthro-archaeological perspective. 

[Mikewww]

In think the Anatolian vicinity is still a very viable consideration.

What on earth makes you think that?  It is not just the date that is wrong with the idea that PIE spread from Anatolia. It is everything. PIE evolved in direct contact with Proto-Uralic. That did not happen in Anatolia.

All the evidence, linguistic, archaeological and genetic, points to the Pontic-Caspian steppe as the PIE homeland. There is no archaeological or linguistic evidence of movements from Anatolia in the Copper Age towards cultures which are clear antecedents of IE-speaking ones. The IE Anatolian languages were intruders into Anatolia. They picked up words from languages there earlier.

Genetically and archaeologically we have a clear trail from Andronovo to Scythian (known to be IE speaking). We now have both R1a and R1b in  ancient DNA from cultures both European and Asian which we can assume were IE-speaking (on cultural continuity from the Bronze Age) and which have clear cultural antecedents in Yamnaya. How clear do we need this to be?

Good points. I was thinking we have to write off the early Neolithic expansions completely but not necessarily Anatolia, or at Transcaucasia.  As you noted, if the Proto-Uralic influence is clear on PIE itself then that pulls the center of gravity for the probable origin north, if not both north and east.

[Mkk]

If Semetic influence could come from south of the Caucasus, why couldn't Uralic influence go south?

[Heber]

I think the extreme Neolithic expansion ( as shown by Tyler Smith) came from Anatolia and the Womb of Nations. M269.
I have no problem with PIE possibly coming from The Steppes, as R1a.
However I believe the point of contact was the Balkens, L23.
How PIE could then have transferred on masse to the vastly superior in numbers Erbins (R1b), I do not know.
I believe next there were different routes, one by the Danube and the LBK, the other by the Meditteranean and Cardial Ware (Cunliffe). They reached a hub at the Rhone and Loire source as the transition from L51 to L11  and spread to Italy,  the Rhine and Rhone source and met their cousins who migrated via the Danube.
They then controlled all the great rivers of Europe and the littolal and therefore most of inland Europe.
It was just a short journey through the Pillars of Hercules or across the Garonne Valley to Iberia and the expansion of the Bell Beakers (Patterson)  from the Tagus Valley and the great copper mines of Rio Tinto.
The trail of the Stelae people could represent a hybrid of the Danube and Meditteranean route.
The Celtic language grew from there (Cunliffe and Koch) and was essentially an Atlantic expansion with P312 and L21.
From a hub at Morbihan they spread inland again via the Loire to the source or the great rivers to form the Halstatt and Le Tene cultures and U152 and P-Celtic. The rest as they say is history. L21 and DF13 went on to expansion in the Isles (Busby) under the Gaelic Clan system. Other iron Age Celts decendented from the Halstatt and La Tene cultures entered the Isles via the Channel crossing speaking p Celtic.
Busby shows that the highest frequency for M269, L51, L11, P312, L21 and M222 is in the Ireland and in particular in the West of Ireland.

[Jean M]

If Semetic influence could come from south of the Caucasus, why couldn't Uralic influence go south?

Uralic was spoken by hunter-gatherers in the forest-steppe. Generally the language of a culture at a simple level of technology will borrow words from the language of a more complex culture. That is the case here. 

Proto-Uralic borrowed farming vocabulary from PIE and from Proto-Indo-Iranian. That places PIE to the south of Proto-Uralic, a bit closer to the Neolithic heartland. PIE in turn seems to have borrowed some vocabulary from more southerly languages such as Proto-Semitic via Kartvelian (a South Caucasian language). Exactly how much vocabulary is disputed. Some linguists think (surprisingly) that the word for "bull" that is similar in PIE and Proto-Semitic was borrowed from the former, while others argue vice-versa.

Either way there was a link between the two languages in the form of the Maikop culture trading between the North Caucasus and Sumer.

Proto-Semitic was a Copper-Age language of the Levant which seems to have spread into Sumer. It is not an Anatolian language. So it is strange that anyone would think that borrowed words from Proto-Semitic bolster the notion of an IE homeland in Anatolia millennia earlier. Wishful thinking really.

[alan trowel hands.]

If Semetic influence could come from south of the Caucasus, why couldn't Uralic influence go south?

Uralic was spoken by hunter-gatherers in the forest-steppe. Generally the language of a culture at a simple level of technology will borrow words from the language of a more complex culture. That is the case here.  

Proto-Uralic borrowed farming vocabulary from PIE and from Proto-Indo-Iranian. That places PIE to the south of Proto-Uralic, a bit closer to the Neolithic heartland. PIE in turn seems to have borrowed some vocabulary from more southerly languages such as Proto-Semitic via Kartvelian (a South Caucasian language). Exactly how much vocabulary is disputed. Some linguists think (surprisingly) that the word for "bull" that is similar in PIE and Proto-Semitic was borrowed from the former, while others argue vice-versa.

Either way there was a link between the two languages in the form of the Maikop culture trading between the North Caucasus and Sumer.

Proto-Semitic was a Copper-Age language of the Levant which seems to have spread into Sumer. It is not an Anatolian language. So it is strange that anyone would think that borrowed words from Proto-Semitic bolster the notion of an IE homeland in Anatolia millennia earlier. Wishful thinking really.

Is it certain though that Uralic only existed among the hunters in the forest steppe?  How are we certain that in the hunter-gather period that Uralic didnt extend into non-forrest steppe groups?  Also, was the bit of the steppes that Cucuteni-Trypole extended into not forrest steppe?  I f it was forest steppe and if hunters in that environment were Uralic, would that not place Uralic in contact with the C-T farmers?  Seem to as far as I can see unless you are only talking about specific parts of the Forrest steppe for Uralic.

Also, I have read a few recent articles that showed that steppe and forest steppe fluctuated wildely in the Ukraine etc leading to absolutely mind bending complexity of displacement and reoccupation phases of various cultures.

[Jean M]

Is it certain though that Uralic only existed among the hunters in the forest steppe?

Proto-Uralic was a hunter-gatherer language. There are linguistic links to Central Asian languages which suggest that a precursor language was spoken in the Sayan region of south-central Siberia. That would fit with the archaeological evidence of a drift westwards from the Lake Baikal region in the Mesolithic, bringing the first pottery to enter Europe. This need not have been one sudden migration, but a gradual spread,  meeting and mixing with hunter-gatherers spreading eastwards from the Urals.

Certainly the ancient DNA found at Sopka, Siberia of the Ust-Tartas culture (4000-3000 BC) showed a mixture of mtDNA haplogroups suggestive of hunter-gatherer mixture from west and east prior to any mixture with Neolithic types. (That would fit with one linguist's estimate that the people who created Proto-Uralic may have arrived at the Urals as late as 3000 BC.) They did not get Y-DNA from this site, but Y-DNA N1c is of Asian origin and correlated with Uralic-speakers.  

[Bren123]

Genetically and archaeologically we have a clear trail from Andronovo to Scythian (known to be IE speaking). We now have both R1a and R1b in  ancient DNA from cultures both European and Asian which we can assume were IE-speaking (on cultural continuity from the Bronze Age) and which have clear cultural antecedents in Yamnaya. How clear do we need this to be?

Where can I obtain that data?

[Jean M]

Where can I obtain that data?

Ancient Western Eurasian DNA. In the right-hand column you will find links to the bibliography. From there are links to the published sources. For the archaeological links eastwards, see David Anthony The Horse, The Wheel and Language (2007). For the archaeological link from Yamnaya to Bell Beaker particularly see Richard Harrison and Volker Heyd, The Transformation of Europe in the Third Millennium BC: the example of ‘Le Petit-Chasseur I + III’ (Sion, Valais, Switzerland), Praehistorische Zeitschrift, vol. 82, no. 2 (2007), pp. 129–214.

[Bren123]

Is there any evidence of pre Indo-European placenames in Europe?

[Jaska]

Is there any evidence of pre Indo-European placenames in Europe?

Yes, a lot: many known non-IE languages in the Mediterranean coast (Iberian, Tartessian, Etruscan, Pelasgian, Sican, Siculian etc.) and many unknown in the more northern areas (language of geminates, language of bird names etc.). Unfortunately the information does not seem to be available in Internet.

[Jean M]

We have to be careful, because non-IE does not necessarily mean pre-IE. There have been a lot of assumptions made about non-IE languages that may not be true. I once went to a lecture where the speaker showed a slide of his vision of European prehistory. This had Basque, Etruscan, Finnish and Indo-European as developing in the Last Glacial Maximum refuges of southern Europe!

• Basque appears to be a Copper Age language which may not have evolved where it is now spoken.
• Etruscan appears to have arrived in Italy from NW Anatolia c.1,200 BC.
• Finnish is a Uralic language. Proto-Uralic seems to have evolved c. 3000-4000 BC in the vicinity of the Urals. Jaska proposed that the people who evolved it had only just arrived there from somewhere in the Sayan Region of Central Asia. That fits the archaeological and genetic evidence.
• Proto-Indo-European has been realistically dated to c. 4000 BC, not the early Mesolithic.

[Richard Rocca]

Is there any evidence of pre Indo-European placenames in Europe?

Yes, a lot: many known non-IE languages in the Mediterranean coast (Iberian, Tartessian, Etruscan, Pelasgian, Sican, Siculian etc.) and many unknown in the more northern areas (language of geminates, language of bird names etc.). Unfortunately the information does not seem to be available in Internet.

Correction - Siculan was an IE language usually grouped with Q-Italic.

[Jean M]

Well I didn't want to be picky, but Koch recently argued that Tartessian is Celtic, and if we assume that the language which left the -ossos and -inthus place-names is Pelasgian, then some linguists argue that was Luwian.  

[alan trowel hands.]

Is it certain though that Uralic only existed among the hunters in the forest steppe?

Proto-Uralic was a hunter-gatherer language. There are linguistic links to Central Asian languages which suggest that a precursor language was spoken in the Sayan region of south-central Siberia. That would fit with the archaeological evidence of a drift westwards from the Lake Baikal region in the Mesolithic, bringing the first pottery to enter Europe. This need not have been one sudden migration, but a gradual spread,  meeting and mixing with hunter-gatherers spreading eastwards from the Urals.

Certainly the ancient DNA found at Sopka, Siberia of the Ust-Tartas culture (4000-3000 BC) showed a mixture of mtDNA haplogroups suggestive of hunter-gatherer mixture from west and east prior to any mixture with Neolithic types. (That would fit with one linguist's estimate that the people who created Proto-Uralic may have arrived at the Urals as late as 3000 BC.) They did not get Y-DNA from this site, but Y-DNA N1c is of Asian origin and correlated with Uralic-speakers.  

I am curious though about what the language in the western forrest steppe of the Ukraine and west might have been.  That at least partly was the environment that Cucuteni-Trypolian farmers intruded into and assimilated the Bug-Dneister hunters.   Any thoughts on the possibility that R1b was in the Bug-Dneister group and got absorbed into farming as a substrate among the C-T farmers (who I assume were not R1b)?  It would seem to make a lot of sense in terms of geography and in terms of the variance dating for the initial take off of R1b in the east.  I understand that the L23* in the lower Danube area has high variance.     

[princenuadha]

who I assume were not R1b

Im sure this has already been addressed at this site, but why do you assume the 'original' C-T weren't R1b? Also, do you have a guess as to how the hunters at the western end of the steppe forest would have gotten r1b?

[Jean M]

@ Alan

We have no idea what other languages might have been spoken by foragers in this region apart from pre-PIE. That is the only one to survive, apart from the late-arriving Proto-Uralic. Nor do we know what language was spoken by the first farmers (Cris/Koros) to mix with foragers. David Anthony guessed that it might have been an Afro-Asiatic language and that was where PIE borrowed the tauros-type word for "bull". However, as far as I can gather, said word is specifically Proto-Semitic, which is Copper Age. Some linguists even argue that the borrowing was in the other direction.

Frankly I think it would be more useful to focus on C-T, which was the survivor of the Balkan Neolithic cultures, and the one that blended with Yamnaya.

...  the C-T farmers (who I assume were not R1b)?

I'm assuming that they were. The first farmers were most likely carrying the haplogroups that have turned up elsewhere in Neolithic Europeans, while the D-D foragers were most likely carrying Y-DNA R1a1a and mtDNA U5, U4 and U2. By the time the D-D foragers had turned into farmers, they were mixed. (We have mtDNA.)   

Dairy farming brings the possibility of a new wave of people who could be carrying R1b as a majority or minority haplogroup, plus the gene(s) for lactase persistence. It seems to have come from the Sea of Marmara area via the Hamangia and Boian cultures. A group of the latter combined with some earlier farmers to create C-T. These are the people who could be speaking a linguistic precursor to Basque.   

  

[alan trowel hands.]

who I assume were not R1b

Im sure this has already been addressed at this site, but why do you assume the 'original' C-T weren't R1b? Also, do you have a guess as to how the hunters at the western end of the steppe forest would have gotten r1b?

Simply because R1b does not display the patterns of a lineage that was within the farming population until rather late.   The C-T culture originated from other older farming cultures.  If R1b had been within the ancestors of C-T then would have expanded and branched much earlier.  However it is possible that C-T may have absorbed R1b Bug-Dneister elements and that could be a way where R1b could have been in the C-T culture when its huge population was dispersed. 

[Jean M]

The C-T culture originated from other older farming cultures.  

No it didn't. That is what I'm saying. It originated from a new wave out of Anatolia. This is not at all apparent from many current works. The whole continuity assumption makes it hard to fathom what was actually happening. Some archaeologists will admit to new influences out of Anatolia. For others it is heresy. :)

[princenuadha]

The first farmers were most likely carrying the haplogroups that have turned up elsewhere in Neolithic Europeans, while the D-D foragers were most likely carrying Y-DNA R1a1a and mtDNA U5, U4 and U2. By the time the D-D foragers had turned into farmers, they were mixed. (We have mtDNA.)

This is a very random idea, but would if it was R1a1a people of a more neolithic and/or Central European type who mixed with the D-D around that time?

Bare with me. German CW has forms of r1a but every sample we have lacks u2e, which seems to be connected to chalcolithic Russia and by extension probably D-D. Even the small group of Beakers in Germany had U2e which is likely connected to the steppes. We also have a map on cranial morphology, by DiMarco, whereby the German-Polish CW shows a closer relation to the neolithic than the mesolithic, yamanaya, or even the Estonian-Russian CW. This might emphasise the neolithic spread of R1a1a. You (jean) have connected the TRB to the neolithic balkans. Though hardly suggestive, it is interesting that Swedish TRB seems to be more neolithic than Danish TRB, while modern Sweden has more r1a1a than Denmark. Evidence for the Swedish TRB being closer to the European neolithic migrations comes in the form of mtdna and the cranial morphology map referenced above.

It is also worth noting that "The Transformation of Europe in the Third Millennium BC", by Richard Harrison and Volker Heyd, seems to suggest that yamanaya demic input into Central European CW is not that great, which cooberates U2e not being found yet. This might also suggest that r1a1a was prominent in eastern-central Europe before the yamanaya migrations west.

For easy evaluation, what I have said in a nutshell, is that German-Polish CW (GPCW) is fairly neolithic-related and probably abundant in r1a1a. I don't think a whole lot of people, or r1a1a, went from the steppes to GPCW during its' development. There's a small chance that Scandinavian r1a1a has some connections to the TRB and the neolithic balkans.

And so, given the neolithicness of r1a1a (I argue) maybe r1a1a spread from eastern Central Europe to the east, including the D-D.

Is what I said even possible? Could r1a1a spread from eastern Central Europe post neolithic to Central Asia (mummies), anatolia, and other Eastern IE peoples?

Ok, I'm sure I'm wrong but I just wanted to play with some ideas.

[Jean M]

@ princenuadha

No. I just don't know where to start.

Look - the TRB people were similar to Balkan Neolithic types genetically. The Balkan Neolithic types are nothing like the steppe foragers or the Siberian foragers with mtDNA U2e who were more robust types. The TRB farmer who was tested for a large chunk of her genome was like Southern Europeans, not modern Scandinavians. So modern Scandinavians with all that R1a1a cannot be unalloyed descendants of the TRB. Corded Ware types were more robust and rather like Mesolithic types.

The idea that CW was local and just absorbed ideas from Yamnaya was orthodoxy for a long while, but does not meet the test of genetics. It is Yamnaya-derived, not just culturally, but genetically.