search this blog

Tuesday, May 19, 2015

The Origins of Proto-Indo-European: The Caucasian Substrate Hypothesis


Some interesting stuff here from Allan R. Bomhard, especially in light of the ancient DNA we've seen recently from Late Neolithic/Early Bronze Age Europe.

ABSTRACT: There have been numerous attempts to find relatives of Proto-Indo-European, not the least of which is the Indo-Uralic Hypothesis. According to this hypothesis, Proto-Indo-European and Proto-Uralic are alleged to descend from a common ancestor. However, attempts to prove this hypothesis have run into numerous difficulties. One difficulty concerns the inability to econstruct the ancestral morphological system in detail, and another concerns the rather small shared vocabulary. This latter problem is further complicated by the fact that many scholars think in terms of borrowing rather than inheritance. Moreover, the lack of agreement in vocabulary affects the ability to establish viable sound correspondences and rules of combinability. This paper will attempt to show that these and other difficulties are caused, at least in large part, by the question of the origins of the Indo-European parent language. Evidence will be presented to demonstrate that Proto-Indo-European is the result of the imposition of a Eurasiatic language — to use Greenberg’s term — on a population speaking one or more primordial Northwest Caucasian languages.

Allan R. Bomhard, The Origins of Proto-Indo-European: The Caucasian Substrate Hypothesis. Paper to be presented at “The Precursors of Proto-Indo-European: the Indo-Hittite and Indo-Uralic Hypotheses”. Workshop to be held at the Leiden University Centre for Linguistics, Leiden, The Netherlands, 9—11 July 2015.

See also...

Support for linguistic macrofamilies from weighted sequence alignment

Modeling Yamnaya with qpAdm

Massive migration from the steppe is a source for Indo-European languages in Europe (Haak et al. 2015 preprint)

177 comments:

Aram Palyan said...

Nice work.
It is what I wanted to say on that other thread.
Some Eurasian population came from East ( North or South is not important for me ) that superimposed himself upon early Neolithic language ( or languages ) and created IE.
In this case Armenians are what they should be. Very high in Transcaucasian component + some South-West Asiatic ( maybe from Mesopotomia-Levant). No need for Balcanic trip at all. And the Armenian R1b came from East (north or south i don't know ).

Aram Palyan said...

About Maykop that was in North-West Caucasus.

--------------

The foreign objects in the North Caucasus reveal no connection to the upper reaches of the Euphrates and Tigris or to the floodplains of Mesopotamia, but rather seem to have ties to the Iranian plateau and to South Central Asia. Recent excavations in the Southwest Caspian Sea region are enabling a new perspective about the interactions between the “Orient” and Continental Europe. On the one hand, it is becoming gradually apparent that a gigantic area of interaction evolved already in the early 4th millennium BC which extended far beyond Mesopotamia; on the other hand, these findings relativise the traditional importance given to Mesopotamia, because innovations originating in Iran and Central Asia obviously spread throughout the Syro-Anatolian region independently thereof.

http://dienekes.blogspot.com/2013/05/origins-of-maykop-phenomenon.html

Aram Palyan said...

I will go further and say that perhaps the Kartvelian and Hurro-Urartean have same kind of genesis. A Eurasian language upon a neolithic one. But with little bit different admixtures.
Johanna Nichols is suggesting this idea for Kartvelian. This will explain very well Transcaucasia-Gedrosia connection.
The same can be suspected for Hurro-Urarteans. Alan Bomhard has an interesting study about IE and Hurrian.


Davidski said...

I think the above mentioned Eurasiatics are EHG or their ancestors.

Alberto said...

So does this say that Pre-Indo-European came from Central Asia, and it transformed into Proto-Indo-European around the Caucasus (Maykop?)?

Davidski said...

Yes, although by Central Asia he doesn't appear to mean South Central Asia, but rather the Eurasiatic steppe or even Siberia (ie. ANE country) 7000 BCE.

And the Caucasus region he's referring to is basically Abkhazia.

Alberto said...

Interesting hypothesis. Johanna Nichols, to whom he refers when mentioning the Central Asian origin, places PIE around North Afghanistan-Tajikistan-East Uzbekistan. The place that genetically seems to make more sense (don't forget that this population had some 50% ENF, though an eastern kind quite different from the one present in EEF).

Whether they migrated west through the north Caspian or the South Caspian is not evident at this point. The south Caspian is an easier route from there, but it could be the north Caspian as the author seems to suggests (especially if they arrived to Samara before than to the Pontic Steppe).

bellbeakerblogger said...

Basically, these guys are moving into the West and imposing on Near Eastern farmers.

http://bellbeakerblogger.blogspot.com/2015/02/r1-and-dispersal-of-ceramic-from-far.html

Davidski said...

Yep, like I said, the "Eurasiatic" Pre-Proto-Indo-Europeans are EHG from the steppe, not farmers from Afghanistan.

Gökhan said...

I had read this article before. I am totally agree with that. PIE came from central Asia and formed Yamnaya with interacting with Local North caucasians who have connection with Mesopotamian-Anatolian people.

Davidski said...

It doesn't say that PIE came from Central Asia. It says that PPIE came from Central Asia, and then PIE developed in Eastern Europe near the North Caucasus.

Nirjhar007 said...

I can say that Its going in the right direction and sometimes after Central Asia will be accepted as the PIE source imo The area around S Caspian, and its obvious that the Connection of PIE and Uralic is exaggerated.

Alberto said...

David, I think you are getting confused.

He says he agrees with Nichols on the Central Asian origin, and he proposes a "somewhat earlier" date of arrival to the western steppe, around 5000 BC.

These cannot be EHG. This is a Central Asian population, and we know they arrived to Samara somewhat later than 5000 BC (though not much, if what you said about Khvalynsk culture is correct). Farmers or not, they had some 50% ENF.

Nichols places the origin in Central Asia, around Tajikistan. Not in Siberia or the Kazakh steppe.

What we don't know, is when they arrived to the Pontic Steppe. Before or after Samara? Did they come from the Caucasus or from the north Caspian? Or both?

Nirjhar007 said...

Oh yes another thing that they don't know for example what PIE-Sumerian connection is and they take for granted Pontic-Caspian as PIE by default like approach, i consider that a serious error anyway there were movements from SC Asia-Iran area to Maykop as Evident so an Early IE influence there is not surprising at all.

Nirjhar007 said...

Ah I would say probably Both Alberto Similar situation is also given by Giacomo as you have read long before...

Nirjhar007 said...

and yes for those who are unaware of Nichols works here a reading-
http://www.scribd.com/doc/150388032/The-Epicentre-of-the-Indo-European-Linguistic-Spread-Useful

Davidski said...

Alberto,

Bomhard puts the entry of the Pre-Proto-Indo-Europeans into Eastern Europe at 7,000 BCE, while Nichols at 5,000 BCE.

Neither of them is talking about any Afghan or Pashtun-like population with ENF admixture, whether they know it or not.

They're obviously talking about EHG.


Nirjhar007 said...

Dude Nichols Places dispersal from around Tajikistan or Bactria who are you kidding?
Read the Research i gave. Bomhard agrees with that but assumes an older date.

Nirjhar007 said...

BTW Nichols don't assume them as PPIE but IE folks.

Davidski said...

This paper is not about Nichols' hypothesis.

It's about a Uralic-related Eurasiatic population mixing on the Eastern European steppe with North Caucasians to form the Proto-Indo-Europeans.

Alberto is reading too much into Bomhard's quote about Nichols. There was no migration of Afghans or anyone similar to Eastern Europe 7,000-5,000 BCE. Forget it.

Alberto said...

You didn't read correctly. I'll quote:

"In a paper published in 1997, Johanna Nichols argued that the earliest Indo-European speech community was located in Central Asia. She proposes that Pre-Proto-Indo-European spread westward across the steppes, eventually arriving on the northeastern shores of the Black Sea. I support this scenario. I would place the Pre-Indo-Europeans in Central Asia at about 7,000 BCE, and I would date their initial arrival in the vicinity of the Black Sea at about 5,000 BCE — this is somewhat earlier than the date Nichols assigns. No doubt, the immigration occurred in waves and took place over an extended period of time."

So he says the Pre-Indo-Europeans where in Central Asia around 7000 BC (which makes sense), and he places their initial arrival to the vicinity of the Black Sea around 5000 BC (which might be correct, though it looks a bit early. He himself points out that he refers only to an INITIAL arrival, a first wave, not the bulk of the migration).

Nirjhar007 said...

David,
And then you miss our Logic again its not about the dates but the pattern! though Bomhard is speaking of PIE in around E Europe but he accepts its ultimate origin is from central Asia.
And Nichols also followed similar approach though she showed PIE has likely locus around Bactria.

Davidski said...

Did you guys even bother to read the paper?

"As a result of the socio-cultural interaction with and resultant borrowing from Caucasian languages, especially primordial Northwest Caucasian languages, Proto-Indo-European developed unique characteristics that set it apart from the other languages of northern Eurasia."

He's obviously talking about a North Eurasian people. You know, like ANE/EHG.

This has nothing to do with any farmers from Afghanistan, who didn't even exist 7,000 BCE anyway.

Alberto said...

Honestly, this is starting to sound silly.

This population was in North Afghanistan and Tajikistan 7000 BC. Most likely, they were still ANE. But soon after, ENF arrived to the area (Jeitun, since at least 6500 BCE).

When this population moved west, they obviously had ENF, because we have ancient DNA that shows it!

You just said the Khvalynsk had already 25% of this Armenian-like population, and that's the 5th millennium BC.

EHG were in Karelia at least from the 6th millennium BC.

Are you really proposing that EHGs migrated from Central Asia to Eastern Europe from 5000 BC onwards?

a said...

Interesting linguist ideas/evidence. Perhaps distribution pattern of R1b-M73 and it's possible split from R1b-M269 in Circum-Uralic region[elevated ANE/EHG] also point to some migration reaching into Southern regions and paucity of R1b-M73 among Indian/Southern Caucasus populations.

Davidski said...

Alberto,

No, the North Eurasian population here has to be EHG, probably originally from Siberia.

That's why the Yamnaya/Corded Ware males carry Y-HG R1 plus mtDNA from the North Caucasus.

It makes perfect sense. There's no need for any mysterious ENF group from Central Asia. It's unnecessary.

a said...

R1* and subsequent distribution of R1a-R1b lineages.Mixing Proto-Uralic & Proto-North Caucasian & Dravidian and Munda.

Alberto said...

Ah, so the paper is about EHGs who were R1 and came from Siberia to Eastern Europe and took Caucasus wives?

Ok, I was missing those points completely. I'll read it more carefully when I have more time.

Nirjhar007 said...

Here is the synopsis of his suggestions-
''The precursor of Proto-Indo-European came from Central Asia. Proto-Indo-European proper is the result of the imposition of a Eurasiatic language on a population speaking one or more primordial Northwest Caucasian languages''
It doesn't say that Proto-PIE came from Northern Eurasia just Central Asia obviously as Nichols specified the area Bomhard didn't specify anything! just that it came from C Asia and ''Became'' PIE Proper with fusion with NW Caucasian languages, i think this is an wishful enforcement in simple words NW Caucasian had early PIE speakers from N Iran and it gave rise to Maykop and others, Nichols has shown in her paper why that area with P Caspian can't be the source of PIE as she firmly states-
''that the locus accounting for the distribution of loanwords, internal innovations, and genetic diversity within IE could only have lain well to the east of the Caspian Sea''
Alberto,
Yes Exactly those points you mentioned is as accurate as it can get.

Kristiina said...

It looks an excellent paper. The list of morphological similarities between Proto-Indo-European and Proto-Northwest Caucasian is impressive. At home, I have a paper copy of Colarusso's "A Grammar of the Kabardian Language" and "Phyletic Links between Proto-Indo-European and Proto-Northwest Caucasian", and I agree that he really is on to something.

It is intriguing that many of the so called Nostratic (=North Eurasian?) parallels are missing in the list, as if confirming the north – south admixture:
/na/ ”originally a locative particle”
/t/ ”in the meaning from”
-/t̕V/ "inanimate plural" (if IE ”s” is accepted to be derived from ”t”)
'm' ”nominative/oblique for I” (the correspondence in the list above is much weaker)
't' ”nominative/oblique thou” (IMO the form PNC w- corresponds much better with IE plural form v-)
't' "this"

I am also seeing proto-languages as mixtures of earlier languages, and I think that all languages are mixed in the same way as our genes.

spagetiMeatball said...

So you think the EHG guy in karelia was just wandering around there?

Skilur said...

@Alberta and Nirjhar
Do you can show any evidences for ANE/ENF farmers migrating into the northern Central Asian steppe? Archaeologically there are no evidences for the presence of farmers cultures in Kazakhstan north of BMAC and Yamnaya could only get their Near Eastern ancestry directly from South Central Asia if non-farmer South Central Asians rich in ANE/ENF moved into the Asian steppe. It is much easier to explain the armenian-like admixture in Khvalynsk and Yamnaya with genetic influx from the Caucasus and most southern Steppe.

Nirjhar007 said...

Kristiina,
IMO Pre-PIE has clear Semitic substratum a very strong Relation with ancient Asian languages like Sumerian and dispersal into the Caucasus and SC asian area at a very early time say even middle 5th millennium BC and it also started to spread towards E Europe around the same period which intensified after the 5.9 KYO event.

Nirjhar007 said...

Skilur, Ah yes there are Jarmo related migrations for Example in North Eurasia up to Urals from S Caspian area and those areas are of Herding type from ancient times not farming, C Asia-SC Asia was a ANE rich area with enough ENF type ancestry also from archaic times.

Alberto said...

@Skilur

I'm not sure what you're trying to say. That you disagree with the author in the route that this population took from Central Asia to the Pontic steppe?

The author suggests it was through the steppe (Kazakhstan). I think that's an option, and the other option is through the south Caspian to the Caucasus.

Personally I prefer the second option, but I don't discard the first one. Maybe it was both.

The Armenian-like admixture in Samara is proved by ancient DNA, so I'm sure you're not asking about that part.

Skilur said...

A route through Kazakhstan makes not much sense for me. This region were sparsely populated by HGs and some kind of Proto-Pastoralists but not any traces of farmer cultures where found there. It is likely that they got ENF admixture from South Central Asia just like EHG got it by mixing with Caucasian women but Davidski and others have already shown that the Near Eastern ancestry of Yamna is linked to the Caucasus even when it is quite similar to the Near Eastern ancestry of South Central Asians. Both ancient Caucasians and South Central Asians seem to share the same kind of ANE+ENF ancestry but nobody knows where this "teal" component was formed. Maybe Iran prior to the migration of new Near Easterners who lowered ANE there so that only in the Caucasus and in the Hindukush/Pamir this "teal" people survived

Skilur said...

What? There were no farmers north of BMAC and farming is even today uncommon in the Kazakh steppe. The ancient North Central Asians got probably some ENF admixture from the South because of "trade", war and cultural interactions but neither they adopted agriculture nor left any traces of farming there. They were HGs in most cases and in later periods maybe cattle herders but nothing Jarmo relates existed North of Tajikistan/Usbekistan

Nirjhar007 said...

Skilur,
Nor they should as the area is not ideal! but certainly a movement after ANE-ENF type admixture in C Asia is logical and there is for example Neolithic Kelteminar culture, dated 5500-3500 BC, spread from northern Turkmenistan to Kazakhstan. Its stone industry had common elements with the Caspian, Ferghana and Hissar cultures and moreover It is notable that all these areas, when they first appear in history, belong to Indo-European cultures. According to wiki the Kelteminar people were Mesolithic groups coming from the Hissar area (South Tajikistan/North Afghanistan), with bow, arrow and dog.the Hissar culture has been derived from the Zarzian the area where i consider PIE came to existence.
In an Important though old book Titled ''Prehistoric Russia'' by Tadeusz Sulimirski it is suggested how the Pre-Cursors of cultures like of Yamnaya were influenced from the Northern Central Asian cultures originating in the south Including Kelteminar culture ultimately coming from the Zarzian horizon.

a said...

Interesting North Caucasian result's using K36 & a couple Yamnaya samples.

Yamnaya1-M951285
Fennoscandian 18.07%
North_Caucasian 17.09%
Eastern_Euro 14.50%
North_Sea 12.89%
South_Central_Asian 12.51%
East_Central_Euro 10.17%
North_Atlantic 7.51%
Volga-Ural 3.51%
Central_Euro 1.90%
Amerindian 1.85%

Yamnaya2-M020637
North_Caucasian 20.00%
Eastern_Euro 16.34%
North_Atlantic 15.65%
North_Sea 14.04%
Fennoscandian 11.09%
South_Central_Asian 10.74%
Volga-Ural 5.36%
East_Central_Euro 5.41%
Amerindian 1.36%

Nirjhar007 said...

Ah yes some more suggestions the Jeitun culture is a sort of ancestor of the BMAC probably came to existence in Time when ANE-ENF mix in Central Asia was in the mature phase, it belongs to the Neolithic period, starting around 6500 BC and ending around 4500 BC, while the BMAC is of the middle and late Bronze Age (2500-1700 BC). In the middle, there are the subsequent Namazga phases of Chalcolithic and early Bronze Age.
And also guys Mariya Ivanova have worked on the links of Maykop with Iran and Central Asia, that you can find on Academia.edu, in German, but there is also her book in English: http://books.google.it/books?id=wNsgAwAAQBAJ&vq=yamna&dq=Telmankend&hl=it&source=gbs_navlinks_s

Alberto said...

@Skilur

If this population came from Central Asia (which is what the paper is proposing), I do favour a southern route to the Caucasus as the main one. But the author suggests a north Caspian one through the Kazakh steppe.

I don't think that's the main point to debate about this paper, anyway.

The paper itself is about the formation of PIE in the Caucasus, from a Pre-IE brought by a population from Central Asia (this based on Nichols' theory which is based on linguistic evidence itself) and the local contact with NW Caucasus languages.

The question of when did ANE arrived to the Caucasus is not an easy one with the data we have. We can only speculate.

a said...

One point of interest with Yamnaya and K36 South Central Asian found are the copper hoards so-called bar-celts [/“Cudgel-scepter”] found around Ganges-Yamuna Doab region.
“Oh, Indra, getting your support Let us take cudgels,Like (…) vajra, And we will gain victory over all enemies”RV
I.

8.3
“The owner of golden horses takes upThe weapon - golden vajra in his hands.”RV 3.44.4 “Desiring (…), the glistening vajra,Covered by light (…)”.RV 3.44.5


http://www.academia.edu/3836804/An_Indo-Iranian_Symbol_of_Power_in_the_Earliest_Steppe_Kurgans

tew said...

With all due respect to the hard work of Bomhard and others from the "Eurasiatic" and "Nostratic" camps, their ideas are not mainstream and are considered highly speculative at best by the majority of linguists (partly because of a lax application of the traditional comparative method, "lumping" bias and reliance on mass comparison). Nichols' work in typology is mainstream and well respected, but her theories on deep linguistic phylogeny are somewhat more controversial. I say this to add a bit of perspective, lest people without a background on linguistics might take research such as this as the latest big news in historical linguistics, when in fact some of the premises it starts with are highly contentious (and then a few downright fringe).

Kristiina said...

Nirjhar, when you say that "Pre-PIE has clear Semitic substratum and a very strong Relation with ancient Asian languages like Sumerian", can you back it up with a list of morphological similarities between either Sumerian - Proto-IE or Proto-Semitic-IE?

The basic characteristics of Sumerian are the following:
split ergative
human/non-human distinction (no male/female distinction)
perfective and imperfective aspects (no past/future distinction)
cases: ergative (-e), genitive (-(a)k), dative/allative (-r(a) for human nouns, -e for non-human nouns), locative (-a, only with non-human nouns, -ne), comitative (-da), equative ("as, like") (-gin), directive/adverbial ("towards") (-š(e)), ablative ("from") (-ta, only with non-human nouns).

Genitive is possessed + possessor –(a)k

personal pronouns: ĝe26-e (1st p. sing.), ze2-e (2nd p. sing.), a-ne or e-ne (3rd p. sing. human), and a/e-ne-ne (3rd p. pl. human), -b(i) (3rd p. sing./pl. non-human, also demonstrative and collective), -me (1st pers. pl.), -zu-ne-ne (2nd p. pl.), and -(a)-ne-ne (3rd pers.pl. animate)

Basic word order is subject–object–verb.

IMO, there are some shared characteristics with Afro-Asiatic, with several Caucasian language families and North Eurasian languages, but if you are aware of a more detailed comparative morphological analysis between Sumerian/Semitic and IE, I would be happy if could share it.

Ryan said...

I have to say I find this very interesting and very plausible.

Hattic is thought to have been related to Northwest Caucasian languages, no?

If Indo European is the result of an EHG language intruding onto a Northwest Caucasian language, could Hittite and the Anatolian branch of IE be the result of a separate but equivalent mixing?

It would reconcile the evidence for the Anatolian hypothesis into the Steppe hypothesis rather elegantly. Hittite would be the most diverged Indo European language because it would actually be the result of an EHG language on to a separate but related branch of NW Caucasian languages.

a said...

Aram Palyan said...

"The foreign objects in the North Caucasus reveal no connection to the upper reaches of the Euphrates and Tigris or to the floodplains of Mesopotamia........."

So does grape/wine derive from Proto-Kartvelian-Armenian or Proto-Indo European?

Also no word for wine in Akkadian :)
I thought perhaps Akkadian's produced wine from dates, not grapes, any ideas?
http://en.wiktionary.org/wiki/Appendix:List_of_Proto-Semitic_stems

Skilur said...

We have R1b in Neolithic Samara and we have it in Yamna Samara but with armenian-like admixture, which both Karelia and Samara HGs lacked so there is really no point anymore in debating a Near Eastern/Caucasian origin of R1a or R1b (I myself believed that some months ago). EHGs indeed have (North) Central Asian ancestors but they brought ANE to far eastern Europe without carrying any ENF and this migration happened more than 10000 years ago. EHGs represent the paternal ancestry of PIEs and certainly spoke Proto-PIE so if we go enough back in time the very distant ancestor of PIE was spoken in Central Asia/Siberia for sure.

Roy King said...

The new YFull Tree has the M269* lineage with a TMRCA of 3000 BCE (includes a sample from Trabzon, Turkey and an Ashkenazi Jew). This date could be consistent with the first appearance of the Anatolian languages (Luwian/Palaic/Lydian/Hittite) in Anatolia.
Also check out the new Nature paper from Jobling on Y sequences from Europe/Turkey and Palestine: http://www.nature.com/ncomms/2015/150519/ncomms8152/fig_tab/ncomms8152_F1.html

Mike Thomas said...

I have read throught the paper briefly, and it seems above there is some debate as to what it is saying.

In my understanding, the author clearly suggests that

* pre-proto-Indo European developed in central Asia, explicitly linking it to J Nochol's theory. She quite clearly locates this c. Bactria. So however we might define 'central Asia' this is clearly not Siberia, north Kazakhstan or "EHG" territory

* Moreover, he clearly stipulated that the original peoples of the north Ponto - Caspian steppe were **Caucasian**. He argues that later ("propper") Proto-IE then developed on the steppe through the effect of what is essentially imperfect language shift of a Caucasian population adapting to PPIE.

* Overall, i think its a lovely paper. His level of knowledge of phonology and morphology us rgeat and far suprpusses anything I'd be capable of critiquing. But my only criticism is:

- he perhaps too readily assumes a 'simplicity' of Hittite due to its greater antiquity. Acutally, recent persepctives on Hittie do not see it to be that divergent viz-a-viz "Nuclear (core) IE",, and we must account for langauge contacts in Anatolia for its apparently queer position. Indeed, all other Anatolian languages (Luvian, Palaic) are entirely within normal IE variation.

- he takes at face value Mallory's and Anthony's account of Tocharian. This still doesn;t sit well with me. Tocharian is not attested until Late Antiquity, and assuming that it just melled in the Tarim basin for 3000 years, evoliving linearly and gradually seems highly unlikely.

* He also takes at face value the date of 3500 BC of the waves of different 'nuclear' PIE langauges (like GReek, Iranian) dispersing. I've argued previously, this is far too early IMO, and should rather be placed to c. 2500 BC - from what we know about Mycenean Greek.


* all in all, a great paper. It confirms with what I have thought myself. The north Pontic region and beyond, was the homeland of Uralic and Caucasian (& 'para-Caucasian') languages, and not PIE/ PPIE - or whatever we wish to term it.

Davidski said...

The point here is that a linguistics paper identifies two main groups that made up the Proto-Indo-Europeans in Eastern Europe: a North Eurasian group and a North Caucasian group.

This gels very nicely with the ancient DNA results we've seen to date.

Bomhard obviously hasn't looked closely at this ancient DNA yet, so he shouldn't be expected to pinpoint the origins of the population expansions with any great accuracy. But he's generally got things right.

The North Eurasian Eurasiatics identified by Bomhard are obviously EHG foragers carrying R1. These are the people who imposed themselves on a North Caucasian substrate population somewhere near the North Caucasus to create the Proto-Indo-Europeans.

This paper isn't about Nichols' hypothesis. It just very broadly agrees with it. Nichols didn't have any ancient DNA to work with back in 1997.

Davidski said...

Mike,

What the hell are you mumbling about there at the end?

There are only two groups involved here: the North Eurasians and North Caucasians.

One of these groups has to be the PPIE.

Mike Thomas said...

Davidski

This is a linguistics paper which needs to be analysed on its own merit -
Don't muddy the waters with "EHG"- which is a genetic category.

But to put it your terms- EHG people spoke several languages by 8ky BP, and not just one, single PPIE as you'd believe. This is a simple fact if one understands forager group densities and sociolinguistics.

And the paper is more explicit than you claim: it argues that PPIE came from central asia (specifically linking it to Nichols' theorem linking it to Bactria and the surrounds, and *not* Siberia).

Then, due to substratum effects, PPIE evolved into a later form of PIE on the steppe, which had otherwise been the home of **Caucasian ** related languages.

FYI the term "para-"language- which is unfamiliar to you - is a relatively recent term introduced by the likes of Jaha Janhunen to describe para phyletic language groups distantly but not directly related to a main branch. So Hunnic or Bulgar can be described as "para -Turkic".

Davidski said...

He's talking about two main groups that made the Proto-Indo-Europeans in Eastern Europe.

EHG + North Caucasian farmers = Proto-Indo-Europeans.

But we've known this for a while now, haven't we?

Mike Thomas said...

well yes, that is if Bactrians c. 7 ky BP were EHG type.

Davidski said...

See that's what you and Alberto keep missing here.

The North Eurasian population that Bomhard is talking about wasn't from anywhere near Bactria, whether he actually knows it or not.

Aram Palyan said...

I read once a more Johanna Nichols paper.
I see one great point. but also weak points.
The weak point that she moves the Proto-Kartvelian with IE into South Central Asia. This is speculative and needs more serious proofs. Without this move her IE locus will not work because the highest similarity is between IE and Kartvelian.
Another possible weak point is that the locus of IE needs to be close to sea. Which is impossible in Tajikistan.

The great point is that Steppe works like spread zone. So Steppe is important for spreading like in the case of Scythian, Turkic and Russian later but it is not a good place for origin.

Mike Thomas said...

Well, no we're not misindersranding him
It seems you're putting words in his mouth, based on your interpetation of the genetic evidence.

Rightly or wrongly, based on his historical and linguistic criteria, he places it squarely in central Asia, strictu sensi, north of Iran.

And I think that is more probable. Little came from Siberia, except perhaps Uralic and some moose.

Aram Palyan said...

BTW Georgians has a medieval story about their original homeland called Arian-Kartli.
https://en.wikipedia.org/wiki/Arian-Kartli

Kristiina said...

Mike, at the moment I am fairly sure that Uralic contains a similar dichotomy as Bomhard’s Proto-Indo-European and Proto-Northwest Caucasian fusion. On the one hand, there are northern/West Siberian microblade cultures and on the other hand, new technology coming from the south and bringing in the so called IE words. The spread of Proto-Uralic in the north could well be linked with mining activities and trade in the Urals. Considering the behaviour of women, the more powerful yDNA’s were probably most succesful.

Mike Thomas said...

Kristiina
Do you mean to say that Uralic spread across the forest and forest steppe zone during the early Bronze Age "opening up" of mining in the Urals, etc ?

Kristiina said...

Yes. When languages spread, they need a "motor", which is usually a better life or at least a hope for a better life. :-)

And it is not only moose, you should check ”Prehistoric Copper Mining in Europe: 5500-500 BC, William O'Brien”, which says that another example is the mining of several million tonnes of copper ore at Kargaly (Orenburg Oblast) during the Bronze Age, leading to the production of an estimated 150,000 tonnes of metal. The smelting of this oxidized ore required large quantities of charcoal fuel. It was necessary to burn 300-500 cubic metres of pine or birch to produce enough charcoal to smelt a tonne of copper. Rovira estimates that 75 million tonnes of wood would have been consumed by the Bronze Age production. The nearest true forest is located 200-250 km to the north-east in the mountainous zone of the southern Urals.

Mike Thomas said...

I'd agree with that time frame.
I think most of northwest eurasia was Uralic speaking, or now extinct para-Uralic languages. I know some Uralicists would even place this as far south Ukraine.
Only problem is, scholars see Uralic as a mesolithic language, based on the Samoyed -Uralic split

Kristiina said...

It is a question of how you define things. Of course, you can put all Mesolithic stuff and ancient ways of living in the "Samoyed" category and say that this is Uralic and then claim that Uralic people are underdeveloped hunter-gatherers.

However, IMO, mainstream understanding is that Proto-Uralic developed in the Urals and spread from there to the surroundings and Samoyed-speaking people ended up in the tundra after a tour in Altai postdating their origin in the Urals.

FYI Orenburg Bashkirs’ yDNA:
Western Orenburg: http://ftp.anrb.ru/molgen/Lobov_AS.PDF

R1a 40%
R1b 23%
C-M48 12%
N1c 7%
E-M35 7%
C-M130(xM48)5%
G-P15 2%
I*-M170 2%
L-M20 2%

Eastern Orenburg

N1c 65%
R1a 18%
R1b 9%
O-M175 6%
I-P37 2%

Основная доля Y-хромосом в изученных субпопуляциях приходится на три гаплогруппы (R-M269, R-SRY10831.2 и N-Tat), которые в сумме составляют от 70% до 100%.

Burzyansky Bashkirs http://www.ncbi.nlm.nih.gov/pubmed/25857199

R1a 34.5%
R1b1b2 37.9%
N 19%
R1b1b1 5.2%

Mike Thomas said...

Thanks Kristiina
I agree with you.
Interestingly, Thomas & Kaufmann- leading contact Linguists- claim there is a Uralic influence on all slavic languages(ie not just northern russian). If true, then a case can be made that much of EE was Uralic at one point, prior to PIE arriving there. This could be as late as 1000-600 BC.

Skilur said...

@Mike
Neither there are Uralic river names in the steppe nor are they in Poland, Ukraine, Belarus and South Russia. Even in Latvia and Lithuania just few of them exist. Some Finno-Ugrian HGs moved probably much more south but they hardly dominated any area Soutwest of Russia.

Mike Thomas said...

Skilur
There aren't *any* river names in EE which aren't slavic , baltic, "northwest IE block" or Iranian on the steppe.
Contrary to what three hundred years of misguided linguists claim, river names aren't a crystal ball to the past; esp when that past was illiterate.
So according to these river names you'd hold so dearly, europe was uninhabited until 600 BC

Skilur said...

@Mike
It just means that Uralics had at least since the Bronze Age no dominating role in any of this regions. The Uralic substrate in Russian is even almost non-existing (very very few loanwords) if we exclude North Russian dialects. So I doubt that Proto-Slavic was heavily influenced by Finno-Ugrian languages. Nobody doubts that Uralics dominated large parts of Russia once and that some of them moved southwest(Ukraine, Belarus,..) but they hardly dominated all of EE

Mike Thomas said...

Well yes I agree.
I was not stating that it is my belief / theory that Uralic was the common tongue of EE for most of prehistory, but was merely iterating one interesting proposal by a team of rather well known linguists.

But the point I want to impress is that we cannot use river names as guides to deep prehistory. The earliest river names come from the turn of the 6th century BC- when greek Historians first started naming things.

Most of the information they got about the pontic region seems to have come via Iranic speaking intermediaries, no doubt linked to the depth of The Persian empires, and the presence of "Scythian" mercenaries all around the near East and Caucasus.

And even then, the river names we have (eg Borysthenes) are confined to the steppe. Who knows exactly what languages were spoken further north in Belarus or Latvia. But i can almost certainly assure you, it wasn't Slavic or Baltic- at least not yet. And who knows how they called the rivers? What psycholinguistic processes shaped their naming of the landscape with which they interacted ? The "archaic" hydronyms that modern scholars have studied ad nauseum are mere recent constructs

Mike Thomas said...

And my point about Uralic was that undoubtably at some point in pre-history perhaps prior to the Bronze Age it was spoken further down south compare to its current range

Skilur said...

@Mike
I am certain that north of Iranians and South of Finno-Ugrians various Proto-Slavic and Baltic languages were spoken else we could not explain the Iranian loanwords in Slavic languages. The oldest Slavic hydronyms are found between the Vistula and Dnjepr and archeologically this is the most plausible place where Proto-Slavic involved having contacts with Iranians in the southeast, Baltics in the North and East Germanics in the West. Old Balts populated large parts of Belarus and Southwestern Russia since the Bronze Age and were just the descendants of eastern Corded Ware folks. There was no major migration from Central Europe into EE and Balto-Slavic presence in Eastern Europe is the result of Corded Ware expansion. Later Slavs expanded from a small region somewhere between eastern Poland/Western Ukraine/Southwestern Belarus and assimilated genetically very similar East Germanics and Balts

Nirjhar007 said...

Kristiina,
Yes my conclusions are based on seeing Bernard Segeants works first of all, to the Semitic-Hamitic linguistic family, because of root structure (with 2 or 3 consonants), grammar (the fusional system instead of agglutinative), phonetics (including laryngeals), and a very important common vocabulary, especially with Semitic for instance the number seven with this he also demonstrates that there is a rich comparative material showing that all seven first numbers are common between Indo-European and Semitic!
And about Sumerian and Indo-European relation Giacomo has published a short article on the issue as i linked you guys on the other thread though more work is coming up which will focus more on Maykop and Asia.
http://new-indology.blogspot.in/2015/05/sumerian-and-indo-european-surprising.html
BTW i couldn't read the paper with time as i'm travelling but Some parallels seem good, but it is difficult to say if there are ancient loanwords or a relation.
Probably NW Caucasian people came from the south, someone says (also cited in the paper)
that Hattic of Central Anatolia is NW Caucasian, in that case there could be an ancient relation
with some IE people in Eastern Anatolia. But then, what about his theory of relation with Hurrian?

Mike Thomas said...

@ Skilur

" am certain that north of Iranians and South of Finno-Ugrians various Proto-Slavic and Baltic languages were spoken...""

Yea I'm certain of it also- c. 1000 BC - 500 AD .
Prior to this, one cannot talk of slavic or baltic, or even Uralic; at least not apart from some pre-pre-proto languages which could not be tracked down to any specific locality or specific permutation of spatial relations

Nirjhar007 said...

Mike, PIE Originates in ~3000 BC according to you, please tell us if already haven't when it reach the Eastern European Area?

Nirjhar007 said...

a,
//So does grape/wine derive from Proto-Kartvelian-Armenian or Proto-Indo European?//
I have tried to point this before That the word doesn't Exist in Indo-Iranian or Tocharian, i clearly doubt that its a real PIE word! but a borrowing from Near East.
See here-
http://starling.rinet.ru/cgi-bin/response.cgi?root=config&morpho=0&basename=%5Cdata%5Cie%5Cpiet&first=1&off=&text_proto=&method_proto=substring&ic_proto=on&text_meaning=&method_meaning=substring&ic_meaning=on&text_hitt=&method_hitt=substring&ic_hitt=on&text_tokh=&method_tokh=substring&ic_tokh=on&text_ind=&method_ind=substring&ic_ind=on&text_avest=&method_avest=substring&ic_avest=on&text_iran=&method_iran=substring&ic_iran=on&text_arm=&method_arm=substring&ic_arm=on&text_greek=&method_greek=substring&ic_greek=on&text_slav=&method_slav=substring&ic_slav=on&text_balt=&method_balt=substring&ic_balt=on&text_germ=&method_germ=substring&ic_germ=on&text_lat=&method_lat=substring&ic_lat=on&text_ital=&method_ital=substring&ic_ital=on&text_celt=&method_celt=substring&ic_celt=on&text_alb=&method_alb=substring&ic_alb=on&text_rusmean=&method_rusmean=substring&ic_rusmean=on&text_refer=&method_refer=substring&ic_refer=on&text_comment=&method_comment=substring&ic_comment=on&text_any=wine&method_any=substring&sort=proto&ic_any=on

Mike Thomas said...

Nirj

Yes, more or less. The form of PIE from which modern IE languages, incl known but extinct ones like may of the Celtic, and even Tocharian, spread rapidly over a wide area c. 4000 - 3000 BC, but at the tail of this period, this "PIE" still existed more or less as a unity.

Whilst widepsread, it was not necessarily all encompassing, dominant, exclusive, etc.
Moreover, the forms of languages like Celtic, Baltic, etc, almost certainly evolved later within Europe, and did not come as bone fide languages with the advent of cultures like CWC. No sir.

Kristiina said...

Skilur, you say that ”Old Balts populated large parts of Belarus and Southwestern Russia since the Bronze Age and were just the descendants of eastern Corded Ware folks”, however, N1c and R1a were found in Smolensk area in the pile dwelling context with pottery and no metal, 2500 BC. They did not necessarily speak any IE or Uralic language and their languages were not necessarily even related to the language spoken by the Karelian hunter gatherer.

You also say that you ”doubt that Proto-Slavic was heavily influenced by Finno-Ugrian languages. Nobody doubts that Uralics dominated large parts of Russia once and that some of them moved southwest (Ukraine, Belarus,..)”.

However, I do not think that Uralics moved southwest in any meaningful numbers. Instead, the Uralic language in the mainstream linguistic meaning travelled north, both to the northwest and northeast. N1c you have in the Southwest is probably autochtonous.

According to the definition that I gave to the Uralic languages (i.e. a fusion between West Siberian microblade cultures and people who were acquainted with new technology and animal husbandry and spoke a language related to or comparable to proto-IE), Uralic languages could only develop in the former areas of the microblade cultures, so they obviously did not exist in more southern areas.

Mike Thomas said...

Where do you think the process of fusion occurred, Kristina, for Uralic ?

Kristiina said...

MIke, I get back to your question later, but here is another opinion! :-)

At the moment (however, I may change my mind!), I think that I1 and N1c-VL29 travelled to Finland from the Baltic area during the Corded Ware and spoke a Corded Ware language. Thereafter, perhaps during the Iron Age due to a technological advantage, there was another male expansion with N1c Z1936 spreading from the East that brought a Uralic type of language to Finland. In the end and after several severe famines and epidemies, all remaining hunter gatherer yDNA’s were practically lost and compensated by a number of more recent and varied foreign y-lines.

a said...

@Nirjhar007



Your quoted source.
Proto-IE: *wein- (Gr w-)
Meaning: wine
Hittite: wijan(a)- c. 'Wein' (Friedrich 255)
Armenian: gini `Wein'
Old Greek: ói̯no-s, dial. woi̯no-s m. We∈',ó i̯nǟf.Weinstock'
Slavic: *vīno
Germanic: *wīn-a- n., -m.
Latin: vīnum, -ī n. `Wein'
Other Italic: Volsk, Falisc vinu We∈';Umbrv∈uWein'
Celtic: OIr fīn We∈';Cymrgw∈Wein'
Albanian: tosk. verɛ, geg. venɛ Wein
Russ. meaning: вино (виноградное)
References: WP I 223 f

Does the word exist in Armenian? Proto-kartvelian? How about Akkadian? or Sumerian?

Nirjhar007 said...

No not Sumerian, Yes in Armenian.
See also this one-
http://www.etymonline.com/index.php?allowed_in_frame=0&search=wine&searchmode=none

a said...

Nirjhar007

"wine (n.) Look up wine at Dictionary.com
Old English win "wine," from Proto-Germanic *winam (cognates: Old Saxon, Old Frisian, Old High German win, Old Norse vin, Dutch wijn, German Wein), an early borrowing from Latin vinum "wine," from PIE *woin-o-, related to words for "wine" in Greek (oinos), Armenian, Hittite, and non-Indo-European Georgian and West Semitic (Arabic wain, Hebrew yayin), probably from a lost Mediterranean language word *win-/*woin- "wine."

http://edition.cnn.com/2010/WORLD/europe/04/20/georgia.wine.heritage/


http://edition.cnn.com/2010/WORLD/europe/04/20/georgia.wine.heritage/

"Haplogroup R1b had considerably high frequencies (>25%) in the
southern part of Georgia, in Armenia, and at the Caspian Sea Coast, peaking at
50% in Georgian speakers of the Kartvelian"

"Archaeologists suggest that the first Indo
-
Europeans
expanded to the Southern Caucasus from the south
-
west in 3
rd
Millennium BC
(Melikishvili, 1959; Melaart, 1970). "

a said...

@Nirjhar007

8k
http://digitalcommons.wayne.edu/cgi/viewcontent.cgi?article=1053&context=humbiol_preprints

Nirjhar007 said...

a,
Thx a for that paper yes i knew it and i have discussed it, the paper actually reflects how complicated is the Caucasian situation, but it gives also some useful data. Association of J1 with poorly forested mountains confirms that it was brought by shepherds, it is generally associated with Semitic people, maybe they had to change their language in the Caucasus??. R1a in Turks is a heritage of the Scythians, I guess. I am not convinced by the proposed places of origin, it is incredible that still says that R1a originated north of the Black sea. The Hg G is found also in Europe up to Portugal and much in the Levant, which is the homeland of Western Semitic. I assume it can be suggested that G2a of Early European Neolithic (LBK and Cardial) came from the Levant, which is the most probable origin of the Cardial ware which reached Portugal: http://en.wikipedia.org/wiki/Cardium_Pottery. So I say that if Hg G came from the Caucasus it was in a pre-Neolithic period, and apparently it did not spread the Kartvelian languages which remain isolate in the area of Georgia, although as we see they have some deep connection with IE (for instance the root mkerd 'breast'-IE kerd 'heart', and ablaut: http://en.wikipedia.org/wiki/Proto-Kartvelian_language), and phonologically they are also quite close to Semitic like in case of uvular stops, velar/uvular fricatives.

Kristiina said...

I have come up with 2 possible etymologies for the word ”wine”: one is North Caucasian and the other one is IE/mixed. In many languages ”bier” is derived from a word meaning drink, e.g.
Slavic pívo - cfr. pivát drink
Pan-European bier - cfr. boir drink
Swedish öl - cfr Irish olaim drink
In Lezgi, the verb drink is qʷa and the masdar form is qun (the Proto-Tsezian construction is χ:ʷi-). Wine could then be a kind of Masdar form of χ:ʷi, i.e. a drink.

The other etymology is the verb ”bend” and the word for ”withe”, and in this case the etymology would not be ”a drink” but ”a climbing plant”:
Sanskrit vinam bend, vaṅkya curved, pliant
Polish ginać bend
Italian vinco withe
Mari vəńək withe
Kabardian gwaana branch

Kristiina said...

Mike, according to Jaakko Häkkinen, the oldest IE words in the Uralic language family are dated to 2800 BC.

William O'Brien observes that the oldest mining period in Kargaly is dated to 3800-2500 BC and the more recent phase to 1900-1300 BC. There are at least two metallurgy words that can be reconstructed to proto-Uralic: *äsa and *wäśka both meaning different metals. On the basis of this, I would suggest that the Uralic proto-language developed in the Ural forest area c. 2800 BC.

Jaakko Häkkinen’s map of the area of origin is on the last page of this article http://www.sgr.fi/susa/92/hakkinen.pdf

Nirjhar007 said...

Kristiina,
Thx for your effort! but i just think its simpler to assume a Non-IE source from Near East or Caucasian area where in My Theory IE branches borrowed when they went there from East:).

a said...

Kristiina thank-you
is the information with regards to wine very accurate re- pages 561&563?
https://books.google.ca/books?id=M2aqp2n2mKkC&pg=PA561&lpg=PA561&dq=hittite+laws+grapes&source=bl&ots=POq7zZsgaL&sig=KTDjGgZRsJsI3pwoE_idjDUIbt0&hl=en&sa=X&ei=wrpcVZHaIoGVyATC6YDQBg&ved=0CC4Q6AEwAw#v=onepage&q=hittite%20laws%20grapes&f=false

Nirjhar007
Soma/homa or wine which came first?


Indra is a drinker of the intoxicating beverage soma, whose influence leads him to ...

http://en.wikipedia.org/wiki/Transcaucasia
"ranscaucasia, in particular where modern-day Turkey, Georgia, Armenia and Iran are located, is one of the native areas of the wine-producing vine Vitis vinifera.[14] Some experts speculate that Transcaucasia may be the birthplace of wine production.[15] Archaeological excavations and carbon dating of grape seeds from the area have dated back to 7000–5000 BC.[16] Wine found in Iran has been dated to c. 7400 BC[14] and c. 5000 BC,[17] while wine found in Georgia has been dated to c. 6000 BC.[18][19][20] The earliest winery, dated to c. 4000 BC, was found in Armenia.[14]"

Nirjhar007 said...

Ah i have not much time now at all but Wine and Soma are totally different things and the word for Wine in Indo-Aryan is Sura as shown from a different root .
And the argument for the ''loss of word wine'' in Indo-Iranian and Tocharian proposed in the book is in one word laughable:), ok now i go keep the discussion heated guys;)...

a said...

@Nirjhar007
I respect your opinion. Can you show a connection with Sura and Sumerian and or Akkadian god's?
http://www.wineintro.com/history/religion/sumerian.html

a said...

@Nirjhar007

Any comments?
Anthony 2007, p. 454-455.

"At least 383 non-Indo-European words were borrowed from this culture, including the god Indra and the ritual drink Soma.[8] According to Anthony,"

http://en.wikipedia.org/wiki/Soma

Kristiina said...

Nirjhar, what I proposed is not against what you say. My first proposal was a Lezgi or a Proto-Tsezian word, i.e. I proposed a non-IE origin in the Caucasian area. My second proposal is not necessarily IE either, although the root is attested in several IE languages.

Mallory and Adams list the IE root *uendh as a possible etymology for the word wine: Old Norse vinda, Old High German wintan, Greek kannathron [?], Armenian gind (ring), Old Indian vandhura (basket). However, I agree that a north European origin is not very plausible for this word and Greek and Armenian cognate words are not so strong.

Kristiina said...

I forgot to add that the Persian word wine "sharāb" should be derived from the arabic word drink "šariba".

Grey said...

Not ure how this fits with my pet theory which was

group of people (ENF?) somewhere near apple valley (almaty) in S. Kazakhstan domesticate goats and spread west and south as foot pastoralists (not necessarily full farmers yet) and *become* near eastern, balkan etc displacing the existing foot HGs in those regions.

one branch splits off and heads onto the steppe north of the Caspian where they tussle with mounted HGs and lose, one part retreats south into the Caucasus while the other becomes the commoner layer in a stratified PIE society with the mounted HGs as the upper layer.

Grey said...

Uralic and copper

If the giant copper field around the south Urals was a starting point for copper working and so Uralic speaking people became early copper workers and spread along the trade routes then could the Uralic influence in other languages not be a substrate but a (whatever the word is).

How many Uralic loan words are connected to metal working or mining, trading etc?

(If so I'd suggest R1b spread the same way - copper.)

Grey said...

Wine or Soma

Could the derivation of the word(s) have anything to do with fermented apples or pears?

Colin Welling said...

@david

The North Eurasian population that Bomhard is talking about wasn't from anywhere near Bactria, whether he actually knows it or not.

No he isn't, He is talking about linguistic groups. We can't expect linguistic borders to match the genetic borders. For all we know EHG in pontic steppe spoke Caucasian and the Eurasian speakers Bomhard is referring to were a mix of ANE, EEF, and EHG/WHG from Central Asia.

However, I do lean towards PIE being strongly influenced by EHG on the steppe, in pre PIE times, along with EEF related people who came from the south bringing the Caucasian linguistic influence. In that sense, I agree with you about the EHG and EEF divide. But i don't for one second think that EHG as a whole spoke this ancient derived language that is somehow closer to PIE than Caucasian. There is no way some unified language would exist in EHG who are wide spread and are probably evolving their languages so quickly, with their hunter gather lifestyles, such that their language "memory" is weak.

I must admit that I am highly skeptical of a PPIE language even existing. I think PIE evolved either by internal evolution or by a mixing of north and south. Obviously PIE is real its descendants are around today but think PIE was a unique situation of huge population expansion which helped preserve much of the structure. Thats something I wouldn't expect of hunter gatherers wandering the steppe area.

Davidski said...

Read the paper again.

It's mainly trying to explain why Indo-European looks like such a North Eurasian language without actually being part of a clade with Uralic.

The stuff about Nichols' hypothesis and Central Asia is only relevant to the point that this North Eurasian group ancestral to Indo-European came from the east.

This of course makes perfect sense, because it seems that ANE and/or EHG arrived in Eastern Europe from the east. There's a big fucking map in Haak et al. strongly hinting at this, which basically looks like it could well have come from Bomhard's paper.

Here it is...

https://drive.google.com/file/d/0B9o3EYTdM8lQTFZCWjFCR2lFODg/view?usp=sharing

What other Central Asian group do you mean? The imaginary one? Bomhard doesn't know or care about ANE, EEF, and EHG/WHG. But he's obviously talking about a NORTH EURASIAN people...you know, like from the north.

Mike Thomas said...

David

"This of course makes perfect sense, because it seems that ANE and/or EHG arrived in Eastern Europe from the east".

Yes, during the last glacial maximum I bet. So what's that got to do with PIE in the copper age ?

Davidski said...

No Mike, they were probably in the Altai-Sayan Mammoth steppe refuge during the LGM.

Now, with that in mind, and also that the paper is from a conference on the Indo-Uralic hypothesis, read this part of the abstract again, this time more carefully.

"Evidence will be presented to demonstrate that Proto-Indo-European is the result of the imposition of a Eurasiatic language — to use Greenberg’s term — on a population speaking one or more primordial Northwest Caucasian languages."

Now look at this map again.

https://drive.google.com/file/d/0B9o3EYTdM8lQTFZCWjFCR2lFODg/view?usp=sharing

Still doesn't click?

It should, because you're not going to sound very convincing if you claim that these Eurasiatic Uralic-like people weren't the ancestors of the EHG, but some mysterious, as yet undocumented farmers who came from Afghanistan or somewhere nearby.

terryt said...

"Proto-Indo-European is the result of the imposition of a Eurasiatic language — to use Greenberg’s term — on a population speaking one or more primordial Northwest Caucasian languages".

To me that makes complete sense and explains a considerable number of problems regarding the 'origin' of the Indo-European languages.

Davidski said...

Yes, it makes perfect sense to me too, as I've already said above.

Can you elaborate why it makes sense to you? Because so far I feel like I've ventured into the Twilight Zone by taking part in the discussion here.

Mike Thomas said...

Dave

What are you getting at ??

EHG/ ANE was present already in Mesolithic Sweden.

So regardless of where ANE hid during the ice age, it was already present in Europe by say 16 or 14, or 10 ky BP, and has nothing to do with PPIE.

On the other hand, c. 5 y BP, we see clear evidence of some central Asian component arriving on the steppe, which happens to coincide with the broad dating of PIE.

Kristiina said...

A, it is true that Gamkrelidze is not very specific with respect to ”wine”, as he goes: ”the ancient derivatives from PIE *wei- twist, plait, the source of the words for grape and wine in Indo-European, are used in Indo-Iranian only for various species of vines and trailing plants, with no obvious connection to wine and grapes.”

According to Mallory PIE *wei is probably this:
?*ueik- 'bend a pliable object'. [7EW 1130 (*ueik- ~ *ue/g-); Wat 75 (*weik-)}. Lat vincioo 'bind, tie', Lith vykis 'tape-worm', Latv viikt 'bend, fold', Grk eiko 'yield *ye/g-: Olnd viici- 'deceit (?)'; ON viikja 'bend, turn', OE wice 'wych elm'. The relationship between the Germanic forms with final *-g- and the other forms with final *-k- is unclear.

I am not so sure that Gamkrelidze is right when he claims that ”the western Iranian languages preserve traces of the early IndoEuropean viticultural terminology” [if here reference is made to ”wine”], specifically as he goes on that ”in the eastern Iranian languages the ancient terms for wine and grape have been completely replaced. The eastern languages have an entire set of new terms of new terms referring to other beverages prepared from other plants, and the meanings of the ancient viticultural terms have changed.”

I would put my money on an etymology based on a word ”drink”, but nobody knows what happened …

Davidski said...

No Mike, what happens 5,000 years ago is that a particular group of EHG start to impose their language on groups from the North Caucasus, who may or may not have some sort of Central Asian admixture.

The data we have at the moment shows nothing more than this. And this just happens to fit very nicely Bomhard's Caucasian substrate hypothesis.

See that's why he calls it the Caucasian substrate hypothesis.

Davidski said...

Actually, not 5,000 years ago, more like 5,000 BCE, because the Khvalynsk people already have some of that Caucasian admixture.

Helgenes50 said...

When do you think that the the Khvalynsk genomes will be available

Davidski said...

I don't know when the Khvalynsk samples will come out. Probably not for a while.

But they're basically like Yamnaya except with 25% of the Armenian-like admixture, as opposed to 50%.

So Mike's claims about a migration from Central Asia onto the steppe 5,000 years ago is just nonsense.

Alberto said...

David, I understand that your theory makes sense to you. But you're trying to match your theory with Bomhard's one by twisting what he says. And that's wrong.

Bomhard is very explicit about the place of origin and time of arrival of this population to the eastern shores of the Black Sea. When he says they came from Central Asia (Bactria-Sogdiana) and started arriving from 5000 BC (and continued coming over a long period of time, probably 5000-3500 BC), he means that exactly. He doesn't mean that they came from Siberia 5000-8000 years earlier.

And there are very specific (linguistic, not genetic) reasons for it.

So stop trying to change Bomhard's theory to match yours. They are different. Just say that you disagree with him, expose your arguments and be done with it.

Re:ancient DNA, a population with a Central Asian genetic profile did start to arrive to the Blask Sea around that time (or slightly later, we don't know yet exactly). Why do you keep saying that this is a mysterious, non-existent and unnecessary population?

Davidski said...

Alberto,

The mixing process between the Georgian or Armenian-like population and EHG started around 5,000 BCE, during the Khvalynsk period or before.

That means the relevant migrations had to have happened earlier, because different human groups don't just start mixing when they meet.

So where in any of the data are you seeing a migration here from Central Asia 5,000 BCE or later?

There's nothing like that in any of the data. All I'm seeing is mixing between Eurasiatic EHG and Northwest Caucasians.

That's what Bomhard is seeing too, whether he knows it or not. If he thinks that his Eurasiatics came from Bactria then he's obviously wrong.

Mike Thomas said...

I thi k it's early to tell
Ie I don't think the mixing must have happened in isolated, discreet events
Yes it might have begun c. 7 ky BP, but accelerated by 5 ky BP. Whether this has any bearing on PPIE directly is yet another issue

Davidski said...

Holy shit, you guys obviously aren't Sherlock Holmes fans.

Bomhard says there are two groups mixing on the Eneolithic steppe: the Northwest Caucasians and the Eurasiatics, who imposed their PPIE language on them.

Ancient DNA says there are two groups mixing on the Eneolithic steppe: a Georgian-like population with a high level of the teal component and Near Eastern/Caucasus mtDNA lineages, and the EHG.

So who are the Northwest Caucasians and who are the Eurasiatics? Just think about it for a few seconds...

Mike Thomas said...

Well I like the one with Rob Downey Jr.

I see what yr getting at; but your making the assumption that the (genetically) "Georgian like" pop spoke Caucasian, whilst the EHG spoke Eurasiatic. A fair assumption.

But that's flipping Bornholms Central thesis 180'. ie assuming that it was actualy a Caucasian group (Georgian like) intruding into Eurasiatic (EHG).

Nirjhar007 said...

Ah finally i have some time to wander in Web.
a,
//Can you show a connection with Sura and Sumerian and or Akkadian god's?//
I would try words as gods will be rather vague:D! i will check.
''"At least 383 non-Indo-European words were borrowed from this culture, including the god Indra and the ritual drink Soma.[8] According to Anthony,"''
Total nonsense and though many Etymologist exist for indra there are some better ones like this-
http://starling.rinet.ru/cgi-bin/response.cgi?root=config&morpho=0&basename=%5Cdata%5Cie%5Cpiet&first=1&off=&text_proto=&method_proto=substring&ic_proto=on&text_meaning=&method_meaning=substring&ic_meaning=on&text_hitt=&method_hitt=substring&ic_hitt=on&text_tokh=&method_tokh=substring&ic_tokh=on&text_ind=indra&method_ind=substring&ic_ind=on&text_avest=&method_avest=substring&ic_avest=on&text_iran=&method_iran=substring&ic_iran=on&text_arm=&method_arm=substring&ic_arm=on&text_greek=&method_greek=substring&ic_greek=on&text_slav=&method_slav=substring&ic_slav=on&text_balt=&method_balt=substring&ic_balt=on&text_germ=&method_germ=substring&ic_germ=on&text_lat=&method_lat=substring&ic_lat=on&text_ital=&method_ital=substring&ic_ital=on&text_celt=&method_celt=substring&ic_celt=on&text_alb=&method_alb=substring&ic_alb=on&text_rusmean=&method_rusmean=substring&ic_rusmean=on&text_refer=&method_refer=substring&ic_refer=on&text_comment=&method_comment=substring&ic_comment=on&text_any=&method_any=substring&sort=proto&ic_any=on
Soma surely has IE root as that book you linked refers,

Nirjhar007 said...

Kristiina,
Yes of course its just that some academics for Some reason just don't accept the obvious the word Wine most likely has non-IE origins, i don't know why they are so stubborn on this.
//I am not so sure that Gamkrelidze is right when he claims...//
I have also so far not able to find the words in Indo-Iranian when it is stated that ”the ancient derivatives from PIE *wei- twist, plait, the source of the words for grape and wine in Indo-European, are used in Indo-Iranian only for various species of vines and trailing plants, with no obvious connection to wine and grapes.”

Nirjhar007 said...

Grey,
Yes I already told you that i like your ideas on Copper I think BMAC went close to Urals areas in Arkaim and Sintasta to yield more copper to export in south this is how i think the Indo-Iranian elements came to the Uralic.
''Could the derivation of the word(s) have anything to do with fermented apples or pears?''
Soma and wine are different stuff totally, Soma is most likely the Ephedra and Sura which is Wine according to the Monier-Williams Dictionary-
surA f. (ifc. also %{sura} n. ; prob. fr. 3 %{su} , `" to distil "' , and not connected with %{sura} , `" a god "') spirituous liquor , wine (in ancient times `" a kind of beer "') ; spirituous liquor (personified as a daughter of Varun2a produced at the churning of the ocean) RV. &c. &c. ; water Naigh. i , 12 ; a drinking vessel L. ; a snake L.

Grey said...

Nirjhar

Could be. If Uralic did spread like that then it would explain why there weren't any river names - more of a superstrate(?) than substrate.

.

Davidski

Maybe I am misunderstanding something but the PPIE being a mixture including a language now spoken in the NW Caucasus doesn't mean that's where they were at the time the language mixture took place (or only there at the time).

If the losing side were herders (higher pop. density) and the winning side originally HGs (lower pop. density usually) but they had a winning advantage in horses then I'd say the actual conflict was more likely to have been on the flat and the losers retreated to the mountains.

a said...

@Kristiina said...

A, it is true that Gamkrelidze is not very specific with respect to ”wine”, as he goes: ”the ancient derivatives from PIE *wei- twist, plait, the source of the words for grape and wine in Indo-European, are used in Indo-Iranian only for various species of vines and trailing plants, with no obvious connection to wine and grapes.”

According to Mallory PIE *wei is probably this:
?*ueik- 'bend a pliable object'. [7EW 1130 (*ueik- ~ *ue/g-); Wat 75 (*weik-)}. Lat vincioo 'bind, tie', Lith vykis 'tape-worm', Latv viikt 'bend, fold', Grk eiko 'yield *ye/g-: Olnd viici- 'deceit (?)'; ON viikja 'bend, turn', OE wice 'wych elm'. The relationship between the Germanic forms with final *-g- and the other forms with final *-k- is unclear.

Thank you.

Yes and also he qualifies PIE words>> as one from Anatolian and one from I.E>

For example in following video @5:10 he defines.
https://www.youtube.com/watch?v=oERPeM1A_GM

Anatolian in first split from PIE. Armenian is separate branch.Correct?

Plus, now you have another possible clue with Kartvelian. Just how old is proto Kartvelian and the languages discussed in this paper; comparing PIE origin?

tew said...

@Collin
quote: "I must admit that I am highly skeptical of a PPIE language even existing"


There must be a bit of confusion there. Assuming a standard comparative-method-based historical linguistics model, a pre-proto-language is simply a reconstructable stage of a proto-language that is considered older than the most recent common ancestor of its daughter languages. While the latter is usually reconstructed on the basis of attested data in the descendants, pre-proto-languages can be partially reconstructed by relying on internal evidence in the proto-language, such as morphological irregularities, lexical variants, loanwords in neighbouring [proto-]languages, etc.

By extension, and more generally, a pre-proto-language can be whatever stage one can identify that is still genealogically continuous with a reconstructed proto-language "but not yet it", i.e., before certain changes that one supposes must have occured. Also, we should keep in mind, in standard historical linguistics, languages canNOT be considered simply "mixed", excluding very few cases, usually artificial. There is always a single genealogical predecessor, whether identifiable or not.

The conclusion is that, by definition, PIE must have come from PPIE, whatever may have happened between the two stages. So what your and others here are denying is a specific hypothesis about how we can get from PPIE to PIE, not the existence of a PPIE itself, which is something of a logical mistake, given the starting assumptions.

A different objection, of course, would be to deny the possibility of reconstructing a PPIE stage because of insufficient data or too great of a time depth, but apparently that's not what is at stake here.

a said...

@ Nirjhar007

Having 7500+/- samples of Kareilan R1a and Samara R1b being so similar at a autosomal level shows some derived continuity on the Steppe. Then when you add the fact that Ma1-R* at 24k has similar component/s, is also quite informative, again showing continuity setting the stage for a southern migration/mixing. IMO it is reasonable to think that Samara Sok river R1b H.G. 7650+/-B.P. spoke a language closely related to his 5000+/- B.P. R1b-z2103+ relatives/descendant's found in Yamnaya kurgans. Who then went on to expand/migrate and mix with North Caucasus, type populations.

Nirjhar007 said...

I don't trust Autosomes much:) and we should wait to see the aDNA from important areas of Asia and even Europe and we can't correlate dna with Language unless we have Archaeological and linguistic correspondence also.

Nirjhar007 said...

ah yes i have checked for Wine like word in Sumerian-Akkadian here-
http://psd.museum.upenn.edu/epsd1/nepsd-frame.html
Its not there.

a said...

@Nirjhar007
That's your choice. I look at all the data. Here is a little test using K36 amongst all samples below. 1 of these is Punjabi 1 is Polish and 2 are from Yamnaya can you tell which ones are which?


Amerindian 0.84%
Armenian 1.04%
East_Central_Euro 2.91%
Eastern_Euro 4.74%
North_Atlantic 2.14%
North_Caucasian 6.88%
North_Sea 1.58%
South_Asian 39.69%
South_Central_Asian 37.26%
Volga-Ural 2.01%
West_Caucasian 0.90%

East_Central_Euro 22.93%
Eastern_Euro 17.15%
North_Sea 11.06%
North_Atlantic 9.18%
Fennoscandian 8.36%
Central_Euro 7.12%
Italian 6.17%
Iberian 4.94%
French 4.80%
North_Caucasian 2.75%
East_Balkan 2.53%
Volga-Ural 2.10%
West_Med 0.90%

Fennoscandian 18.07%
North_Caucasian 17.09%
Eastern_Euro 14.50%
North_Sea 12.89%
South_Central_Asian 12.51%
East_Central_Euro 10.17%
North_Atlantic 7.51%
Volga-Ural 3.51%
Central_Euro 1.90%
Amerindian 1.85%

North_Caucasian 20.00%
Eastern_Euro 16.34%
North_Atlantic 15.65%
North_Sea 14.04%
Fennoscandian 11.09%
South_Central_Asian 10.74%
Volga-Ural 5.36%
East_Central_Euro 5.41%
Amerindian 1.36%

Nirjhar007 said...

Uh Last 2 are from Yamnaya? 1st Punjabi and second Polish?:I

a said...

Yes. I wanted to make it a little more challenging. Like which ones are R1a-93 and R1b Z2013+. As you probably already know- both exist within Steppe/BMAC regions and Bashkir/Yamnaya regions- projects. However, you guessed right. I agree with you, and really nice/informative to getting more genomes from areas/regions not yet covered.

terryt said...

"so far I feel like I've ventured into the Twilight Zone by taking part in the discussion here".

I think you have. But in explanation I read the 'Eurasiatic language' hypothesis years ago and have had it in the back of my mind ever since. Greenburg suggested that the majority of Amerind languages originally formed part of it, although since then most have claimed such languages are too diverse to fit. I disagree, especially when we consider the diversity within the proposed Eurasiatic language family. Besides which MA-1 offers some level of support for an early presence of an appropriate ancestor population in Central Asia.

Further to the idea of a Indo-European language over an (or more than one) original substrate language is the difficulty in sorting out the actual time of branching for the language groups. They diversify much earlier than they 'should have'.

"Maybe I am misunderstanding something but the PPIE being a mixture including a language now spoken in the NW Caucasus doesn't mean that's where they were at the time the language mixture took place (or only there at the time)".

NW Caucasian could have been much more widespread at the time. The fact it is now confined to a mountainous region even supports such an idea. Davidski pointed out that, 'Bomhard says there are two groups mixing on the Eneolithic steppe: the Northwest Caucasians and the Eurasiatics, who imposed their PPIE language on them'.

tew said...

Whether Eurasiatic "makes sense" from a genetic point of view or not is actually immaterial, since it has not been demonstrated using mainstream linguistic methodology and as such remains a fringe theory, which can turn out to be true in the future but is still highly speculative. And genetically it would make perfect sense for Hungarians, Slovaks and Romanians to speak closely related languages but we know that is far from true.

terryt said...

"And genetically it would make perfect sense for Hungarians, Slovaks and Romanians to speak closely related languages but we know that is far from true".

The language a people speak has very little to do with genetics. It is a matter of history. A hugely diverse genetic collection speaks English these days. In many cases people who are the same genetically and tribally speak, say, English or French depending on what side of a political border they live. The 'Eurasiatic' language group, if it ever really existed, is very ancient and so connections would be extremely difficult to discern today. It remain a hypothesis although to me it seems a very likely hypothesis.

Mike Thomas said...

Tew

Well put. I have argued that ad nauseum here . However, this common sense is often lost on amateur and academic genealogists, who see unfortunately see genetics as a means to an end to "prove" their story, for whatever reason...

And I agree that it's sloppy terminology, and just a general lack of nuanced knowledge by bloggers, but what the paper discusses here is substratum effects and language shifting, which is very different to a mixed language (which are rare)

tew said...

"although to me it seems a very likely hypothesis"

That is the whole point. Why does it seem likely to you (or to others who commented to the same effect)? To linguists the evidence as it stands is not compelling since it is not adduced using mainstream accepted methodology, period.

Davidski said...

I'm looking at Eurasiatic here as a geographic term more than anything.

It doesn't matter if Indo-European and, say, Uralic–Yukaghir are closely related or not, what matters is that both essentially look like North Eurasian languages, which is what is clearly stated in the paper above.

Now, considering we already have North Eurasians on the steppe, known as Eastern hunter-gatherers (EHG), then why the hell do we need a migration from Bactria, of all places, to bring more North Eurasians to the steppe?

Early steppe people look like a mixture of EHG and Northwest Caucasians, and that's what the Proto-Indo-European language looks like too.

terryt said...

"Why does it seem likely to you (or to others who commented to the same effect)? To linguists the evidence as it stands is not compelling since it is not adduced using mainstream accepted methodology, period".

It has always interested me that Greenburg's original proposal has come in for so much criticism, and yet his partitioning of African languages into groups by using the same methodology is still largely accepted by the mainstream. I see no real reason why the hypothesis must be 'wrong'. Perhaps it is just that people are more possessive over the exclusivity of their own language rather than the African languages with which they have no emotional connection.

tew said...

@Mike
Yes, I see now the main issue seems to be about strata. Still, as you said, there is a lot of terminological confusion throughout this discussion, particularly concerning the misused word "mixed".

tew said...

@terry
Well, first of all, Greenberg's (not Greenburg) mass comparison is considered acceptable as a starting point, but an incomplete alternative to the traditional comparative method when it comes to drawing conclusions. Such analyses can sometimes be confirmed and sometimes not, but the hard work of using the comparative method with the data at hand and finding regularities must be done before that. So it is not that the hypothesis "must be wrong" at all. It is simply that a) there is insufficient data, according to technical criteria used by most professionals in the field; b) the methodology employed by proponents is not acceptable according to technical criteria used by most professionals in the field. Simple as that.

Now, the reason why his classification of African languages was well received at the time while his other comparative work was not is partially sociological, indeed, but of a very different sort. When it comes to African languages, Greenberg had far less competition (still has), with even more amateurish analyses, some of which used to classify languages more on the basis of cultural or racial criteria than any technical data. The same cannot be said about the languages under some of his other macro-families when he proposed them.

Also, you may not be aware of this, but his classification of African languages has in fact been revised and many changes have occurred. For example, the old "Khoisan" group is not considered a valid language family anymore, being replaced by as many as 5 or 6 different families and isolates; and there are huge doubts about the reality of Nilo-Saharan and Niger-Congo (which have actually been broken up in many recent classifications).

Anyway, few people would say that Greenberg and his intellectual descendants in the Eurasiatic, etc. camp are complete crackpots, really. His work in typology (which non linguists mostly ignore) is well respected and still today an important reference. So it is not like this is a crusade of some sort. I see very little emotion involved in the fact that say an American or Dutch linguist would deny that you can definitively lump everything from Chukotko-Kamchatkan to Dravidian to Etruscan based on a few lexical resemblances that have not been studied using sound standard methodology; however, you are free to believe any conspiracy theory.

tew said...

@David
"I'm looking at Eurasiatic here as a geographic term more than anything."

Ok, but that seems like a category mistake to me, and one that has curious consequences. Consider this: to most proponents of megalocomparison, "Eurasiatic" includes everything from Dravidian to Tungusic to Indo-European to Japanese (Bomhard's vague, never defined term "Northern Eurasia" somehow accommodates all that).

Northwest and North Caucasian languages, on the other hand, are assigned by them to another macrofamily, Dene-Caucasian or Sino-Caucasian, which includes everything from Basque to Burushaski to Sinnitic to Na-Dene (that's why he couldn't ever accept Colarusso's data as evidence of genealogical relation, only of contact).

As you can see, there is no geographical consistency, and I suspect one would be hard-pressed to find any consistency in terms of genetic similarity, either, however remote that may be. That is a big problem, since to use the fact that PIE appears to be the result of a Sino-Caucasian substrate on an Eurasiatic language as an argument in genetics, you would have to somehow make sense of these linguistic groups genetically. In other words, I doubt the hypotheses at hand are of much use to you.

BTW, the more modest claims that there is a North-West Caucasian substrate in PIE and that, if true, this can be taken as support for the Kurgan Urheimat theory, has been made many times before by linguists who do not subscribe to Bomhard's megalocomparison, and nobody needs to believe in the reality of Eurasiatic (nor to develop grand theories based on it) to accept them. But I can see how those reasonable claims are perhaps too modest for some people.

Mike Thomas said...

Tew
Frankly, I think theories like Nostratic are "stretched " at best and contrived at worst by a few trying to make a name or a stance.

These are then grasped by other disciplines in an attempt to add further "evidence" to their convictions.

Aside from whether they are linguistically (comparativist) sound or not, I find it hard to fathom that some Eurasiatic language could only have been spoken exclusively by "EHG people". Such a notion appears utterly divorced from any kind of sociolinguistic reality, on multiple levels. It pleads that language is genetically conditioned, and some how sticks to the world Atlas (ie North of the 41st parallel).

Arch Hades said...

I'm curious to what the Y chromosomal lineages would be of this Caucuses-like population that migrated into the steppe and mixed with the EHGs, that is if we assume the EHGs or at least ANEs are the fathers of the R1 clade.

We need more genomes and samples of the Yamnaya horizon.

Davidski said...

tew,

As per the paper abstract, Bomhard appears to be using Joseph Greenberg's definition of "Eurasiatic", which looks specifically North Eurasian, and fits with his statements that the PPIE Eurasiatics were North Eurasians.

Altaic
Chukchi-Kamchatkan
Eskimo–Aleut
Etruscan
Indo-European
Korean-Japanese-Ainu
Nivkh
Uralic–Yukaghir

Arch,

Maybe only the Caucasian women moved into the steppe? This is possible if the two groups practiced exogamy with each other.

I'm not saying that we won't ever find any other Y-haplogroups on the Eneolithic/Early Bronze Age steppe apart from those typical of EHG, but I think they'll prove to be very rare and hard to locate.

terryt said...

"I find it hard to fathom that some Eurasiatic language could only have been spoken exclusively by "EHG people".

Although MA-1's Y-DNA is now extinct its close relations are extremely widespread through Central Eurasia from Europe to America. To me it seems obvious that at some time a single population containing Y-DNA R and Q (K2b2 or P if you prefer) was able to expand widely through Central Asia at some time after earlier populations had become extinct or forced south by deteriorating climate. It is reasonable to suppose that this population, represented by MA-1 at least, spoke a group of closely related languages which diversified once they had spread geographically.

"Northwest and North Caucasian languages, on the other hand, are assigned by them to another macrofamily, Dene-Caucasian or Sino-Caucasian, which includes everything from Basque to Burushaski to Sinnitic to Na-Dene"

My best guess is that eastern haplogroups (N, O, C2) and languages later spread over this K2b2 population. Sino-Tibetan and Na-Dene are eastern languages. Although it is difficult to make a geographic connection between Basque and Na-Dene Burushaski sits halfway. And the Ket languages are related to Na-Dene evidently.

Mike Thomas said...

No Terry

That is a naive assumption. Even if one accepts that MA-1, EHG and native Americans were in part derived from one met-population (an extremely simplistic view in itself), to claim that they further must have spoke a set of closely related dialects in terms of something realistically reconstructable is not only utterly devoid of sociolinguistic grounding, but verges on pseudo-science. Not least, To claim any coherent continuity -linguistic or cultural between 25 kya and 13 kya, let alone 7 kya is beyond reason.

Kristiina said...

I am probably the person to be blamed for using the word ”mixed”. I have been using it in a general way without any clear definition or theory. However, I insist that languages change and it is probably due to new speakers not using correctly the forms of the language they use. In any case, I do not believe in a mere internal logic which would lead to changes.

I still insist that for example the fact that English is very different from Proto-Indo-European is due to its evolution in the mouths of non-natives: recently, a contrast between the celtic language(s) and the anglic language; before that, whatever ancient form of IE interacting with a northern proto-European language to form the Germanic proto-language; before that a possible Germanic-Balto Slavic form of IE interacting with a Proto-European language; and so on.

The main characteristics of Proto-Indo-European are:
variable lexical stress;
eight or nine cases (nominative, accusative, genitive, dative, instrumental, ablative, locative, vocative, and possibly a directive or allative);
three genders: masculine, feminine and neuter
complex system of morphology that included inflections (suffixing of roots) and ablaut (vowel alterations, as in sing, sang, sung).
nouns used a complicated system of declension and verbs used a similarly complicated system of conjugation.

The English language has lost almost everything. The only thing that remains is ablaut.

Grey said...

terryt

"NW Caucasian could have been much more widespread at the time."

Yes I was just stressing that point.

If PPIE was a North Eurasian language imposing itself on a substrate similar to NW Caucasusian then (at least on the face of it) that imposition wouldn't have taken place in the NW Caucasus itself or they'd be speaking the composite language there.

So the language event most likely took place on the steppe north of the Caucasus itself.

Secondly there's no reason to assume the population now in the NW Caucasus are necessarily indigenous to that region. They might be but they are just or more likely to have come from somewhere else onto the steppe: either from the south through the Caucasus or around the Caspian/Black Seas and then retreated back into the NW Caucasus.

So basically there's no need for EHG to have come from anywhere and it seems more likely to me they were there already and it was this second population who moved onto the steppe and retreated again.

Simon_W said...

Mike Thomas: "we must account for langauge contacts in Anatolia for its (Hittite's) apparently queer position. Indeed, all other Anatolian languages (Luvian, Palaic) are entirely within normal IE variation."

What sources are you drawing upon? As far as I know all Anatolian languages are closely related to each other.

Simon_W said...

I think you fell prey to a misunderstanding. Because Hittite is by far the best attested language of the Anatolian family, linguists often speak of Hittite when they actually mean Anatolian in general.

Simon_W said...

From Wikipedia:
Hittite preserves some very archaic features lost in other Indo-European languages. For example, Hittite has retained two of three laryngeals (*h₂ and *h₃ word-initially). These sounds, whose existence had been hypothesized by Ferdinand de Saussure on the basis of vowel quality in other Indo-European languages in 1879, were not preserved as separate sounds in any attested Indo-European language until the discovery of Hittite. In Hittite, this phoneme is written as ḫ. Hittite, as well as most other Anatolian languages, differs in this respect from any other Indo-European language, and the discovery of laryngeals in Hittite was a remarkable confirmation of Saussure's hypothesis.

Simon_W said...

A very useful summary on the numerous PIE-Uralic similarities:

http://en.wikipedia.org/wiki/Indo-Uralic_languages

In my opinion it's not really that crucial if IE and Uralic are genetically related, that is, if the Indo-Uralic hypothesis is true, or if the correspondences are merely the consequence of contact and Sprachbund. Even if the latter mechanisms are the whole explanation they would presuppose very longstanding and intimate contacts with clear consequences for the location of the PIE homeland.

Certainly it's possible to stay sceptical despite all of this, but it has to be said, really compelling arguments don't exist, there is no killer argument so to say, and I doubt there can be more conclusive evidence of a different category locating the PIE homeland than its relationships to other language families.

Mike Thomas said...

@ Simon W

Im very well aware that there are more than one Anatolian languages (I've clearly mentioned them) .

My point was - much of Hittite's divergence could be due to substratum effects and language contact with other non-IE languages in the Near East, rather than solely due to how "old" it is.

I further highlighted that all other Anatolian languages look very much like "normal" IE languages, ie non-deviant. So all in all, Anatolian languages aren't that peculiar ...

Further, "Anatolian" languages are rather diverse internally, so much so that Ivanov actually argued that they are not a true family. We needn't agree or disagree with his proposal, but his point that they less internally related to each other than are Celtic, italic and certainly slavic languages is definitely correct.

I wholly agree language relationships are useful clues as to where IE originated: somewhere in north-west eurasia is the only honest answer we can derive, and contrary to what claims have been made by palaeolinguistics, this needn't have been exclusively within a steppe environment; for reasons I've already explained numerous times.

Kristiina said...

Mike, I wholly agree with you when you say that "my point was - much of Hittite's divergence could be due to substratum effects and language contact with other non-IE languages in the Near East, rather than solely due to how "old" it is."

If you compared for example English and Russian with Sanskrit without any knowledge of their internal history, you would say that English is more deviant and therefore conclude that it diverged from Proto-IE before Russian.

Moreover, with many languages it is challenging that we cannot be sure of the original form and our knowledge of the history of a language is usually not complete.

terryt said...

"That is a naive assumption. Even if one accepts that MA-1, EHG and native Americans were in part derived from one met-population (an extremely simplistic view in itself), to claim that they further must have spoke a set of closely related dialects in terms of something realistically reconstructable is not only utterly devoid of sociolinguistic grounding, but verges on pseudo-science".

You would be correct if the expansion north had been by multiple separate populations spread along the whole northern margin of populated Eurasia. But that is not what seems to have happened. In fact the movement into previously uninhabited regions would have required the development of improved technologies. Closely sewn skin clothing would have been essential, as well as improved mega-fauna hunting skills. It is most probable that the combination of factors happened in a single region on the northern margin of human habitation. That population, containing Y-DNAs Q and R, was able to expand into the wide region of Eurasia that had become uninhabited, and had the region to themselves for a considerable time. That population can only have at least started out with a single language and culture.

Later populations from west (containing apparently Y-DNA I) and east (containing Y-DNAs C2 and N) adopted and improved the technology and began their own expansions. In the case of N it spread right across Arctic Eurasia much as Q and R had done further south some years before.

"To claim any coherent continuity -linguistic or cultural between 25 kya and 13 kya, let alone 7 kya is beyond reason".

Yes. Any original link would have become very tenuous by now, but that doesn't mean the link is impossible.

Mike Thomas said...

Terry, to entertain such a link one has to venture into the realm of suspended belief and pseudoscience. I suggest u merely consider what happens to language in small bands of hunter gatherers separated not only by large distances, but thousands of years. The cohesive forces of language convergence are minuscule, and the languages spoken would have been divergent *from the outset*, let alone after thousands of years of drift, separation; not to mention extinction. Leading to my next point:..

your version of the population history of North eurasia is an extremely crude, broad brushed one. For instance, the was a big thing called the LGM which peaked c.22-18 kya, and resulted in the near complete abandonment of Siberia and the East european plain. So really, Mal'ta boy- whilst fascinating- was an evolutionary dead end, even though I wholly accept that more of the met-population from whence he came survived further south. And this was not an isolated phenomenon, but had already occurred earlier after the CI (c 40 ky BP). Combine this with further local extinctions, replacements and re-colonisations, then the picture is the complete opposite to that you claim

Finally, not only is your proposal flawed from a theoretical sociolinguistic perspective, and devoid of an appreciation of settlement dynamics, but it also sounds like an early 2000s Nat Geo understanding of what the Y DNA data actually tells us; as if drawing simple little arrows of singular dispersals associated with mega-haplogroups Q and R, which again is completely opposite.

terryt said...

" the was a big thing called the LGM which peaked c.22-18 kya, and resulted in the near complete abandonment of Siberia and the East european plain".

Yes. And several occasions before that. Several times humans have been able to move into regions that became unoccupied for one reason of another. That is presumably why just a few haplogroups became widespread. They were able to expand rapidly.

"So really, Mal'ta boy- whilst fascinating- was an evolutionary dead end"

Less of a dead end than was Ust-Ishim and Kostenki. Mal'ta genetic element is one of those widespread elements that survive in many modern populations. Although his particular Y-DNA haplogroup is extinct close relations are still found in the region.

Mike Thomas said...

Yes, I agree - as stated the metpopulation from which her derives undoubtedly continued, and contributed to formation of EHG.

But in the 10 000 years plus of drift, isolation, admixture, and internal evolution, the two speech communities must have transmutated 100 times over, so much so that they'd basically show no resemblance to each other- even if we make the potentially unwarranted assumption that there was linear language inheritance from Mal'ta to his 40, 000 th removed EHG "ancestor".

Kristiina said...

Terry, you say that "Mal'ta genetic element is one of those widespread elements that survive in many modern populations." Yes, the figure Shaikorth posted shows indeed this MA1 ancestry if I understand the figure correctly. It has survived best in the following populations approximately in the following order: Karitiana, Pima, Colombian, Mayan, East Greenlander, Ket, Selkup, West Greenlander, Koryak, Chukchi, Mari, Shors, Yukaghir, Khakass, Chuvash, Nganassan, Altaian, Tuvinian, Moksha, Erzya, Finnish, Estonian, Kargopol Russian, Lithuanian, Kalash etc.

(http://s28.postimg.org/wuj44klpp/Kostenki14_Ust_Ishim_MA1ls.png)

terryt said...

"But in the 10 000 years plus of drift, isolation, admixture, and internal evolution, the two speech communities must have transmutated 100 times over"

I agree. But it becomes difficult to see language family relationships at more than 10,000 years and we're talking at least 20,000 here. Assuming for now that a proto-Eurasiatic language did exist it would have mutated into the very diverse north Eurasian families that survive today. As well we would have the influence of other langauges on the individual Eurasiatic languages.

"MA1 ancestry if I understand the figure correctly. It has survived best in the following populations approximately in the following order: Karitiana, Pima, Colombian, Mayan, East Greenlander, Ket, Selkup, West Greenlander, Koryak, Chukchi, Mari, Shors, Yukaghir, Khakass, Chuvash, Nganassan, Altaian, Tuvinian, Moksha, Erzya, Finnish, Estonian, Kargopol Russian, Lithuanian, Kalash"

Nearly all (except for Ket and the Amerinds) present in the list of Eurasiatic Davidski provided above:

"Altaic
Chukchi-Kamchatkan
Eskimo–Aleut
Etruscan
Indo-European
Korean-Japanese-Ainu
Nivkh
Uralic–Yukaghir"

And many having some level of Y-DNA Q. Even though any 'original expansion may have been pushed south we know some sort of 'R' had reached the Baikal region by 25,000 years ago and so presumably some branch had reached the coast somewhere near Japan. In fact Japan appears to have been settled by 35 kya. but probably by eastern Y-DNAs such as C and D, probably not by Q.

Simon_W said...

@ Mike Thomas

I know that you were aware that there are more than one Anatolian languages, because as you said, you mentioned them, and I'm not as stupid as to be able to overlook that. My point was that your belief that only Hittite is „awkward“ as you called it, and the other Anatolian languages completely in the range of normal IE variation, is wrong and likely the result of a misunderstanding.

As for Ivanov's claim that the Anatolian languages are very divergent from each other, this is obviously in strong disagreement with what I thought to be correct. Now it would be interesting to know why he arrived at such a conclusion. Because the linguist Craig Melchert arrived at exactly the opposite conclusion: That Luwian and Hittite of the empire period, c. 1400 BC, are sisters about as different as modern Welsh and Irish.

Also consider these excerpts from Anthony's book:
http://jpst.it/z4XH
All Anatolian languages are strongly divergent from other IE languages.

Lastly, take a look at the main isoglosses seperating Hittite from other IE languages:
http://www.bmanuel.org/corling/isoglo-IE++_.jpg

If only Hittite was thus separated, and all the other Anatolian languages not, then, as you rightly said, they wouldn't be part of the same family. But my wikipedia link showed that at least the laryngeal isogloss included all Anatolian languages, and according to Craig Melchert Hittite and Luwian are close relatives. Most other Anatolian languages are from a later period, though, which inevitably means that they are more divergent, simply because they are later.

Simon_W said...

Apart from the above cited grammatical and the few phonological differences, Hittite isn't very awkward, as the comparison of IE consonants shows:

http://webspace.ship.edu/cgboer/indoeuropean.html

Simon_W said...

@ Kristiina

Well, Germanic (English) really does seem to have split from the IE mainstream earlier than Slavic (Russian). This doesn't mean that they split from PIE at different dates, but Germanic lost contact with the remaining center of the IE world at an earlier date, hence it missed the satem innovation for instance.

Simon_W said...

By the way, on an unrelateed note, it's interesting that Anthony (in his recent 2015 paper with Don Ringe) has suggested that the Corded people might be derived from a migration of Tripolye C2 and Usatovo people to southeastern Poland. This would be in agreement with archeological convictions that the Corded Ware was rather from somewhere near Poland than from far eastern Europe.

And in the same paper he suggests that Italo-Celtic is from the Yamnaya wave into the Carpathian Basin, not a new proposal. From a genetic point of view these two proposals seem unlikely at first sight. Because the Corded people from Germany are close to Samara Yamnaya people, whereas BR1 from the Carpathian Basin differs from that quite a bit, because of very low Caucasus-like ancestry and stronger WHG relations.

I would propose that this seeming contradiction may be resolved if BR1 was descended in large part from earlier steppe migrations down the Black Sea into southeastern Europe (and that it hence may be related with the early ancestors of the IE Anatolian branch), and that it was more culturally Yamnaya than genetically.

Simon_W said...

As for Bomhard's paper, I'm highly sceptical that the North Pontic hunter-gatherers (such as the Dneper Donets culture) had ever spoken a Northwest Caucasian language. The most common y-haplogroup in modern speakers of Northwest Caucasian is G2a. R1a is present as a minority, and so is J2. Since the EEF haplogroup G2a was replaced by other haplogroups in most parts of Europe, already starting with the ascent of the I2 variants, it may have some significance that G2a is still the most common haplogroup in Northwest Caucasians.
But more generally, if the PIE homeland was on the PC Steppe, then contacts with Uralics on the one hand and with Northwest Caucasians on the other hand make perfect sense. The influential Maikop culture was most likely Northwest Caucasian speaking. In fact, Maikop, the town, is nowadays Russian speaking, but almost encircled by Adygei.
But others have stressed that there were also (perhaps indirect) contacts between PIE and a Kartvelian-related language, and the ancient DNA evidence seems to suggest a rather basal Georgian or Armenian-like influence from South of the Caucasus.

Davidski said...

Bomhard isn't saying that hunter-gatherers spoke Northwest Caucasian languages, only that someone living on the steppe did.

I think the Northwest Caucasian speakers were women from the Caucasus and surrounds who brought those Near Eastern mtDNA lineages to the steppe (via female exogamy).

Mike Thomas said...

Simon W

I wasn't dichotomozing Hittite vs all other Anatolian languages, anyhow I think we're now on the same page

Re: BR1 - it's possble what you state. But what "earlier" migrations ? Are we resorting to making up history just to breath some CPR into a failed hypothesis ? Maybe we need to just analyse the data from scratch and come up with hypotheses which match rather than vice versa ? I guess we'll just have to wait for data from the NW pontic region

David :

It's more than just women, but im not suggesting necessarily large migrations. I'll happily wait for more ancient data. But on the big scale of things the answer is staring us in the face. One just needs to be well-read enough to know what was happening in west eurasia c. M4

Mike Thomas said...

But I can tell you EHG had little to do with it. They were just caught up in the whole affair

Davidski said...

Yes Mike, Caucasian women colonized the steppe, and they spread Indo-European languages to Western Europe and India.

The EHG R1 men just got caught up in it all.

a said...

Interesting 1 Yamanaya sample was L23*[51+2103]
Yfull experimental tree gives some estimates.
http://www.yfull.com/tree/R1b/
R-Z2103Z2103/CTS1078 * CTS9416 * Z8130/S20902... 7 SNPsformed 6200 ybp, TMRCA 6100 ybp

R-L277.1M12141/Y4364 * Y4369 * Y4368... 8 SNPsformed 6100 ybp, TMRCA 4100 ybp

R-L584FGC14589/Y13369 * FGC14586/Y13370 * FGC14588/Y13368... 2 SNPsformed 6100 ybp, TMRCA 4900 ybp

R-Z2110S12460 * Z2110/CTS7822 * S17864formed 6100 ybp, TMRCA 6100 ybp

Nice variance among the above in both ancient 5k+/- and modern day projects. Georgians with L584+L277 and Ossetians with CTS-7822>predicted 9219.

https://www.familytreedna.com/public/Georgia/default.aspx?section=yresults

https://www.familytreedna.com/public/Ossetian?iframe=yresults

Mike Thomas said...

Dave
I'm not saying that R1b Z2103 came from Georgia, but your explanation of mass wife swapping would imply that none of the local men married a local girl- to account for the 4o% west asian signature in Yamnaya. Seems far fetched

It also misses the point of what was actually happening in the proto-Bronze Age world

a said...

Using MLDP-world9 calc and comparing Caucaus_Parsia component among some ancient/modern r1b-z2103+r1a samples. Interesting results, who knew?

# Population Percent
1 North_and_East_European 53.15
2 Caucaus_Parsia 29.66
3 South_and_West_European 10.81
4 Arctic_Amerind 3.86
5 Indian 1.84



# Population Percent
1 North_and_East_European 50.53
2 South_and_West_European 38.68
3 Caucaus_Parsia 7.67
4 Indian 1.42
5 Middle_East 1.36




# Population Percent
1 North_and_East_European 72.00
2 Arctic_Amerind 11.85
3 Mesoamerican 6.12
4 Caucaus_Parsia 4.20
5 South_and_West_European 2.19
6 Indian 1.79
7 North_Asian 1.35



# Population Percent
1 North_and_East_European 50.15
2 Caucaus_Parsia 30.76
3 South_and_West_European 12.52
4 Arctic_Amerind 3.56
5 Mesoamerican 1.73

Davidski said...

Mike, you should acquaint yourself with the concepts outlined in this abstract. They'll be very useful in understanding what happened on the steppe during the Eneolithic and Bronze Age.

"Ancient DNA reveals patterns of residential continuity and mobility at the onset of the Central European Bronze Age

Mittnik et al.

At the transition from the third to the second millennium BC, the introduction of bronze for the manufacture of tools, weapons and personal ornaments marked a major step in European prehistory. Trade of the metal raw materials and manufactured goods required regular and organized contacts among communities. On the other hand, local population continuity was a prerequisite for the accumulation of wealth, the establishment of enduring social differentiation, and the formation of regional elites.

The archeological record in the Lech Valley in southern Bavaria, Germany, shows a rapid and gapless transition from the Late Neolithic Bell Beaker Phenomenon to the Early Bronze Age. To investigate social and demographic changes associated with the appropriation of the bronze technology, we studied nearly eighty individuals from six burial sites in the region with respect to their maternally inherited mitochondrial DNA (mtDNA).

The results indicate both local genetic continuity spanning the cultural transition, and, following the onset of the Early Bronze Age, a major influx of mtDNA types previously not found in this region. Integrating stable isotope data with the genetic data reveals a picture of a patrilocal society with remarkable mobility in women. While crucial for understanding the change of local demographics, these findings also have implications for the spread of major technological and societal changes across Europe at the beginning of the Bronze Age."

Davidski said...

a,

I've already analyzed the ancient steppe samples.

http://eurogenes.blogspot.com.au/2015/03/bell-beaker-corded-ware-ehg-and-yamnaya.html

http://eurogenes.blogspot.com.au/2015/03/modelling-yamnaya-with-qpadm.html

Mike Thomas said...

David

That looks like a great paper
I never deny exogamy, I just question the extent and exclusivity which you imply
But hey maybe my doubt is incorrect

Simon_W said...

I've now read Johanna Nichols's chapter on „The epicentre of the Indo-European linguistic spread“. Thanks @ Nirjhar for the useful link. In light of the recent discussions about the „Central Asian Teal“ component it's certainly an intriguing theory.

I have made an interesting observation: Nichols claims that the Kartvelians have migrated to their modern seats in the southern Caucasus from somewhere in the east, presumably from the south Caspian, while the Northeast and the Northwest Caucasians were in the Caucasus for a longer time.
Now, looking at the K9 Teal component in the various Caucasus populations, this prediction is precisely verified:

Georgians 58.8%
Abkhasians 58.5%
Lezgins 55.6%
Chechens 52.7%
Adygei 52.3%
North Ossetians 51.1%
Balkars 50.8%
Armenians 48.5%

Just a coincidence? Maybe, but these results are perfectly in line with Nichols's predictions. Also, on the uniparental side, one rather central-southern Georgian group has R1b as its most common y-haplogroup. And by the way, the Abkhasians are Northwest Caucasians who live south of the great Caucasus chain, near the Georgians.

It's also interesting that a broadly „Kartvelian-like“ or rather Teal (?) influence was already present in the Khvalynsk culture. Khvalynsk is actually quite far away from the Near East and the Caucasus. But, in fact, the North Caspian coastal area was a broad belt of semi-desert and desert landscape just like in Central Asia. It would be natural if people from Central Asia migrated there, it was the kind of environment they were adapted and used to. And this North Caspian area was not very far from Khvalynsk. I don't know the archeological details, mabe someone knows it better, but this may be a more reasonable explanation for the admixture in the Khvalynsk HGs than migrations from the Caucasus.

Simon_W said...

I think the main problem for Nichols' theory is that at the crucial time, before the emergence of the Yamnaya culture, Central Asia was covered by the Kelteminar culture, which was rather primitive - hunter-gatherers with no farming and apparently no animal husbandry. Thus Kelteminar had no connection with Mesopotamia and certainly didn't mediate Semitic and Sumerian loanwords to the PIE. I think that's really the coffin nail for her theory.

However, as she has shown, PIE and Kartvelian have the largest number of sharings. And PIE stood more in the sphere of Mesopotamian influence than the North Caucasian languages did. I think an alternative explanation for this situation is that a culture from the Transcaucasus related to Kartvelian really moved to the steppe, largely bypassing the North Caucasians.

capra internetensis said...

@Simon_W

The connection to Mesopotamia wouldn't be via Kel'teminar, but to the south. The Namazga culture of Turkmenistan actually had connections to Sarazm in Tajikistan in the 4th millennium, and in the opposite direction to Iran (and south to the Early Harappan in Pakistan too).

Going from her proposed Bactria-Sogdiana homeland to the Pontic-Caspian region is a problem; as you say, at that time northern Central Asia seems to have been inhabited by foragers, with pastoralism earlier in the west than in the east.

Simon_W said...

@ Capra

True, I had this in mind and didn't want to deny Mesopotamian connections in the south. But the lack of connection with the Pontic-Caspian region is important, because if the PIE origin was in southern central Asia, then the Yamnaya culture cannot have been IE, which is very implausible given the recent findings of Yamnaya-related influence in Europe. Unless it was early IEs who crossed the Caucasus from the south, which isn't quite in line with Nichols' theory.

Simon_W said...

(And which would have a hard time explaining the IE-Uralic similarities.)

PF said...


I’ve previously suggested an Eastern imposition on an early Caucasian language as the basis for PIE. Glad to see this paper supports my contentions (admittedly based almost entirely on intuition rather than extensive scholarship).

What still throws me off are the haplogroup discontinuities. NW Caucuses are exactly where we see some of the highest concentrations of G2a around the globe, and not that much R1a. I mean, shouldn’t it be exactly the other way around?

Now I know we shouldn’t use modern hg distributions to make assertions about the past. That said, I’d love someone to propose an explanation.

Davidski hints that the collision was mediated by exogamous women. I can’t speak to the truth of this but I also can't help but feel skeptical. I mean, when had this type of wife-swapping ever happened!? What are the historical precedents? Is there an obvious and significant influence of Eastern mtDNA in NW Caucasian populations?

It contradicts the later pattern seen all across Europe of R1a exploding and G2a being obliterated. Hard to imagine a scenario in which invaders from the east arrive in the north Caucuses and leave so many G2a heads attached to shoulders, even if these G2a men are for some reason throwing their women at them.

It makes sense if the initial contact was more or less peaceful, and the militaristic style IE invasions began only once the population fully formed (and perhaps gained some sort of major advantage in the process?). Again it’s difficult to believe this... all in all it’s difficult to reconcile everything into a clear picture, even with all the ancient DNA that’s added so much to our understanding…

Davidski said...

R1a is an EHG marker. Why would it or need to be common in the Caucasus?

In regards to wife swapping, or rather female exogamy, it was very common during the Bronze Age.

http://eurogenes.blogspot.com.au/2015/05/high-female-mobility-in-bronze-age.html

PF said...

My question is not so much about the spread of R1a, but rather about the discrepancy between G2a survival in the Caucuses vs. the rest of Europe. It's not like it's even close: the former still retains extremely high levels of G2a, while in the latter the levels approach zero.

I don't see how the G2a demise in core Europe can't be both fairly sudden and military driven. So again, why don't we see this pattern in the Caucuses, presumably where EHG-like eastern invaders first met with Abkhaz-like NW Caucasians and PIE developed?

I think the papers you mention about high female mobility in the Bronze Age have very little to do with exogamous practices between two very different cultures. Women moved around -- so what? If anything the second paper supports my greater claim: that is, the R1a led invasions were both sudden and militaristic.

Why would there be mtDNA continuity into the Early Bronze Age, and almost no Y-DNA continuity??? This doesn't sounds like exogamy or any sort of cultural fusion -- it sounds like a lot of guys got the axe.

My speculative proposition is that Abkhaz-like people had a significantly greater range back then. The "meeting" which led to admixture and PIE probably happened just north of the actual mountains, and the people we see living today are largely descended from survivors who ran uphill, as we see in the Alps and other similar places. That's why they're still G2a-heavy and speak a non-IE language.

Davidski said...

You're confusing two very different issues:

1) the formation of the steppe pastoralists

2) the massive genetic shift across much of Europe.

The first process was the result of patrilineality and patrilocality on the steppe, which were necessary for maintaining wealth and power, and female exogamy, which was used to keep down inbreeding, but also to strengthen long distance alliances.

This is why ancient DNA is showing such strong local Y-chromosome continuity on the steppe, but a significant shift in mtDNA.

We don't know yet what happened in the Caucasus, but probably something similar, except the character of the mtDNA didn't shift as much there possibly because population densities were much higher in the south than on the steppe.

The steppe pastoralists then invaded central Europe. They didn't swap women with the Central European farmers, they actually moved into their territory, took it over, and then started practicing patrilineality and patrilocality and female exogamy among themselves.

This is why Y-DNA G mostly disappeared from Central Europe but not from the Caucasus.