PDA

View Full Version : Languages and Y-DNA lineages



Pages : [1] 2 3

bolek
10-27-2013, 03:31 PM
I am trying to understand the relation between languages and Y-DNA lineages.

From R1a tree we see that R1a-Z645 was split into R1a-Z283 and R1a-Z93.

http://s22.postimg.org/5loobmfrl/ra1_treee.jpg

Both Z93 and Z283 correlate with satem languages. Z93 with Indo-Iranian languages and Z283 with Baltic and Slavic languages. Indo-Iranian, Baltic and Slavic languages are moreover closely related and originated in Eastern Europe, some linguists say that some sort of Indo-Iranian-Baltic-Slavic linguistic continuum existed in the past.
So it seems natural to me to put those R1a lineages on language tree by Warnow as follows:

http://s22.postimg.org/fuh5hg3td/languages_genes.png

I would like to learn about correlations between other languages and Y-DNA lineages. Could you make some proposals.

Michał
10-29-2013, 12:31 PM
I would like to learn about correlations between other languages and Y-DNA lineages. Could you make some proposals.
First, I would like to make an assertion that although the subject of this thread is indeed very interesting for most people investigating the deep ancestry of particular Y-DNA lineages, it is very hard to avoid excessive simplification when constructing a theory that explains most of those very intriguing questions that arise when studying the prehistory some the IE-speaking populations. Nevertheless, I know very few people who would definitely reject any potential relationship between different Y-DNA haplogroups/clades and some IE and non-IE languages, hence I would consider most of such speculations as at least partially justified, if only people were aware of all those limitations that need to be taken into account before jumping to any final conclusions. Another thing that makes such discussions “dangerous” is some strong animosity between the R1a and R1b members who fight for their right to call its own haplogroup a ”true Indo-European” one, which is of course a very childish attitude, to say the least.

I have already presented my point of view on this subject when posting in another thread on this forum a while ago :
http://www.anthrogenica.com/showthread.php?1026-what-is-the-latest-thinking-on-were-R1a-originated/page17&p=10587#post10587

So this time I will only make a brief summary of my presumptions.

The R1a-Z645 branch shows indeed strong association with the potential Satem branch of the IE languages, although we should keep in mind that such hypothetical close relationship between the Balto-Slavic (associated mostly with the R1a-Z280 and R1a-M458 sub-branches of R1a-Z282) and Indo-Iranic (showing clear association with R1a-Z93) languages is rejected by many linguists (mostly by those who consider the Satemization process to be caused by a relatively late areal influence, instead of seeing it as something that indicates a common origin of those two families of languages. Additionally, many linguists are convinced that the ancient Indo-Iranic languages show many similarities to the ancient Greek, thus suggesting some genetic relationship between these languages. Personally, I would rather see it as indicating some early contacts between Proto-Indo-Iranic and another early IE branch that included Pre-Greek, which could have likely happened in the North-Pontic/North-Caspian region (already after the separation between Indo-Iranic and Balto-Slavic).

The R1b-Z2103/Z2105 branch is associated in my scenario with the very hypothetical Graeco-Armenian-Thracoid branch of IE that is currently represented by Greek, Armenian and Albanian.

The R1b-L51/L11 branch is representing another hypothetical branch of IE called “Italo-Celto-Germanic” (not shown as a separate entity in your linguistic tree), where R1b-P312 is associated with Italo-Celtic while R1b-U106 shows strong association with Germanic.

The association of some early separated R1b species (like R1b-M73 or some unknown R1b(xM239) subclades) with the Tocharian branch of IE is not considered by me as well supported, but I would, nevertheless, include this presumption to my hypothesis.

In the case of the Anatolian family of IE languages, I would assume that their ancestors were initially also associated with some early separated R1b subclades, like R1b-L389(xP297) and maybe R1b-V88. However, this was long before these Pre-Anatolians came to Anatolia, and this relationship has been lost due to some intensive interbreeding with some neighbouring South Asian and Middle Eastern non-R1b populations.

Silesian
10-29-2013, 01:56 PM
R1a & R1b examples, in regions of interest for myself, Armenian Iranian Anatolian- plateau, Himalayan plateau, Crete plateau.

Grugni et al study. http://www.ncbi.nlm.nih.gov/pubmed/22815981
http://dienekes.blogspot.ca/2012/07/huge-study-on-y-chromosome-variation-in.html region around Armenia ancient Mede[ substrate] with also possible ancient Zoarostian/Avesta.

Himalaya Plateau. Nepal - written Sanskrit. R1a and R1b-m269- Newar

R1a1-M198 from the north-western region of India before the arrival of the Indo-
European speakers.23Although a much earlier entry ofR1a1-M198.........wiki "Newars are a multi-ethnic, linguistic and cultural community of Indo-Aryan and Tibeto-Burman ethnicities."
http://www.cell.com/AJHG/retrieve/pii/S0002929707609446

I have a feeling also Lashiti plateau in Crete has some old variants of R1b and R1a that might be connected with the above mentioned regions.
http://www.nature.com/ncomms/journal/v4/n5/full/ncomms2871.html

Bulgaria R1a-458 & R1b L23 Razgrad
http://www.ncbi.nlm.nih.gov/pubmed/23483890
http://dienekes.blogspot.ca/2013/03/y-chromosomes-of-bulgarians-karachanak.html


wiki...Razgrad was built upon the ruins of the Ancient Roman town of Abritus on the banks of the Beli Lom river. Abritus was built on a Thracian settlement of the 4th-5th century BC with unknown name

newtoboard
11-14-2013, 06:56 PM
Would you connect the possible satemization of Thracian and Dacian to an influence of R1a-Z93+ or R1a-Z280+? I know some people still believe in the Thraco-Cimmerian theory.

Baltimore1937
11-15-2013, 05:57 PM
My L664 (R1a) may have spoken a language(s) unrelated to PIE. PIE evolved later?

newtoboard
11-15-2013, 06:11 PM
My L664 (R1a) may have spoken a language(s) unrelated to PIE. PIE evolved later?

L664+ certainly seems like a very early seperated branch but if it originated somewhere in NW Corded Ware it is ppossible it was also IE speaking. Modern IE languages linked to R1a are certainly linked only to Z645+ clades.

newtoboard
11-20-2013, 01:19 PM
The association of some early separated R1b species (like R1b-M73 or some unknown R1b(xM239) subclades) with the Tocharian branch of IE is not considered by me as well supported, but I would, nevertheless, include this presumption to my hypothesis.


What do you think R1b-M73's original linguistic identity was then?

Michał
11-25-2013, 11:04 PM
My L664 (R1a) may have spoken a language(s) unrelated to PIE. PIE evolved later?
I agree with newtoboard that they likely spoke an IE language, and I would also suspect that it was an extinct Satem dialect.

Michał
11-25-2013, 11:16 PM
What do you think R1b-M73's original linguistic identity was then?
I really don't know, but if I had to guess, I would suspect an Early PIE dialect, most likely ancestral to Tocharian. The idea that Tocharian was from the very beginning associated with R1a (and more specifically with R1a-Z93) is simply not consisted with the linguistic data, unless it was an early separated R1a(xZ645) lineage that was nearly completely displaced by some other R1a and R1b lineages in more recent times (but we simply don't know such Asian R1a(xZ645) lineage that would fit this scenario).

newtoboard
11-26-2013, 01:57 PM
I agree with newtoboard that they likely spoke an IE language, and I would also suspect that it was an extinct Satem dialect.

What about Z284+?

newtoboard
11-26-2013, 01:59 PM
I really don't know, but if I had to guess, I would suspect an Early PIE dialect, most likely ancestral to Tocharian. The idea that Tocharian was from the very beginning associated with R1a (and more specifically with R1a-Z93) is simply not consisted with the linguistic data, unless it was an early separated R1a(xZ645) lineage that was nearly completely displaced by some other R1a and R1b lineages in more recent times (but we simply don't know such Asian R1a(xZ645) lineage that would fit this scenario).

What about the possibility of R1b-M73 fitting with Botai? I don't think anything supports a M73 link to Tocharian giving that Tocharian likely descended from an Eastern Yamnaya variant, there is no R1b likely to be found in Afanasievo/the early Tarim and M73 is too old for it to have been speaking a PIE or IE language.

Michał
11-26-2013, 03:34 PM
What about Z284+?
Like in the case of L664, an early Satem dialect, but more closely related to a dialect ancestral to Balto-Slavic.

Michał
11-26-2013, 04:03 PM
What about the possibility of R1b-M73 fitting with Botai?
I find this idea very interesting and I have already mentioned this in one of my previous posts in another thread.

Is there anything that would speak strongly against Botai being associated with a very early dialect of IE (or the so-called Early PIE)?


I don't think anything supports a M73 link to Tocharian giving that Tocharian likely descended from an Eastern Yamnaya variant, there is no R1b likely to be found in Afanasievo/the early Tarim
I do believe that there was indeed some link between the early Eastern Yamnaya population and Afanasevo, but don't know any data that would strongly support the association of Afanasevo with the Tocharian-speaking people rather than with a population using some Eastern Satem dialect, probably related to Indo-Iranian. Most importantly, any strong association between the hypothetical Pre-Tocharian language and R1a-Z93 is simply implausible when taking into account all the linguistic data we have.


and M73 is too old for it to have been speaking a PIE or IE language.
I would rather say that the age of M73 fits perfectly the scenario in which the two major dialects of Early PIE have been separated, with one of them giving rise to Pre-Tocharian (M73) while the other being ancestral to the so-called Late PIE dialect (M269). Can you see any similar split in the R1a (or R1a-Z93) tree that would so nicely correspond to this commonly acknowledged linguistic division between the two Early PIE dialects?

newtoboard
11-26-2013, 04:30 PM
The argument against Botai being associated with IE languages is that it shows its closest link to Keltiminar and not steppe cultures.

There was no R1b in Xiahoe either which looks like a population that moved from South Siberia to Central Asia (the opposite of Andronovo). Maybe the Tocharian subclade of R1a died out.

What can't R1a-Z93+ correspond to Tocharian? We could have a case where R1a-Z93+ is just a Eastern lineage. Satemization is just a process and likely occurred later on. I haven't seen anything suggesting Tocharian is particularly close to other Centum languages.

Can you post more information on there being two so called PIE dialects? It doesn't make sense if we have R1b in Sredny Stog seperated by R1a in Yamnaya and M73 further out east speaking another PIE dialect.

Michał
11-27-2013, 01:51 AM
The argument against Botai being associated with IE languages is that it shows its closest link to Keltiminar and not steppe cultures.
What kind of link do you have in mind? Are you suggesting that Botai was derived from Kelteminar and that Kelteminar was definitely associated iwth a non-IE language? If you know any data that strongly support both these assumptions, please share it with me.


There was no R1b in Xiahoe either which looks like a population that moved from South Siberia to Central Asia (the opposite of Andronovo). Maybe the Tocharian subclade of R1a died out.
We really don't know when the Tocharian-speaking people arrived to the Tarim basin and where they came from, but what we know is that when Tocharian is first attested there, it is just one of several language families used in this very region (and this includes some non-Tocharian IE language families, like Iranian and Indo-Aryan), while the aDNA samples from Xiaohe predate the first attestation of Tocharian by more than 2500 years. This does not allow us to securely assume that those R1a people from Xiaohe were indeed Tocharian-speaking.


Maybe the Tocharian subclade of R1a died out.
This is of course possible. However, assuming an existence of an extinct R1a lineage associated with the Tocharian language cannot be considered the most likely option when we have an early separated Asian R1b lineage that fits the ancient Tocharian branch relatively well (at least much better than any known subclade of R1a).



What can't R1a-Z93+ correspond to Tocharian? We could have a case where R1a-Z93+ is just a Eastern lineage. Satemization is just a process and likely occurred later on. I haven't seen anything suggesting Tocharian is particularly close to other Centum languages.
This is exactly what makes its early association with R1a-Z93 very unlikely. Tocharian is not closely related to any of the remaining Centum languages because this particular branch has been relatively early separated from the branch ancestral to the so-called Late PIE dialect (ancestral to most known IE language families, including Greek, Italic, Celtic, Germanic, Baltic, Slavic, Iranian and Indo-Aryan). Thus if Indo-Iranian is commonly considered to be more closely related to Slavic, Germanic, Celtic and Greek than to Tocharian, it is simply hard to imagine how R1a-Z93 would be specifically associated with the development of both Indo-Iranian and Tocharian but not Celtic, Germanic or Baltic.


Can you post more information on there being two so called PIE dialects?
The vast majority of the relatively recent "professional" IE trees place Tocharian as a branch that was second (after Anatolian) to split from the main IE branch, so this is supposed to take place before all remaining major IE branches were separated from each other (see for example the linguistic tree shown in the first post in this thread). People use different terms to name all those subsequent stages of the initial IE development. For example, Anthony uses the term "Archaic PIE" for a dialect ancestral to all known IE languages (so it can be called a "true PIE", though "Indo-Hittite" is also frequently used). This initial split resulted in a separate development of Pre-Anatolian and a branch ancestral to the so-called "Early PIE", ancestral to Tocharian and "Late PIE". Finally, the Late PIE language is a dialect ancestral to all known IE families of languages except Anatolian and Tocharian.


It doesn't make sense if we have R1b in Sredny Stog seperated by R1a in Yamnaya and M73 further out east speaking another PIE dialect.

I don't think it would make more sense if only one of the two closely related R1b lineages spoke IE language, especially when we have no better candidate for the language associated with R1b-M73? What would be the most likely language that was initially spoken by the R1b-M73 people in your opinion?

Generalissimo
11-27-2013, 03:30 AM
Is there anything that would speak strongly against Botai being associated with a very early dialect of IE (or the so-called Early PIE)?

Apparently, David Anthony doesn't like the idea. It's in these videos, but I forget where exactly.

http://www.youtube.com/watch?v=d864bwyCAoA&list=PLAXoDomeFLX90fTHi0W8lYBtEoZHSBH2i&index=10

http://www.youtube.com/watch?v=HliaR2Ep24s&list=PLAXoDomeFLX90fTHi0W8lYBtEoZHSBH2i&index=6

His argument is that PIE has words for animal domestication, but there's no evidence of animal domestication east of the Urals early enough. There are just hunter-gatherers living there until the Indo-Europeans cross over from west of the Urals.


I would rather say that the age of M73 fits perfectly the scenario in which the two major dialects of Early PIE have been separated, with one of them giving rise to Pre-Tocharian (M73) while the other being ancestral to the so-called Late PIE dialect (M269). Can you see any similar split in the R1a (or R1a-Z93) tree that would so nicely correspond to this commonly acknowledged linguistic division between the two Early PIE dialects?


We really don't know when the Tocharian-speaking people arrived to the Tarim basin and where they came from, but what we know is that when Tocharian is first attested there, it is just one of several language families used in this very region (and this includes some non-Tocharian IE language families, like Iranian and Indo-Aryan), while the aDNA samples from Xiaohe predate the first attestation of Tocharian by more than 2500 years. This does not allow us to securely assume that those R1a people from Xiaohe were indeed Tocharian-speaking.

Even if we assume that the Tarim Basin mummies weren't Proto-Tocharians, they were obviously an early offshoot of a western population moving to the east. So if they don't need a perfect split in the R1a phylogeny to exist as an overwhelmingly R1a population that early and that far east, then the Tocharians don't need it either. Although there is some R1a-M17* in Tibet, so complete Y-sequences of the Tarim Basin R1a would be very useful. Maybe they belonged to a now extinct or extremely rare Tarim Basin/Tocharian subclade?

By the way, what direct evidence do you have that R1b was present in the Tarim Basin, or even Central Asia, early enough to be a Tocharian marker? There's no R1b in any of these ancient samples from China.

http://img826.imageshack.us/img826/1599/ancientchineseydna.png

http://cdmd.cnki.com.cn/Article/CDMD-10183-1012365432.htm

But we know that during the Iron Age there were European-like populations as far east as the Ordos culture.

http://en.wikipedia.org/wiki/File:OrdosPeopleMap.jpg

http://onlinelibrary.wiley.com/doi/10.1002/ajpa.21585/abstract

So the question is, as usual, where's the R1b hiding? And why is R1a already out and about in all the right places?

Michał
11-27-2013, 12:04 PM
Apparently, David Anthony doesn't like the idea. It's in these videos, but I forget where exactly.
...
His argument is that PIE has words for animal domestication, but there's no evidence of animal domestication east of the Urals early enough. There are just hunter-gatherers living there until the Indo-Europeans cross over from west of the Urals.
You are right that Anthony seems to play down a potential role of those early cultures east of Ural in the process of animal domestication. Given a very significant role he definitely applies to domesticated horses, he must have had some serious reasons to ignore both Tersek-Botai and their potential predecessors (Agidel-Surtanda) as some potential centers of early horse domestication, though he does not provide enough details to make me convinced that he is indeed right in this respect.

By contrast, here is what Mallory writes in his book (In Search of the Indo-Europeans) on this very subject :

“Stretching from the southern Urals across the Volga and on westwards to the Manych depression is the Seroglazovo culture of the Pre-Caspian region. It is attested from nearly 1oo sites, mainly surface finds, which include egg-shaped pots and lithic industries still firmly derived from the Mesolithic. The sites are almost invariably situated along river or lakeside shores which would have been far more abundant then, when the Caspian Sea was much larger than at present. This provided a regimen of deltas and marshes along the southern fringe of the Seroglazovo culture. Radiocarbon dates for the maximum transgression of the Caspian suggest that the Seraglazovo culture dated to the seventh or sixth millennium BC.

To the north, principally in the forest-steppe of the middle Volga and extending eastwards to the southern Urals (the Agidel culture), there also lie a substantial number of recently excavated Neolithic sites. Here we speak not only of the recurrent remains of egg-shaped vessels or lithic traditions reminiscent of the Mesolithic but also of evidence for domestic livestock.

The economy of these eastern sites was based on both hunting-fishing and stockbreeding. The hunted remains included the usual prey encountered across most of Europe and in the Proto-Indo-European vocabulary. According to Gerard Matyushin and Aida Petrenko, the domestic animals included the horse, frequently the predominant species, cattle and ovicaprids. Domestic pig is conspicuously absent. Radiocarbon dates are not numerous and are quite controversial in that they suggest that the Neolithic economy had emerged here by at least the sixth millennium BC, if not earlier.”

As for the potential relationship between Agidel-Surtanda and Tersek-Botai, some information is provided in my earlier post:
http://www.anthrogenica.com/showthread.php?1016-Thoughts-on-what-autosomal-component-was-originally-linked-with-R1b-%28M269-and-down%29&p=11291&viewfull=1#post11291



Even if we assume that the Tarim Basin mummies weren't Proto-Tocharians, they were obviously an early offshoot of a western population moving to the east.
If by “western population” you mean “Western Eurasian” (a term that includes Eastern Europe), I would certainly agree with that.



So if they don't need a perfect split in the R1a phylogeny to exist as an overwhelmingly R1a population that early and that far east, then the Tocharians don't need it either.
Let me rectify it a bit. The presence of R1a (and more specifically R1a-Z93, if rightly assumed) in Xiaohe at about 2000 BC is perfectly consistent with both the phylogeny of R1a and a potential association with some Early Satem dialect. What is much less likely (though not absolutely impossible) is that this potential ancient R1a-Z93 was initially associated with a very early separated branch of IE like Tocharian.



Although there is some R1a-M17* in Tibet, so complete Y-sequences of the Tarim Basin R1a would be very useful. Maybe they belonged to a now extinct or extremely rare Tarim Basin/Tocharian subclade?
This would make sense indeed. However, this would also mean that this potential early “Tocharian” subclade of R1a is much less frequent in the Tarim basin (and in Asia generally) than different species of early separated R1b, which still leaves the involvement of R1b in spreading Tocharian a slightly more likely option (at least IMO).



By the way, what direct evidence do you have that R1b was present in the Tarim Basin, or even Central Asia, early enough to be a Tocharian marker? There's no R1b in any of these ancient samples from China.
We don’t have any aDNA from the Tarim basin for the relevant time period (let’s say 500-1000 AD). Therefore, all we have right now is a substantial presence of different rare R1b species (including R1b-M73 and R1b-M269*) in the modern population of this region. This is of course not enough to make definite predictions about the past, but it is also hard to ignore this data, especially when they make some sense from the phylogenetic and linguistic point of view.



So the question is, as usual, where's the R1b hiding? And why is R1a already out and about in all the right places?
Please notice that a huge number of different Indo-Iranian languages have been attested in different parts of Eurasia during the last 3500 years, while we were extremely lucky to get any evidence for the presence of a single extinct family of IE languages like Tocharian (attested for a period of 1500-1000 years ago in a relatively small region only), so both the modern and Bronze/Iron Age proportions between R1a-Z93 and some early species of R1b in Asia (including the Tarim Basin) are perfectly consistent with this. Also, without any aDNA data from the actual Tocharian-speaking population in the Tarim basin, we will probably be unable to make any significant progress in this respect.

newtoboard
11-29-2013, 04:54 PM
I really doubt Xiahoe is associated with any sort of Indo-Iranian languages. mtDNA does not seem similar to Andronovo at all. It looks like a population that moved there from South Siberia. Indo-Iranian expansion was the other way around (Central Asia to Siberia). Not to mention it is in a location where no sort of Indo-Iranian language is have known to existed.

newtoboard
11-29-2013, 04:57 PM
Do we know what the STR's of the Xiahoe R1a looks like?

Also Keltiminar is associated to be Uralic. Animal domestication east of the Urals where Botai and related cultures seems to be more similar to the farming package of Asia. A lack of pigs is one such example as is the increased importance of sheep and goats.

alan
11-29-2013, 05:27 PM
I dont think we can be sure about Afanasievo and Tocharian or who was involved. However, I would observe again that Afanasievo seems to have taken the classic historical migratory route through the Kazak steppe and essentially stopped around Altai in a position that looks designed to control the Dzangurian gate.

The Tarim route is the historical trade route, a more difficult route and apparently taken to avoid open steppe conditions and of course steppe raiders.

We dont know what the Afansievo culture were like in terms of yDNA. Today there location is R1a dominated in terms of R but the steppe is very hard to read the past from modern populations. We have the Tarim mummies and so far they are R1a. However, they are way too late to link with Afansievo and basically in the wrong place. So, the connection of the R1a tarim mummies to either Afanasievo or Tocharian is speculative. I would go further and say that the link with Afanasievo is tenuous.

Between the period of Afanasievo, the Tarim mummies and the attestation of Tocharian there are huge gaps in time and also, in the case of Afanasievo and the mummies, a geographical/functional gap.

All we can say with confidence is that R1a Europeans were using the Tarim trade route by 1800??BC. How that connects to Afansievo around 3500BC in a different route to the north or Tocharian attested far later may never be clear.

I actually think people need to realise that while we may be able to work out origins, date and connections as ancient DNA builds up we can never be sure what anyone was speaking. That will always just be inference working back from the literate period of a given language and the genes of the people at that time but this is full of pitfalls when several leaps of faith have to be made to back-project that into the 4th millenium.

newtoboard
11-29-2013, 05:40 PM
I dont think we can be sure about Afanasievo and Tocharian or who was involved. However, I would observe again that Afanasievo seems to have taken the classic historical migratory route through the Kazak steppe and essentially stopped around Altai in a position that looks designed to control the Dzangurian gate.

The Tarim route is the historical trade route, a more difficult route and apparently taken to avoid open steppe conditions and of course steppe raiders.

We dont know what the Afansievo culture were like in terms of yDNA. Today there location is R1a dominated in terms of R but the steppe is very hard to read the past from modern populations. We have the Tarim mummies and so far they are R1a. However, they are way too late to link with Afansievo and basically in the wrong place and Indo-Iranian elements had arrived before the date of the earliest excavated mummies. So, the connection of the R1a tarim mummies to either Afanasievo or Tocharian is speculative. I would go further and say that the link with Afanasievo is tenuous.

Between the period of Afanasievo, the Tarim mummies and the attestation of Tocharian there are huge gaps in time and also, in the case of Afanasievo and the mummies, a geographical/functional gap.

All we can say with confidence is that R1a Europeans were using the Tarim trade route by 1800??BC. How that connects to Afansievo around 3500BC in a different route to the north or Tocharian attested far later may never be clear.

I actually think people need to realise that while we may be able to work out origins, date and connections as ancient DNA builds up we can never be sure what anyone was speaking. That will always just be inference working back from the literate period of a given language and the genes of the people at that time but this is full of pitfalls when several leaps of faith have to be made to back-project that into the 4th millenium.

That is speculation as well. Xiahoe is not in a location where any Indo-Iranian population is have known to exist.

newtoboard
11-29-2013, 05:55 PM
All of Andronovo's European, South Siberian and Central Asian extensions are well known so I really doubt anybody missed a spread into the Tarim Basin.

alan
11-29-2013, 07:26 PM
Although he has speculated that they might link with Afanasievo, Mallory mentioned Indo-Iranian elements arriving in the Tarim before the date of the earliest of the mummies. The problem with the mummies is that culturally they are later and have moved on from any classic cultural associations. He gives some possible echos of Afansievo and notes physical types etc but I dont think even he is totally confident on a straight forward link- he never says anything without qualification unless its pretty well certain. They could have by then been a mixed group. What seems clear is that they were traders along a very non-steppe type route, a major change from migrants sweeping across the steppe such as Afanasievo's distribution suggests.

Certainly the main problem is the tarim mummies are too late to tie to Afanasievo (and outwith the range of that culture) while Tocharian is too late attested to tie with either with confidence. They could be three separate things.

I think we may have three things going on here.

1. Afansievo sweeping across the Khazak steppe to the Dzungarian gates, probably interested in that important trade node and the metals in the Altai area. That appears to be a very early thrust west c. 3500BC when Yamnaya was only just close to coming into existence. It should in theory be an early branching off PIE but that depends on what date one places PIE at in the 4300-3500BC range. I tend to think full blown non-Anatolian PIE had basically developed by around 4000BC and so an offshoot east 500 years later may not correspond with the first branchings. Whoever they were in terms of language branch, there were Suvorovo type steppe settlers in the Balkans perhaps 700 years before Afanasievo appeared. Anthony links them with Anatolians and perhaps a tribe or two were but I suspect that most were early PIEs ancestral to many of the languages of the Balkans and western Europe. They remained not only adjacent to their original homes east of the Dniester but did return trips back that way in the centuries after, a scenario in which further aerial language development would have taken place. I dont believe for one minute that most of these early Suvorovo groups would not have picked up the word for the wheel.

Anyway, if those Suvorovo groups were full PIEs (not Anatolians) c. 4000BC then Afanasievo breaking off from the core C. 3400BC is not especially early (although they probably set off them the eastern extremity of the PIE zone) and that to me throws some doubt on the Afanasievo-Tocharian link. It is possible that Afanasievo was a more developed form close to the dialects during the Yamnaya expansions west.

A lot depends on how one interprets the centum-satem thing. Some of it could be original and some could be aerial - its probably complex. The roots of Balto-Slavic are traced through a chain of cultures that seems to lead back to the Middle Dnieper culture c. 3200BC and at least some (I am not saying they are right) do see the satem split as being carried by that cultural chain. If that were true then satemisation could have been underway by or before 3200BC. So, although its usually linked to the earliest centum branch, it doesnt seem impossible to me that Afanasievo was actually a satem speaking tribe using the steppe route rather than the Tarim trade route. If that was the case then the Tocharians might have been something else altogether than used the Tarim trade route.

This map of the later trade routes is thought provoking

http://www2.kenyon.edu/Depts/Religion/Fac/Adler/Asia201/Maps/SilkRoad2.gif

I have little doubt that R1a dominated the very northern steppe route that landed about Dzungaria at the gates to Mongolia and I would imagine R1a's very decent showing among the Mongol-Turkic groups in that area owes much to this. This route could have been used for trade but in general it was a route where nomadic steppe peoples could both travel and trade with weight of numbers for safety. It was avoided by specialist traders from what I have read. This was, broadly speaking the Afansievo type route and the end of the route is where Afansievo is found.

The route that M73 could have been linked to is the stretch that led from the north-east Caspian-Aral Sea area south-east bypassing the desert around the Aral Sea and down pas Toshkent into a position close to both Tarim and western trading areas that accessed the south. Or the other slighly different route from the east Caspian portage point then east. I have commented recently that M73 probably arose within 500miles of the Aral Sea area, although not in the deserts themselves. I do not believe that the southermost route east through Iran and Bactria is viable simply because M73 seems to die off at the mountain fringe of central Asia and is practically absent in Iran according to academic papers. It is also practically absent in the Hindu Kush study area around the mountain fringe of central Asia. It also

I am not trying to guess periods here. These are natural route so could have been used many times. However, I would say on Balance that the northermost route may have been taken by r1a and the southern route may have also involved r1b

That is speculation as well. Xiahoe is not in a location where any Indo-Iranian population is have known to exist.

newtoboard
11-29-2013, 07:40 PM
There is nothing suggesting R1b took any sort of root in Central Asia (at least not west from the steppe). That is just speculation. If R1b passed through Tashkent which is in the Ferghana valley we would see plenty of R1b-M73 in Tajiks. We don't. It is indeed likely that Tocharian passed through the deserts though since they somehow ended up engaing in oasis farming , something which only really has its parallels in Central Asia.

If satemization started that early we would see Balto-Slavic dialects reaching full satemization. That is not the case.

The Indo-Iranian elements were always along the edge of the Tarim and the Southern Ridge. That likely was true during silk road times as well.

Jean M
11-29-2013, 07:42 PM
Although he has speculated that they might link with Afanasievo, Mallory mentioned Indo-Iranian elements arriving in the Tarim before the date of the earliest of the mummies.

I'm not sure which work you are taking that from, but it is incorrect. The mummies in the Tarim Basin from whom we have DNA are from the earliest found there, in the Xiaohe (Small River) Cemetery, dated c. 2000 BC. You can see the exact dates in my list http://www.ancestraljourneys.org/ancientdna.shtml . That is before the arrival of groups believed to be Iranian speakers. See Romgard 2008 for the dating of the various sites. The first Andronovo type material is dated 1700–1400 BC.

newtoboard
11-29-2013, 07:47 PM
I'm not sure which work you are taking that from, but it is incorrect. The mummies in the Tarim Basin from whom we have DNA are from the earliest found there, in the Xiaohe (Small River) Cemetery, dated c. 2000 BC. You can see the exact dates in my list http://www.ancestraljourneys.org/ancientdna.shtml . That is before the arrival of groups believed to be Iranian speakers. See Romgard 2008 for the dating of the various sites.


So older than Andronovo by at least 200 years? Hopefully this Tarim mummies=Andronovo myth can go away now. Not that mtDNA and geography didn't make that argument look wrong anyways.

Jean M
11-29-2013, 07:53 PM
@ Newtoboard

This argument seems to crop up periodically. I think we can expect it all over again next year! People forget. I dare say I would myself if I hadn't written it all down. ;)

alan
11-29-2013, 08:42 PM
I sit corrected. My memory reversed the dates.


I'm not sure which work you are taking that from, but it is incorrect. The mummies in the Tarim Basin from whom we have DNA are from the earliest found there, in the Xiaohe (Small River) Cemetery, dated c. 2000 BC. You can see the exact dates in my list http://www.ancestraljourneys.org/ancientdna.shtml . That is before the arrival of groups believed to be Iranian speakers. See Romgard 2008 for the dating of the various sites. The first Andronovo type material is dated 1700–1400 BC.

alan
11-29-2013, 09:11 PM
The middle route that passes south-east past the Aral Sea is largely through Kazakhstan before entering the Tarim area. It only passes through a tiny bit of the interface between Tajikistan, Kyrgyzstan and Uzbekistan in between. The vast majority of the route is in Kazakhstan and China.

http://www2.kenyon.edu/Depts/Religion/Fac/Adler/Asia201/Maps/SilkRoad2.gif
http://www.allrussias.com/images/cis6_big.jpg

I cannot accept without further research that they are just Soviet deportees from the west.


There is nothing suggesting R1b took any sort of root in Central Asia (at least not west from the steppe). That is just speculation. If R1b passed through Tashkent which is in the Ferghana valley we would see plenty of R1b-M73 in Tajiks. We don't. It is indeed likely that Tocharian passed through the deserts though since they somehow ended up engaing in oasis farming , something which only really has its parallels in Central Asia.

If satemization started that early we would see Balto-Slavic dialects reaching full satemization. That is not the case.

The Indo-Iranian elements were always along the edge of the Tarim and the Southern Ridge. That likely was true during silk road times as well.

alan
11-29-2013, 09:23 PM
Also the map shows another route that headed east from a little south of the Aral Sea after a Caspian portage. The majority of this route was in Turmenistan where a lot M73 has been found. Recall too that these routes probably involved relay type systems rather than a complete journey from one end to the other. So a clade doesnt have to follow a route from end to end.

Michał
11-30-2013, 02:13 AM
I really doubt Xiahoe is associated with any sort of Indo-Iranian languages.
Agreed. This is exactly why I suggested that they could have spoken some early Satem dialect related to Indo-Iranian (but probably not Proto-Indo-Iranian or any specific Indo-Iranian dialect).


mtDNA does not seem similar to Andronovo at all.
Firstly, I have never suggested that those people from Xiaohe were derived from Andronovo.
Secondly, does the mtDNA package of the Indo-Iranian speakers in India, Pakistan, Afghanistan, Tadjikistan, Iran, Kurdistan or Osetia reflect the mtDNA composition found in Andronovo?



It looks like a population that moved there from South Siberia. Indo-Iranian expansion was the other way around (Central Asia to Siberia).
This is possible, and I also consider the commonly suggested link with Afanasevo quite likely, so let's wait for the aDNA results from Afanasevo.


Not to mention it is in a location where no sort of Indo-Iranian language is have known to existed.
We don't know it for sure, but it is sometimes suggested that the Yuezhi people who originated from that very region but expanded first east and then south were speakers of some Iranian dialect. At least they spoke an Iranian language when settled in Bactria (before they founded the Kushan Empire). Also, we know that there is a lot of R1a in China and most of it seems to be R1a-Z94 (including Z2122, as recently reported). If this R1a-Z94 is not of some Early Iranian origin, then how would you explain its significant presence in Western China. If most of those presumably Indo-Iranian-associated subclades of Z93 derive from some early Tocharian speakers (and not from some speakers of Indo-Iranian or closely related languages) this wouldn't make much sense from the linguistic point of view, as I have already pointed out on multiple occasions.

Michał
11-30-2013, 02:19 AM
Do we know what the STR's of the Xiahoe R1a looks like?
Nope, so we have really no clue what kind of R1a it was.


Also Keltiminar is associated to be Uralic.
The problem is that I've already seen nearly every Mesolithic, Early Neolithic or Bronze Age culture in North-Eastern Europe, Ural region, Central Asia and Siberia being suggested as associated with the Proto-Uralic or Early Uralic speakers. The truth is nobody really knows where exactly Uralic speakers came from and when exactly it happened. Personally, I really doubt that Kelteminar culture was associated with any group of Early Uralic speakers, as based on the Yukhagir evidence (suggesting early presence of some Pre-Uralic speakers in Siberia) and on some recent linguistic analyzes that date the Proto-Uralic language to just 2000 BC, I would rather believe that the pre-Uralic population came from the Altai region and did not expand into the North-Eastern Europe before 2000 BC.


Animal domestication east of the Urals where Botai and related cultures seems to be more similar to the farming package of Asia. A lack of pigs is one such example as is the increased importance of sheep and goats.
According to my scenario, they did originate in Asia (like all Indo-Europeans), so this would be actually pretty consistent with my hypothesis. :)

Generalissimo
11-30-2013, 04:22 AM
According to my scenario, they did originate in Asia (like all Indo-Europeans), so this would be actually pretty consistent with my hypothesis.

Can't see much evidence of this to date, since even the mtDNA C in Ukrainian Kurgans is European-specific and found in earlier Ukrainian Neolithic remains.

So this would mean that the Indo-European expansion was solely driven by males up to Eastern Europe, and then suddenly it became also a female mediated event, as Corded Ware and Unetice mtDNA data from Central Europe suggest.

So to sum up: this has no support in any data, and instead relies on an unusual turn of events.

Michał
11-30-2013, 11:07 AM
Can't see much evidence of this to date, since even the mtDNA C in Ukrainian Kurgans is European-specific and found in earlier Ukrainian Neolithic remains.

You are talking about the IE-speaking populations that derive from the so-called Late PIE group which represents a relatively late stage of IE development and seems to have been indeed strongly associated with Europe (or specifically with Eastern Europe). However, when I mentioned the hypothetical Asian origin of IE, this was supposed to be related to some much earlier stages, including Archaic PIE (a dialect ancestral to both Anatolian and Early PIE), and Tocharian, one of the two hypothetical major sub-branches of Early PIE (the other being the Late PIE sub-branch in Eastern Europe).

Generalissimo
11-30-2013, 11:34 AM
You are talking about the IE-speaking populations that derive from the so-called Late PIE group which represents a relatively late stage of IE development and seems to have been indeed strongly associated with Europe (or specifically with Eastern Europe). However, when I mentioned the hypothetical Asian origin of IE, this was supposed to be related to some much earlier stages, including Archaic PIE (a dialect ancestral to both Anatolian and Early PIE), and Tocharian, one of the two hypothetical major sub-branches of Early PIE (the other being the Late PIE sub-branch in Eastern Europe).

What evidence is there of any movements from Asia to Europe from the late Neolithic to Bronze Age?

There seem to have been a lot of expansions and migrations within Europe at this time, but these actually resulted in genetic shifts from Near Eastern-like to European-like, especially in Central Europe.

There's no sign of any introgression from Central Asia at this time anywhere in Europe. That sort of thing starts during the Iron Age, and can be clearly seen via the high incidence of East Asian-specific mtDNA in Sarmatian and Scythian remains from Hungary and Southern Russia.

Jean M
11-30-2013, 11:59 AM
Agreed. This is exactly why I suggested that they could have spoken some early Satem dialect related to Indo-Iranian (but probably not Proto-Indo-Iranian or any specific Indo-Iranian dialect).


I can understand how appealing the idea was that R1a = satem and R1b = centum. The idea swept the genetic genealogy forums years ago, but just like the early idea among linguists that satem = eastern IE and centum = western IE, it falls apart on Tocharian. People thought that Tocharian simply must be R1b. I did not. The centum-satem isogloss is chronological, not geographical, while the reverse is true of the R1a/R1b division.


The centum-satem isogloss is a late development in PIE. The departures east and west c. 3500-3000 BC did not carry it, as far as we can deduce from the end results: Tocharian, Celtic, Italic, Greek, etc. The first movement east which could carry it is therefore the Sintashta culture, starting 2100 BC. There is nothing else moving east between the trek to Afanasevo c. 3300 BC and Sintashta.
We have a convincing cultural sequence from Sintashta to Andronovo, and from the latter in the Krasnoyarsk region of Central Asia right through to cultures we can identify as Scythian and therefore Iranian-speaking. R1a1a has been found right through the sequence in the Krasnoyarsk region.
The earliest mummies in the Tarim Basin are of a culture called Gumugou by the Chinese. This culture is not a direct arrival there from the European steppe. Far from it. Genetically there is evidence of mixture with Central Asians. Culturally there is a sequence from the Afanasevo Culture to the Ke’ermuqi Culture (north of the Tianshan Mountains in the Junggar Basin) to Xiaohe. See Jan Romgard, Questions of Early Human Settlements in Xinjiang and the Early Silk Road Trade, with an Overview of the Silk Road Research Institutions and Scholars in Beijing, Gansu and Xinjiang, Sino-Platonic Papers, ed. V. Mair, vol. 185 (November 2008)

Jean M
11-30-2013, 12:08 PM
I should add that deduced Tocharian R1a + Iranian R1a does not rule out R1b of some type moving eastwards at some time. The possibility remains of


R1b in trace amounts among Iranian speakers
R1b-M73 having its own trajectory unconnected to IE, e.g. moving east along the Silk Road from some site in the Caucasus

Michał
11-30-2013, 04:58 PM
I can understand how appealing the idea was that R1a = satem and R1b = centum. The idea swept the genetic genealogy forums years ago, but just like the early idea among linguists that satem = eastern IE and centum = western IE, it falls apart on Tocharian.
I think you still miss the point I am trying to make. What I wrote on this subject is not only related to the Satem-Centum division (which indeed can be questioned) but to the commonly accepted division between the early separated Tocharian branch and the so-called Late PIE dialect that is supposed to be ancestral to all remaining IE languages (except Anatolian of course). Just take a look at the IE tree proposed by David Anthony in his book. How would you explain that those putative “Tocharians” who were separated from the remaining group of the Early PIE speakers long before the subsequent IE branches have been formed (including not only Indo-Iranian and Balto-Slavic but also Greek, Celtic, Italic and Germanic) were represented by the same relatively young subclades of Z93 that are common today among the Indo-Iranian speakers in different regions of Asia. How come such early division between the Tocharian and Indo-Iranian languages (predating the division between Indo-Iranian and let’s say Celtic or Germanic) is not reflected in the structure of R1a-Z93 at all?

Also, I think it is unfair to suggest that we already know which Y-DNA haplogroup was common among the Tocharian-speaking population in China, as this is definitely not the case.



People thought that Tocharian simply must be R1b. I did not.

We probably all agree that it did not need to be R1b (it could have been an early separated branch of R1a instead), but it just doesn’t make sense to associate Early Tocharian speakers with R1a-Z93 (or even with R1-Z94 and some of its subclades, as suggested by the R1a species found in China), as all these are relatively young clades that show quite evident association with the much more recent spread of the Indo-Iranian speakers.

Michał
11-30-2013, 05:00 PM
I should add that deduced Tocharian R1a + Iranian R1a does not rule out R1b of some type moving eastwards at some time. The possibility remains of


R1b in trace amounts among Iranian speakers
R1b-M73 having its own trajectory unconnected to IE, e.g. moving east along the Silk Road from some site in the Caucasus

What is then the main reason for excluding the possibility that those Asian R1b species were associated with some Early IE branches, especially when those R1b clades were found in places previously known to be associated with the Tocharian-speaking people?

Michał
11-30-2013, 05:17 PM
What evidence is there of any movements from Asia to Europe from the late Neolithic to Bronze Age?

Have I ever mentioned that it took place that late? I was rather suggesting that it was in the Early Neolithic, and such early migration from Asia to Eastern Europe (and more specifically from the East Caspian region to the North Caspian region) is supported by some archaeological findings, as discussed elsewhere:
http://www.anthrogenica.com/showthread.php?1016-Thoughts-on-what-autosomal-component-was-originally-linked-with-R1b-%28M269-and-down%29/page9&p=10960#post10960
See also:
http://www.anthrogenica.com/showthread.php?1016-Thoughts-on-what-autosomal-component-was-originally-linked-with-R1b-%28M269-and-down%29&p=11291&viewfull=1#post11291
http://www.anthrogenica.com/showthread.php?1016-Thoughts-on-what-autosomal-component-was-originally-linked-with-R1b-%28M269-and-down%29/page12&p=11232#post11232
http://www.anthrogenica.com/showthread.php?1016-Thoughts-on-what-autosomal-component-was-originally-linked-with-R1b-%28M269-and-down%29/page12&p=11152#post11152

Jean M
11-30-2013, 05:43 PM
How would you explain that those putative “Tocharians” who were separated from the remaining group of the Early PIE speakers long before the subsequent IE branches have been formed (including not only Indo-Iranian and Balto-Slavic but also Greek, Celtic, Italic and Germanic) were represented by the same relatively young subclades of Z93

David Anthony does have Tocharian branching before Italo-Celtic, as do Warnow et al - see the first post in this thread. However the degree of separation was rather stretched by Anthony, as (at the time of writing the book) he had to deal with radiocarbon dates for Afanasievo that seemed "too early" to linguists - c. 3700 BC. Now the outlier dates which were too early have been dismissed, we end up with a better fit. (See his recent paper). Afanasievo comes in at c. 3,300 BC - much more as expected linguistically. Tocharian is centum, as the Anatolian branch is not (it is pre-centum), and it has all the vehicle vocabulary which we expect after c. 3500 BC.

Various dates have been calculated for Z93. It seems to be roughly 3700 BC on the tree by your colleague at http://www.familytreedna.com/public/R1a/default.aspx?section=results . You have been making your own calculations on this thread http://www.anthrogenica.com/showthread.php?1507-Some-provisional-calculations-for-haplogroup-R1a-based-on-the-first-FGC-result . I'm not sure if I am following you correctly there. Are you saying that you are certain that Z93 is younger than 3300 BC?



How come such early division between the Tocharian and Indo-Iranian languages .. is not reflected in the structure of R1a-Z93 at all?

Do we know that it isn't? Looks like there are some deep divisions. But how are we to find out if any such division actually reflects descendants of Tocharian speakers? The culture is long gone. What happened to their descendants I don't know, or even if there are any.

It's not the best question to use modern DNA to answer, I'd say.

parasar
11-30-2013, 08:31 PM
That is speculation as well. Xiahoe is not in a location where any Indo-Iranian population is have known to exist.

"Lop Nur is the site of the bronze-age Xiaohe Tomb complex."
"The Tarim River ends in Lop Nur."

http://upload.wikimedia.org/wikipedia/commons/thumb/c/ca/Map_of_the_Lop_Nor_region_by_Folke_Bergman_1935.jp g/800px-Map_of_the_Lop_Nor_region_by_Folke_Bergman_1935.jp g

Now let us see the names in ~650AD
Sheeta (sanskrit cold) was name for Tarim
http://books.google.com/books?id=883OZBe2sMYC&pg=PA283

The name Shule still survives there.
“The pilgrim Xuanzang tells us that its name in Sanskrit was Śrīkrīrāti which means something like ‘Fortunate Hospitality’ ; the local name was transcribed in Chinese as Shule..., which provides fairly dramatic evidence for what happens when a Chinese tongue tries to articulate Indo-European clusters of sound."
http://depts.washington.edu/silkroad/texts/hhshu/notes21.html

It was the ancient Shambhala, the headquarters of the Kalachakra school.
https://collab.itc.virginia.edu/access/content/group/9f340e95-f808-4bc0-80bc-b23bcadd072e/Copyrighted%20PDFs%20of%20Texts/NewmanBriefHistoryKalachakra.pdf

Michał
12-01-2013, 01:17 AM
Various dates have been calculated for Z93. It seems to be roughly 3700 BC on the tree by your colleague at http://www.familytreedna.com/public/R1a/default.aspx?section=results . You have been making your own calculations on this thread http://www.anthrogenica.com/showthread.php?1507-Some-provisional-calculations-for-haplogroup-R1a-based-on-the-first-FGC-result . I'm not sure if I am following you correctly there. Are you saying that you are certain that Z93 is younger than 3300 BC?
No, I don't think Z93 is younger than 5300 years.



Do we know that it isn't? Looks like there are some deep divisions.
There are indeed some deep divisions in the Z93 branch, but it is hard to find anything that would indicate some potential separation between the hypothetical Tocharian and Indo-Iranian subclades of this branch. The Z93* paragroup seems to be present mostly in Europe and Near East, whereas I don't know any confirmed case of Z93* from India or China. The major subclade of Z93 is Z94, but since it seems to constitute the vast majority of all Z93 members, including nearly all R1a in India, it cannot be specific for Tocharian. Currently, we know three major subclades of Z94 that seem to encompass nearly all Z94 members. One of these subclades (Y40) is relatively small and reported for both Asia (India) and Europe (Italy), which makes its specific association with Tocharian unlikely. Another subclade known as L657 seems to be the major R1a subclade in India, which, again, makes it unlikely to be specifically associated with the Tocharians. Finally, the Z2124 subclade of Z94 is divided into two major sublineages, Z2122 and Z2123, both significantly present in Europe where their distribution seems to indicate some deep Iranian (Scythian or Sarmatian) ancestry (though some more recent Turkic contribution is also likely). Both these subclades are also found in the Near East, while only Z2123 seems to show very significant presence in India. The Z2124* paragroup is relatively large and includes separate subclusters associated with different ancestry, for example Hungarian Szekely, Kirgiz or Pashtun.

The recent study by Shi Yan et al., 2013, reports two R1a cases from China, and both turned out to be Z94+, with one of them representing the F1345 subclade of Z2122 (known to encompass the Ashkenazi-Levite clade CTS6), and more specifically a lineage defined by F2935, previously found in Scotland, Poland/Ukraine and Kazakhstan.

Jean M
12-01-2013, 01:50 AM
Do we have any samples of Uygur R1a tested for the new SNPs?

parasar
12-01-2013, 02:32 AM
No, I don't think Z93 is younger than 5300 years.



There are indeed some deep divisions in the Z93 branch, but it is hard to find anything that would indicate some potential separation between the hypothetical Tocharian and Indo-Iranian subclades of this branch. The Z93* paragroup seems to be present mostly in Europe and Near East, whereas I don't know any confirmed case of Z93* from India or China. The major subclade of Z93 is Z94, but since it seems to constitute the vast majority of all Z93 members, including nearly all R1a in India, it cannot be specific for Tocharian. Currently, we know three major subclades of Z94 that seem to encompass nearly all Z94 members. One of these subclades (Y40) is relatively small and reported for both Asia (India) and Europe (Italy), which makes its specific association with Tocharian unlikely. Another subclade known as L657 seems to be the major R1a subclade in India, which, again, makes it unlikely to be specifically associated with the Tocharians. Finally, the Z2124 subclade of Z94 is divided into two major sublineages, Z2122 and Z2123, both significantly present in Europe where their distribution seems to indicate some deep Iranian (Scythian or Sarmatian) ancestry (though some more recent Turkic contribution is also likely). Both these subclades are also found in the Near East, while only Z2123 seems to show very significant presence in India. The Z2124* paragroup is relatively large and includes separate subclusters associated with different ancestry, for example Hungarian Szekely, Kirgiz or Pashtun.

The recent study by Shi Yan et al., 2013, reports two R1a cases from China, and both turned out to be Z94+, with one of them representing the F1345 subclade of Z2122 (known to encompass the Ashkenazi-Levite clade CTS6), and more specifically a lineage defined by F2935, previously found in Scotland, Poland/Ukraine and Kazakhstan.

HG03705
https://docs.google.com/file/d/0B-b3KeGG3Un1NDRnQ2FBblpnSWM/edit

Jean M
12-01-2013, 02:44 AM
HG03705

HG03705 is a Punjabi in Lahore, Pakistan, I see from http://eng.molgen.org/viewtopic.php?p=15683#p15683

parasar
12-01-2013, 02:59 AM
HG03705 is a Punjabi in Lahore, Pakistan, I see from http://eng.molgen.org/viewtopic.php?p=15683#p15683

Yes that is correct. I was not excluding Lahore from India (after all it was capital of undivided Punjab) and I didn't think Michał was either when mentioning India or China.

Jean M
12-01-2013, 12:36 PM
Yes that is correct. I was not excluding Lahore from India (after all it was capital of undivided Punjab) and I didn't think Michał was either when mentioning India or China.

I understood perfectly. I just specified the location because the Indian subcontinent is a big place! The pattern of migrations into it and within it seems very complex, and I can't claim to have a real grip on it. But if we are looking for remnants of the Kushans, the north-west is where we would look I think.

parasar
12-01-2013, 04:22 PM
I understood perfectly. I just specified the location because the Indian subcontinent is a big place! The pattern of migrations into it and within it seems very complex, and I can't claim to have a real grip on it. But if we are looking for remnants of the Kushans, the north-west is where we would look I think.

And out of it too, which is often neglected.
The regions towards the NW would indeed match the populations emanating from there. As far as Kushans go they are supposed to have originated from the Licchavi-Vajji/Yue-chi*

* http://books.google.com/books?id=CBfN3uhRUvIC&pg=PA254


1. In Dhammapada we are told that in the Buddha period the Vajji were clashing with Ajatasattu:
http://books.google.com/books?id=ppYuAAAAYAAJ&pg=PA166

2. Later the Vajjiputtakas departed after the Second Buddhist Council (of KalAsoka) and were responsible for establishing their schools within and outside the subcontinent.
"The Vajjiputtakas refused to accept the finding of Revata’s Council and formed a separate sect, the Mahāsanghikas, numbering ten thousand monks, who held a recital of their own."
http://www.palikanon.com/english/pali_names/va/vajjiputtakaa.htm

3. "In 349 BCE, several years after the Second Buddhist Council, the Mahasanghika tradition of Hinayana split off from the Theravada. Many Mahasanghikas moved to Gandhara. At Hadda, the main city on the Afghan side, near present-day Jalalabad, they eventually founded Nagara Vihara Monastery, bringing with them a skull relic of the Buddha. A Theravada elder, Sambhuta Sanavasi, soon followed and tried to establish his trad*ition in Kapisha. He was unsuc*cessful, and Maha*sanghika took root as the main Buddhist tradition of Afghanistan.
Eventually, the Mahasanghikas split into five sub-schools. The main one in Afganistan was Lokottaravada, which later established itself in the Bamiyan Valley in the Hindu Kush Mountains."
http://www.berzinarchives.com/web/en/archives/study/history_buddhism/buddhism_central_asia/history_afghanistan_buddhism.html

Michał
12-01-2013, 06:59 PM
HG03705
https://docs.google.com/file/d/0B-b3KeGG3Un1NDRnQ2FBblpnSWM/edit
Thank you, Parasar! I always keep forgetting about this one. :)


Yes that is correct. I was not excluding Lahore from India (after all it was capital of undivided Punjab) and I didn't think Michał was either when mentioning India or China.
Yes, I should indeed mention this sample when writing about India.

Michał
12-01-2013, 07:07 PM
Do we have any samples of Uygur R1a tested for the new SNPs?
Unfortunately not. However, it seems quite likely that those two Chinese samples from the Shi Yan study were of Uyghur origin, though we of course cannot be sure about it.

I would guess that most of the Uyghur R1a is Z2124+, and this probably includes both Z2123 and Z2122. Please note that the Y-DNA regions encompassing Z2124 and Z2122 were not sequenced in that Chinese study (while we don't know if this was the case for Z2123, as well).

Michał
12-02-2013, 10:36 AM
I am trying to understand the relation between languages and Y-DNA lineages.

Below please find a very schematic tree that is supposed to reflect my current views on the most likely links between different IE linguistic branches and some Y-DNA haplogroups. Of course, please keep in mind that I was unable to include all the details and show some peculiarities associated with the mixed character of some IE-speaking populations.
992

Jean M
12-02-2013, 01:07 PM
Golly Michał! That's is very brave of you. The early stages of your model are completely different from mine, as you know, but it is certainly interesting to see how the evidence can be interpreted in different ways.

newtoboard
12-02-2013, 01:14 PM
Firstly, I have never suggested that those people from Xiaohe were derived from Andronovo.
Secondly, does the mtDNA package of the Indo-Iranian speakers in India, Pakistan, Afghanistan, Tadjikistan, Iran, Kurdistan or Osetia reflect the mtDNA composition found in Andronovo?




No. But there are lineages likely from Andronovo in those places. Andronovo lineages are a addition to those gene pools. Whereas the difference between the Andronovo and Xiaohe is that Xiaohe doesn't really have any lineages typical of Andronovo.

newtoboard
12-02-2013, 01:15 PM
Also the map shows another route that headed east from a little south of the Aral Sea after a Caspian portage. The majority of this route was in Turmenistan where a lot M73 has been found. Recall too that these routes probably involved relay type systems rather than a complete journey from one end to the other. So a clade doesnt have to follow a route from end to end.

I hope you are not suggesting the R1b-M73 in Turkmenistan has some sort of ancient presence. If it did why don't we see much R1b-M73 in Iran?

newtoboard
12-02-2013, 01:18 PM
I can understand how appealing the idea was that R1a = satem and R1b = centum. The idea swept the genetic genealogy forums years ago, but just like the early idea among linguists that satem = eastern IE and centum = western IE, it falls apart on Tocharian. People thought that Tocharian simply must be R1b. I did not. The centum-satem isogloss is chronological, not geographical, while the reverse is true of the R1a/R1b division.


The centum-satem isogloss is a late development in PIE. The departures east and west c. 3500-3000 BC did not carry it, as far as we can deduce from the end results: Tocharian, Celtic, Italic, Greek, etc. The first movement east which could carry it is therefore the Sintashta culture, starting 2100 BC. There is nothing else moving east between the trek to Afanasevo c. 3300 BC and Sintashta.
We have a convincing cultural sequence from Sintashta to Andronovo, and from the latter in the Krasnoyarsk region of Central Asia right through to cultures we can identify as Scythian and therefore Iranian-speaking. R1a1a has been found right through the sequence in the Krasnoyarsk region.
The earliest mummies in the Tarim Basin are of a culture called Gumugou by the Chinese. This culture is not a direct arrival there from the European steppe. Far from it. Genetically there is evidence of mixture with Central Asians. Culturally there is a sequence from the Afanasevo Culture to the Ke’ermuqi Culture (north of the Tianshan Mountains in the Junggar Basin) to Xiaohe. See Jan Romgard, Questions of Early Human Settlements in Xinjiang and the Early Silk Road Trade, with an Overview of the Silk Road Research Institutions and Scholars in Beijing, Gansu and Xinjiang, Sino-Platonic Papers, ed. V. Mair, vol. 185 (November 2008)


Source? And what are Central Asians defined as during this period?

newtoboard
12-02-2013, 01:22 PM
"Lop Nur is the site of the bronze-age Xiaohe Tomb complex."
"The Tarim River ends in Lop Nur."

http://upload.wikimedia.org/wikipedia/commons/thumb/c/ca/Map_of_the_Lop_Nor_region_by_Folke_Bergman_1935.jp g/800px-Map_of_the_Lop_Nor_region_by_Folke_Bergman_1935.jp g

Now let us see the names in ~650AD
Sheeta (sanskrit cold) was name for Tarim
http://books.google.com/books?id=883OZBe2sMYC&pg=PA283

The name Shule still survives there.
“The pilgrim Xuanzang tells us that its name in Sanskrit was Śrīkrīrāti which means something like ‘Fortunate Hospitality’ ; the local name was transcribed in Chinese as Shule..., which provides fairly dramatic evidence for what happens when a Chinese tongue tries to articulate Indo-European clusters of sound."
http://depts.washington.edu/silkroad/texts/hhshu/notes21.html

It was the ancient Shambhala, the headquarters of the Kalachakra school.
https://collab.itc.virginia.edu/access/content/group/9f340e95-f808-4bc0-80bc-b23bcadd072e/Copyrighted%20PDFs%20of%20Texts/NewmanBriefHistoryKalachakra.pdf


That is all good information. But that is extremely late. Maybe there were Indo-Iranian speakers during Silk road times but not before that.

Michał
12-02-2013, 01:33 PM
No. But there are lineages likely from Andronovo in those places. Andronovo lineages are a addition to those gene pools. Whereas the difference between the Andronovo and Xiaohe is that Xiaohe doesn't really have any lineages typical of Andronovo.
Do you have some sort of summarized data that would clearly demonstrate such influx of Andronovo-specific mtDNA species into Kurdistan, Iran and Afghanistan? Actually, I am very close to revise my view on the Proto-Indo-Iranian question in relation to archaeology and genetics, but need to work on it a bit more.

BTW, Jean, I cannot access your excellent Ancient DNA page. Is this just a temporary problem?

newtoboard
12-02-2013, 01:44 PM
Do you have some sort of summarized data that would clearly demonstrate such influx of Andronovo-specific mtDNA species into Kurdistan, Iran and Afghanistan? Actually, I am very close to revise my view on the Proto-Indo-Iranian question in relation to archaeology and genetics, but need to work on it a bit more.

BTW, Jean, I cannot access your excellent Ancient DNA page. Is this just a temporary problem?

I said it is likely most of the mtdna U4, U5, and some H/T lineages in those areas trace to Proto Indo-Iranian speakers. But I guess specific subclades need to be looked at. But the Andronovo to Xiaohe question remains. All those areas you mentioned were populated and it is possible that movements south were mostly male mediated. But there is nothing that I can think of that accounts for the gene pool of Andronovo to change so radically in their extremely short journey from Central Asia into the Tarim (from carrying a mix of U, T, H, K and Z to nothing but H and C). They passed through nothing but open steppe really. And as we can see later Andronovo derived cultures did indeed carry at least U and T.

Jean M
12-02-2013, 02:15 PM
BTW, Jean, I cannot access your excellent Ancient DNA page. Is this just a temporary problem?

Some people seem to be having trouble getting it since it moved to a new domain and host. You could try going direct to http://www.ancestraljourneys.org/ancientdna.shtml
If that does not work, try clearing your browser cache or try another browser. If none of that works, you can still see it in Google's cache.

Jean M
12-02-2013, 02:17 PM
Source? And what are Central Asians defined as during this period?

I am referring to the mtDNA C4 in the mummies from Xiaohe. (There was also one sample of M*, which we cannot count as European.) Li 2010 deduced that mixing had taken place in the Altai, before the move to the Tarim Basin.

Generalissimo
12-02-2013, 02:32 PM
There are indeed some deep divisions in the Z93 branch, but it is hard to find anything that would indicate some potential separation between the hypothetical Tocharian and Indo-Iranian subclades of this branch.

So can you find anything on the R1a phylo tree that would indicate a clear separation between the strange and enigmatic Tarim Basin mummies and Indo-Iranian R1a? Or do you think they just belonged to Z93?

But if they belonged to Z93, then why couldn't have the Tocharians? On the other hand, if the mummies were M417*, then why couldn't have the Tocharians been that as well?


The Z93* paragroup seems to be present mostly in Europe and Near East, whereas I don't know any confirmed case of Z93* from India or China. The major subclade of Z93 is Z94, but since it seems to constitute the vast majority of all Z93 members, including nearly all R1a in India, it cannot be specific for Tocharian.

Languages aren't genetic mutations, so you don't need Y-chromosome mutations to represent each language branch. I think that's the fundamental problem with your model. You wouldn't get it past any linguists.

Maybe that's why you've got Baltic and its predecessors on the periphery of the Indo-European tree, and yet linguists think that Lithuanian is the closest thing to PIE today?

In any case, like I've already said, there's M17* in Tibet, near the Tarim Basin.


I am referring to the mtDNA C4 in the mummies from Xiaohe. (There was also one sample of M*, which we cannot count as European.) Li 2010 deduced that mixing had taken place in the Altai, before the move to the Tarim Basin.

But he didn't know that C4 was found in Neolithic and Bronze Age Ukrainian samples, and from Kurgans no less.

parasar
12-02-2013, 02:51 PM
That is all good information. But that is extremely late. Maybe there were Indo-Iranian speakers during Silk road times but not before that.

Do we have any earlier names of these hydronyms? If not, these are the earliest known.
After all the Tarim (Seeta) River had to be important enough to the early Hindus to make it one of the four main rivers (Vakshu, Sindhu, Sheeta, Alakananda) of their cosmology.
http://books.google.com/books?id=Gjp7OFbQGYYC&pg=PA67

parasar
12-02-2013, 03:16 PM
Below please find a very schematic tree that is supposed to reflect my current views on the most likely links between different IE linguistic branches and some Y-DNA haplogroups. Of course, please keep in mind that I was unable to include all the details and show some peculiarities associated with the mixed character of some IE-speaking populations.
992

Looks reasonable.
I would bring the early PIE (but not the archaic PIE) branch out of R1a.
Plus where would you place R2-M479. Its association/correlation with the earliest forms of IE in South Asia is very strong.

newtoboard
12-02-2013, 03:20 PM
So can you find anything on the R1a phylo tree that would indicate a clear separation between the strange and enigmatic Tarim Basin mummies and Indo-Iranian R1a? Or do you think they just belonged to Z93?

But if they belonged to Z93, then why couldn't have the Tocharians? On the other hand, if the mummies were M417*, then why couldn't have the Tocharians been that as well?



Languages aren't genetic mutations, so you don't need Y-chromosome mutations to represent each language branch. I think that's the fundamental problem with your model. You wouldn't get it past any linguists.

Maybe that's why you've got Baltic and its predecessors on the periphery of the Indo-European tree, and yet linguists think that Lithuanian is the closest thing to PIE today?

In any case, like I've already said, there's M17* in Tibet, near the Tarim Basin.



But he didn't know that C4 was found in Neolithic and Bronze Age Ukrainian samples, and from Kurgans no less.


Exactly. There seems to be way too much emphasis on accounting for the origin and spread of different haplogroups based on language branching and geography on this forum.

Also I agree the mixed Gene pool was likely from Ukraine. If it was in the Altai then we would have likely seen a variety of east Eurasian mtdna lines.

Silesian
12-02-2013, 03:35 PM
Languages aren't genetic mutations, so you don't need Y-chromosome mutations to represent each language branch. I think that's the fundamental problem with your model. You wouldn't get it past any linguists.
Is something stopping you, from posting your model?

alan
12-02-2013, 03:35 PM
Very interesting. Also I love the curves - much nicer than straight lines lol. Do you have a text for you model.


Below please find a very schematic tree that is supposed to reflect my current views on the most likely links between different IE linguistic branches and some Y-DNA haplogroups. Of course, please keep in mind that I was unable to include all the details and show some peculiarities associated with the mixed character of some IE-speaking populations.
992

alan
12-02-2013, 03:42 PM
You are puzzling me about M73. Where do you see it if its not in the Hindu Kush area (which the study seems to show), not in west central Asia, not in the steppe, not in Iran (which a study does indicate), not in the south Caucasus (which a study indicates)etc. It had to have got to its present location from somewhere but you seem to be ruling out all possibilities short of teleport. How do you think M73 got to where it is?


I hope you are not suggesting the R1b-M73 in Turkmenistan has some sort of ancient presence. If it did why don't we see much R1b-M73 in Iran?

newtoboard
12-02-2013, 03:43 PM
Is something stopping you, from posting your model?


Why should he have to post a model? Because he disagrees with someone else's? As far as I am aware nobody on this forum is a professional linguist so what is the value in these models. Plus there is no reason a linguistic model should have to account for haplogroups as well as areal features such as satemization. There is no evidence that satemization is not a areal feature or is older than Proto Indo-Iranian so it is unlikely that the Xiaohe mummies were speaking some sort of early separated Satem dialect.

newtoboard
12-02-2013, 03:46 PM
You are puzzling me about M73. Where do you see it if its not in the Hindu Kush area (which the study seems to show), not in west central Asia, not in the steppe, not in Iran (which a study does indicate), not in the south Caucasus (which a study indicates)etc. It had to have got to its present location from somewhere but you seem to be ruling out all possibilities short of teleport. How do you think M73 got to where it is?

Probably somewhere in Eastern Siberia, Mongolia, or the Dzungarian Basin and it got to Central Asia with Turks and Mongols. Tukrmenistan is in Western Central Asia so my point on M73 not having trveled through Turkmenistan and its present day distibution there is very recent is correct. So where is the disagreement?

Jean M
12-02-2013, 03:58 PM
Also I agree the mixed Gene pool was likely from Ukraine. If it was in the Altai then we would have likely seen a variety of east Eurasian mtdna lines.

There are a variety of lineages there now, but some of them arrived later than the period in question, it would appear.

The pre-Neolithic population of Europe is dominated by subclades of mtDNA U. It seems a pretty significant exception that C is found only in peoples in Eastern Europe who can be shown to have had contact with Central Asia. The Dnieper-Donets I people made pointed-based pottery of a type that arrived from Lake Baikal. C and C4a2 are found in east and Central Asia today. C probably expanded from the Ice Age refuges around Lake Baikal and the Yenisei Valley in the Mesolithic. Today 39% of the Tubalar carry C4a2. They are thought to be descendants of hunter-gatherers who found refuge in the Altai-Sayan mountains. See Ancestral Journeys, p. 64 if you want references.

newtoboard
12-02-2013, 04:02 PM
The pre-Neolithic population of Europe is dominated by subclades of mtDNA U. It seems a pretty significant exception that C is found only in peoples in Eastern Europe who can be shown to have had contact with Central Asia. The Dnieper-Donets I people made pointed-based pottery of a type that arrived from Lake Baikal. C and C4a2 are found in east and Central Asia today. C probably expanded from the Ice Age refuges around Lake Baikal and the Yenisei Valley in the Mesolithic. Today 39% of the Tubalar carry C4a2. They are thought to be descendants of hunter-gatherers who found refuge in the Altai-Sayan mountains. See Ancestral Journeys, p. 64 if you want references.


That may all be true but it doesn't rule out the possibility that Xiaohe mtdna C is from Ukraine (even if it arrived there from Siberia during the Mesolithic). Also C and C4a2 might be found in Central Asia but there is no evidence it was found there in ancient times. And it is also found there toady alongside other exotic lineages that likely weren't there during the migration of Tocharians (mtdna A, B, and Z come to mind). The Bronze Age groups of Kazakhstan were fully West Eurasian for the most part.

newtoboard
12-02-2013, 04:05 PM
There are a variety of lineages there now, but some of them arrived later than the period in question, it would appear.

The pre-Neolithic population of Europe is dominated by subclades of mtDNA U. It seems a pretty significant exception that C is found only in peoples in Eastern Europe who can be shown to have had contact with Central Asia. The Dnieper-Donets I people made pointed-based pottery of a type that arrived from Lake Baikal. C and C4a2 are found in east and Central Asia today. C probably expanded from the Ice Age refuges around Lake Baikal and the Yenisei Valley in the Mesolithic. Today 39% of the Tubalar carry C4a2. They are thought to be descendants of hunter-gatherers who found refuge in the Altai-Sayan mountains. See Ancestral Journeys, p. 64 if you want references.


So you think it is more likely the other lineages in the Altai arrived there later but somehow the mtdna C was already present in Central Asia. It is more likely mtdna C arrived in Central Asia more recently than mtdna A, B or Z arriving later to the Altai imo.

alan
12-02-2013, 04:13 PM
All migrations are subsets of the original group so anything is possible. I think the quirky non-wave like distribution of M73 is quite a common phenomenon in R1b in general - something I think is suggestive of its original waves being old and their trials having been broken by later waves which have diltuted and displaced part of the original trail.

M73 in particular seems especially old and seems to have been a player substantially earlier than any of the substantial R1a clades. Given the R1 in general lack of branching from the Palaeolithic to the later Neolithic or later every branching with living descendants is highly significant IMO. As current dating stands, most of the R1a (and L23) clades known in eastern Europe and Asia seem to be telling us about the 3500BC or later periods. M73 appears to be telling us something about a period before this. A period which would in IE terms be in its archaic period. Virtually all other y lines seem to be telling us mainly about the post-3500BC period. There was a long history on the steppe before Yamnaya and it was a very big place and the steppe intrusions into the Balkans began a little before 4000BC. There is little point in trying to relate that sort of period to language groups on the steppe after 3000BC.


I hope you are not suggesting the R1b-M73 in Turkmenistan has some sort of ancient presence. If it did why don't we see much R1b-M73 in Iran?

Michał
12-02-2013, 04:18 PM
I said it is likely most of the mtdna U4, U5, and some H/T lineages in those areas trace to Proto Indo-Iranian speakers. But I guess specific subclades need to be looked at. But the Andronovo to Xiaohe question remains. All those areas you mentioned were populated and it is possible that movements south were mostly male mediated. But there is nothing that I can think of that accounts for the gene pool of Andronovo to change so radically in their extremely short journey from Central Asia into the Tarim (from carrying a mix of U, T, H, K and Z to nothing but H and C). They passed through nothing but open steppe really. And as we can see later Andronovo derived cultures did indeed carry at least U and T.
I agree that it makes some sense. On the other hand, if the C4 haplogroup that dominates in Xiaohe is of East European and not local Asian origin, then we have a problem with explaining its absence in Andronovo (where the apparent heterogeneity of the mtDNA pool makes it unlikely that it was simply related to the bottleneck/founder effect). This is one of the reasons that I consider the mtDNA data much less convincing than the Y-DNA ones.

Let me repeat that I don’t think Xiaohe is derived from Andronovo, only that the mtDNA data are of limited value when trying to verify such potential link.

Michał
12-02-2013, 04:29 PM
So can you find anything on the R1a phylo tree that would indicate a clear separation between the strange and enigmatic Tarim Basin mummies and Indo-Iranian R1a?
No, I cannot see anything like this, and this is exactly why it poses a significant problem from the purely linguistic point of view (unless those mummies were speaking an Indo-Iranian or closely related language when still alive).


Or do you think they just belonged to Z93?
If you mean Z93*, it is of course possible, but some known and very specific subclades of Z94 (that are known to be associated with different Indo-Iranian people) seem to be slightly more likely IMO.


But if they belonged to Z93, then why couldn't have the Tocharians?
Because there is nothing in the Tocharian language that suggests its close shared ancestry with the Indo-Iranian branch of the IE languages. On the contrary, it is commonly suggested that Indo-Iranian is more closely related to other European IE languages (including Balto-Slavic or Greek) than to Tocharian.



On the other hand, if the mummies were M417*, then why couldn't have the Tocharians been that as well?
If the mummies were M417*, this would actually support their potential association with the Tocharian-speaking people very strongly, but if they were Z93+ (or even Z94+ and Z2124+) this would make such association very unlikely.



Languages aren't genetic mutations, so you don't need Y-chromosome mutations to represent each language branch. I think that's the fundamental problem with your model. You wouldn't get it past any linguists.
You are of course right that we cannot simply equate each Y-DNA haplogroup with a specific language. Nevertheless, some correlations are expected to be seen (which you seem to agree with in all those cases when the genetic data fir your favourite hypothesis), so if the Y-DNA data are in strong opposition to the linguistic data (like in this specific case), we cannot ignore this when suggesting the most likely scenario.



Maybe that's why you've got Baltic and its predecessors on the periphery of the Indo-European tree, and yet linguists think that Lithuanian is the closest thing to PIE today?
Why do you think I place Baltic on a periphery of PIE? It is just one of several main sub-branches of Late PIE, and there was no way I could place all of them in the middle of my scheme. Can you show me any other modern IE branch on my scheme that shows a closer distance to the Late PIE node?
If there are any peripheral ancestors of PIE shown in my tree, these are definitely Anatolian and Tocharian, as they have been disconnected from the main branch very early on.



In any case, like I've already said, there's M17* in Tibet, near the Tarim Basin.
If M17* or M417* were found in Xiaohe and/or Afanasevo (not to mention among the actual Tocharian-speaking people of the 6th-9th century AD in the Tarim basin), I would certainly agree with you.

However, if those putative M17* and/or M417* Tocharians were so widely distributed in Siberia, Dzungarian Basin and Tarim Basin, why there is not much left of them in Asia today, especially when assuming that the Yuezhi/Kushan people should bring it to Central Asia and India? Also, if those ancient R1a people from Siberia (Afanasevo) and China (Xiaohe) were indeed R1a(xZ93), how would you explain that both modern R1a Chinese samples from the recent study were Z94+ (representing two distantly related subclades).

Could anyone please give me a link to the study that reports M17* being present in Tibet? Are these SNP data accompanied by any STR haplotypes that would confirm this finding.

Michał
12-02-2013, 04:38 PM
I would bring the early PIE (but not the archaic PIE) branch out of R1a.
Is this because of the "Tocharian evidence" or you see some additional data to support it?


Plus where would you place R2-M479. Its association/correlation with the earliest forms of IE in South Asia is very strong.
I think this is related to the relatively recent inclusion of R2 into the expanding early Indo-Iranian (mostly Indo-Aryan) population. On the other hand, I wouldn't exclude the possibility that at least some R2 members became a part of the Pre-Anatolian (or earlier Archaic PIE) group in the East Caspian region, so they could also be represented among the Anatolian-speaking people who migrated to Anatolia about 2000-1700 BC.

Michał
12-02-2013, 04:43 PM
Do you have a text for you model.
I have nothing that would be written specifically to describe this tree, but all of it is consistent with what I wrote in one of my previous posts in another thread:
http://www.anthrogenica.com/showthread.php?1026-what-is-the-latest-thinking-on-were-R1a-originated/page17&p=10587#post10587

newtoboard
12-02-2013, 04:51 PM
I agree that it makes some sense. On the other hand, if the C4 haplogroup that dominates in Xiaohe is of East European and not local Asian origin, then we have a problem with explaining its absence in Andronovo (where the apparent heterogeneity of the mtDNA pool makes it unlikely that it was simply related to the bottleneck/founder effect). This is one of the reasons that I consider the mtDNA data much less convincing than the Y-DNA ones.

Let me repeat that I don’t think Xiaohe is derived from Andronovo, only that the mtDNA data are of limited value when trying to verify such potential link.



Not if Andronovo groups have most of their origins in European Russia while Afanasievo groups had their origins in Ukraine. Also there is the possibility all the mtdna C migrated east creating a vacuum for pre Andronovo groups to mix with farmers (as seen from the presence of Neolithic groups) from the Balkans.

newtoboard
12-02-2013, 04:54 PM
Is this because of the "Tocharian evidence" or you see some additional data to support it?


I think this is related to the relatively recent inclusion of R2 into the expanding early Indo-Iranian (mostly Indo-Aryan) population. On the other hand, I wouldn't exclude the possibility that at least some R2 members became a part of the Pre-Anatolian (or earlier Archaic PIE) group in the East Caspian region, so they could also be represented among the Anatolian-speaking people who migrated to Anatolia about 2000-1700 BC.


There isn't much R2 in Anatolia and much of it likely has a recent origin from Central Asia or is older than the Anatolian branch of languages.

Silesian
12-02-2013, 04:54 PM
Why should he have to post a model?

So we can exchange ideas.


Because he disagrees with someone else's? As far as I am aware nobody on this forum is a professional linguist so what is the value in these models.

His model is unique. It is the first in attempting to combine two different fields of discipline.


Plus there is no reason a linguistic model should have to account for haplogroups as well as areal features such as satemization..

Can you put forth your model in a rough form? For example does PIE in your model originate in the Baltic region?

newtoboard
12-02-2013, 05:01 PM
So we can exchange ideas.



His model is unique. It is the first in attempting to combine two different fields of discipline.



Can you put forth your model in a rough form? For example does PIE in your model originate in the Baltic region?


Sorry but that is not a good thing imo. There is no reason to have a lingustic branching tree based on Y-DNA and areal features.

alan
12-02-2013, 05:03 PM
The only group in the entire Altai study with much if any M73 is a single Turkic tribe who retreated to Altai c. 1600AD in the face of the Russians. None of the other north Altai groups features it and there was none in south Altai. I was surprised myself by the Altai study but there it is. It seems that the Turks picked it up well to the west.

Add to that that it cannot have reached the area from the Hindu Kush zone or Iran and it sort of bookends it towards west central Asia, the Urals etc. I concluded recently based on all these new studies that present distribution suggests a location within 500miles of the Aral Sea. I admit thats vague and stretches from the south Urals to Turkmenistan but I dont think it can be narrowed down any more.

I am glad you clarrified that you do think M73 is some sort of north Asian group anyway as its pretty well now been ruled out as coming from the south. At least we broadly agree on latitude if not longitude. Anyway, an important aspect of this all is that M269 is M73's closest brother clade and they both lived in the P297* state throughout the Neolithic period of SW Asia and left no descendants. So, it is likely that they shared a similar environment in the Neolithic outside the farming zone. Given M73's absence in Iran and M269's very north-western concentration in Iran it seems very likely to me that M269 made its first entry into Iran from the north down the west side of the Caspian c. 3500BC and IMO by far the best archaeological fit for that is the Maykop kurgans that spill from the steppe interface with the north Caucasus down into NW Iran. Those Kurgans are not as early as the Maykop culture as a whole so they appear to have come from north to south.

I think triangulating between M73 and M269 that a position for P297* close to the northern half of the Caspian is quite likely with M73 staying put or heading east and M269 moving west through the western steppes and south through the Caucasus fits the geography best. Otherwise it is incredibly hard to explain their common P297* phase in the Neolithic 9000-5000BC. Its not possible to give them different starting points as phylogeny shows they shared a single ancestor somewhere c. 9000BC and may have remained close to each other for some time after. So the diverging distribution post dates 9000BC and to explain that we need to explain in a post 9000BC timeframe how that happened. To place M73 in Mongolia (which looks incredibly unlikely given modern studies) you need to explain how it got there and why M269 ended up far more western. I dont think there is any model that can explain this from a starting point so far east. There is no movement in the right sort of timeframe currently envisaged by archaeologists that would explain such a scenario. I cannot think of a movement from the far east to the west in the required timeframe 9000-4000BC that would explain how M269 ended up so far west of its brother clade. I also cannot think of a reverse scenario of a movement from the west towards somewhere like Altai in the 9000-5000BC that would explain M73 being there.

An intermediate point between the current maximum M269*/L23xL51 and M73 distributions seems far more plausible for their common ancestor even if that is a very vague zone. That zone IMO is somewhere around the Caspian.

EDIT-the only way I can remotely see a link to as far east as Altai and adjacent is is we rewind the clock back all the way into the Palaeolithic in pre-P297* times and see some link to the spread of the late upper palaeolithic microblade tradition that reached European Russia around (I vaguely recall) 12000BC and after and had much older roots in Altai and also spread towards Beringia apparently with Q men. However, this spread is not one that readily links to any particular R1 groups as its way earlier than any surviving clades in that zone.

Its infinately more likely that M73 arose much further west closer to where M269 rose.



Probably somewhere in Eastern Siberia, Mongolia, or the Dzungarian Basin and it got to Central Asia with Turks and Mongols. Tukrmenistan is in Western Central Asia so my point on M73 not having trveled through Turkmenistan and its present day distibution there is very recent is correct. So where is the disagreement?

Silesian
12-02-2013, 05:12 PM
Sorry but that is not a good thing imo. There is no reason to have a lingustic branching tree based on Y-DNA and areal features.

If not a branching tree what about associating language with specific species like R1a.Does that mean that Assyrians with less than 2% R1a were originally PIE speaking people, or that R1a in its early form was involved with Semitic speaking peoples ?

Michał
12-02-2013, 05:17 PM
Not if Andronovo groups have most of their origins in European Russia while Afanasievo groups had their origins in Ukraine.
Possible, but you should admit that there is nothing that makes it a more likely hypothesis than a suggested local Asian source of C4 in Xiaohe, especially when your scenario requires the Ukrainian "pre-Afansevo" group to pass through the region occupied by the Russian "pre-Andronovo" group.


Also there is the possibility all the mtdna C migrated east creating a vacuum for pre Andronovo groups to mix with farmers (as seen from the presence of Neolithic groups) from the Balkans.
So where would you exactly see the place of origin for this C-free pre-Andronovo group in Europe, assuming they were located west of the C-rich Dnieper-Donets group?

newtoboard
12-02-2013, 05:18 PM
The only group in the entire Altai study with much if any M73 is a single Turkic tribe who retreated to Altai c. 1600AD in the face of the Russians. None of the other north Altai groups features it and there was none in south Altai. I was surprised myself by the Altai study but there it is. It seems that the Turks picked it up well to the west.

Add to that that it cannot have reached the area from the Hindu Kush zone or Iran and it sort of bookends it towards west central Asia, the Urals etc. I concluded recently based on all these new studies that present distribution suggests a location within 500miles of the Aral Sea. I admit thats vague and stretches from the south Urals to Turkmenistan but I dont think it can be narrowed down any more.

I am glad you clarrified that you do think M73 is some sort of north Asian group anyway as its pretty well now been ruled out as coming from the south. At least we broadly agree on latitude if not longitude. Anyway, an important aspect of this all is that M269 is M73's closest brother clade and they both lived in the P297* state throughout the Neolithic period of SW Asia and left no descendants. So, it is likely that they shared a similar environment in the Neolithic outside the farming zone. Given M73's absence in Iran and M269's very north-western concentration in Iran it seems very likely to me that M269 made its first entry into Iran from the north down the west side of the Caspian c. 3500BC and IMO by far the best archaeological fit for that is the Maykop kurgans that spill from the steppe interface with the north Caucasus down into NW Iran. Those Kurgans are not as early as the Maykop culture as a whole so they appear to have come from north to south.

I think triangulating between M73 and M269 that a position for P297* close to the northern half of the Caspian is quite likely with M73 staying put or heading east and M269 moving west through the western steppes and south through the Caucasus fits the geography best. Otherwise it is incredibly hard to explain their common P297* phase in the Neolithic 9000-5000BC. Its not possible to give them different starting points as phylogeny shows they shared a single ancestor somewhere c. 9000BC and may have remained close to each other for some time after. So the diverging distribution post dates 9000BC and to explain that we need to explain in a post 9000BC timeframe how that happened. To place M73 in Mongolia (which looks incredibly unlikely given modern studies) you need to explain how it got there and why M269 ended up far more western. I dont think there is any model that can explain this from a starting point so far east. There is no movement in the right sort of timeframe currently envisaged by archaeologists that would explain such a scenario. I cannot think of a movement from the far east to the west in the required timeframe 9000-4000BC that would explain how M269 ended up so far west of its brother clade. I also cannot think of a reverse scenario of a movement from the west towards somewhere like Altai in the 9000-5000BC that would explain M73 being there.

An intermediate point between the current maximum M269*/L23xL51 and M73 distributions seems far more plausible for their common ancestor even if that is a very vague zone. That zone IMO is somewhere around the Caspian.

EDIT-the only way I can remotely see a link to as far east as Altai and adjacent is is we rewind the clock back all the way into the Palaeolithic in pre-P297* times and see some link to the spread of the late upper palaeolithic microblade tradition that reached European Russia around (I vaguely recall) 12000BC and after and had much older roots in Altai and also spread towards Beringia apparently with Q men. However, this spread is not one that readily links to any particular R1 groups as its way earlier than any surviving clades in that zone.

Its infinately more likely that M73 arose much further west closer to where M269 rose.


And who did they pick it up from then? Turkic expansion was quite late in history so Indo-Iranians are the only candidates. So what the steppe was like very recently is very relevant. Everything in that area was likely heavily if not entirely R1a dominated. Like Generalissimo has said why is R1b always hiding out in ancient samples from the region?

As for the second bolded statement what a weak argument. As if the world revolves around R1b. North of the Caspian and Black sea is also the perfect position for upstream R1a and Z645+. And I love how its somehow accepted as a fact that M269+ occurred in the steppes without the possibility it occurred in Asia. I wasn't aware anybody had narrowed it down. Its distribution is in the Balkans and around the Caspian. You can argue that means it spread from the steppe but that is just speculation since it isn't very strong in the North Caucasus or East Balkans. Also since when does everything had to fit geography. As if the spread of haplogroups is so perfect.

Btw Turkmenistan is in the South and was a part of the farming world. So according to you it hasn't been ruled out as coming out from the South.

newtoboard
12-02-2013, 05:19 PM
If not a branching tree what about associating language with specific species like R1a.Does that mean that Assyrians with less than 2% R1a were originally PIE speaking people, or that R1a in its early form was involved with Semitic speaking peoples ?

Assyrians are a Semitic people. What is your point? Are you suggesting R1a is associated with Semitic people? Early form? None of the Assyrian R1a is an early form.

Michał
12-02-2013, 05:19 PM
There isn't much R2 in Anatolia and much of it likely has a recent origin from Central Asia or is older than the Anatolian branch of languages.
Is there anything you have just written inconsistent with what I wrote regarding R2? I cannot see it.

parasar
12-02-2013, 05:22 PM
Is this because of the "Tocharian evidence" or you see some additional data to support it?

Yes the Tocharian evidence.

cemeteries at Xiaohe (Small River) and Qäwrighul, both of which are located in the same region as Tocharian C
http://penn.museum/documents/publications/expedition/PDFs/52-3/mallory.pdf



I think this is related to the relatively recent inclusion of R2 into the expanding early Indo-Iranian (mostly Indo-Aryan) population. On the other hand, I wouldn't exclude the possibility that at least some R2 members became a part of the Pre-Anatolian (or earlier Archaic PIE) group in the East Caspian region, so they could also be represented among the Anatolian-speaking people who migrated to Anatolia about 2000-1700 BC.

R2 and R1a1 both correlate well with Indo-Aryans speaking population. R2 correlates best with the older IE populations of eastern and south eastern India - cf. the Andhras. Historically the Andhras (Shatavahanas, the Ikshvakus, the Pallavas , the Salankayanas, the Vishnukundins, the Vengi Chalukyas etc.) were initially a Indo-Aryan peoples but today most are Dravidian speaking.

newtoboard
12-02-2013, 05:22 PM
Possible, but you should admit that there is nothing that makes it a more likely hypothesis than a suggested local Asian source of C4 in Xiaohe, especially when your scenario requires the Ukrainian "pre-Afansevo" group to pass through the region occupied by the Russian "pre-Andronovo" group.


So where would you exactly see the place of origin for this C-free pre-Andronovo group in Europe, assuming they were located west of the C-rich Dnieper-Donets group?


Not if the pre-Afansevo group went through the forest steppe. So far all I have read is about a migration from Yamnaya to the Minusinsk region. Nothing indicates they traveled in a straight line through the steppe.

Isn't it possible this C-free pre-Andronovo group was east of the Dnieper-Donets group but occupying the area of Dnieper-Donets when the pre-Afanasevo groups went east.

newtoboard
12-02-2013, 05:26 PM
Is there anything you have just written inconsistent with what I wrote regarding R2? I cannot see it.


Time frame and direction of migration. The R2 in Anatolia is more likely a recent Central Asian origin as in the past 1500 years (as it got there with Turks imo) or a very ancient movement west through Iran (so a Central Asia->Iran->Anatolia route) rather than the Central Asia-> Steppe->Balkans->Anatolia route you suggested.

Michał
12-02-2013, 05:31 PM
Sorry but that is not a good thing imo. There is no reason to have a lingustic branching tree based on Y-DNA and areal features.
This is not a linguistic tree based on Y-DNA but an attempt to combine different types of data (linguistic, genetic and archaeological) into one consistent model. Your are comparing DNA with linguistics and archaeology all the time, so I don't really understand why you find this approach "not good".

As for the Satem question, I have admitted on multiple occasions that Satemization can be an areal process, as is well documented for some unrelated languages. However, this does not mean that two Satem languages cannot be derived from a common ancestor, especially when they share some additional linguistic innovations (like the Ruki rule) and are associated with the same Y-DNA subclade (R1a-Z645).

Michał
12-02-2013, 05:40 PM
Not if the pre-Afansevo group went through the forest steppe. So far all I have read is about a migration from Yamnaya to the Minusinsk region. Nothing indicates they traveled in a straight line through the steppe.
Do you know any data that would indicate this relatively less likely migration route?


Isn't it possible this C-free pre-Andronovo group was east of the Dnieper-Donets group but occupying the area of Dnieper-Donets when the pre-Afanasevo groups went east.
But this scenario requires some additional assumptions (that are not self-evident at all), which of course doesn't make it impossible, but, nevertheless, makes it somehow less likely than some alternative theories.

newtoboard
12-02-2013, 05:43 PM
This is not a linguistic tree based on Y-DNA but an attempt to combine different types of data (linguistic, genetic and archaeological) into one consistent model. Your are comparing DNA with linguistics and archaeology all the time, so I don't really understand why you find this approach "not good".

As for the Satem question, I have admitted on multiple occasions that Satemization can be an areal process, as is well documented for some unrelated languages. However, this does not mean that two Satem languages cannot be derived from a common ancestor, especially when they share some additional linguistic innovations (like the Ruki rule) and are associated with the same Y-DNA subclade (R1a-Z645).


Because this approach is based on assumptions which actually have a lot of support against them. A common satem language and the existence of R1b Tocharians for one.

There is nothing I have read which derives Indo-Iranian and Balto-Slavic from a common ancestor. The most common grouping is NW Indo-European and SE Indo-European not R1a Satem spakers and R1b Everyone else.

Silesian
12-02-2013, 05:43 PM
Assyrians are a Semitic people. What is your point? Are you suggesting R1a is associated with Semitic people? Early form? None of the Assyrian R1a is an early form.
General has pointed out Baltic is oldest PIE. Assyrian as you point out is Semitic, Elamite is an isolate. You have to reconcile words and examples like below with your species of R1a[ not old], and PIE terms/cognates found in Sumeria[Kur].

http://azargoshnasp.net/recent_history/pan_turkist_philosophy/sumd/sumerianlatvian1.htm
http://en.wiktionary.org/wiki/kur
https://en.wikipedia.org/wiki/Kurios

newtoboard
12-02-2013, 05:44 PM
Do you know any data that would indicate this relatively less likely migration route?


But this scenario requires some additional assumptions (that are not self-evident at all), which of course doesn't make it impossible, but, nevertheless, makes it somehow less likely than some alternative theories.

I agree it is unlikely. It was a suggestion but I don't really believe it. But either way the Xiaohe people are unlikely to be Andronovo derived at least based on mtdna.

Michał
12-02-2013, 05:47 PM
Time frame and direction of migration. The R2 in Anatolia is more likely a recent Central Asian origin as in the past 1500 years (as it got there with Turks imo) or a very ancient movement west through Iran (so a Central Asia->Iran->Anatolia route) rather than the Central Asia-> Steppe->Balkans->Anatolia route you suggested.
What kind of calculations make you believe so strongly that R2 came to Anatolia either before the IE-speaking people or after the IE-speaking people but not with the IE-speaking people. I am not telling that you cannnot be right in this respect, only that I am very surprised that you are so confident about it. Please share these data (calculations) with me, so I could share your confidence.

Michał
12-02-2013, 05:57 PM
There is nothing I have read which derives Indo-Iranian and Balto-Slavic from a common ancestor.
I am quite surprised that you haven't read the book by David Anthony (where Indo-Iranian and Balto-Slavic are shown as closely related) or that you haven't even looked at the well-known IE tree proposed by Warnow that is shown in the first post of this thread. You have a right to question their views, but neglecting the existence of some commonly accepted linguistic theories is simply not fair.

alan
12-02-2013, 05:58 PM
I am pretty sure that the best guess of the roots of Afanasievo do not place it in the Ukraine. I dont keep up very closely with this but I recall more the end of the Repin and Khvalynsk culture in the Ural Volga area in European Russia.


Not if Andronovo groups have most of their origins in European Russia while Afanasievo groups had their origins in Ukraine. Also there is the possibility all the mtdna C migrated east creating a vacuum for pre Andronovo groups to mix with farmers (as seen from the presence of Neolithic groups) from the Balkans.

newtoboard
12-02-2013, 06:04 PM
What kind of calculations make you believe so strongly that R2 came to Anatolia either before the IE-speaking people or after the IE-speaking people but not with the IE-speaking people. I am not telling that you cannnot be right in this respect, only that I am very surprised that you are so confident about it. Please share these data (calculations) with me, so I could share your confidence.

If I recall the majority of Anatolian R2 belongs to a pan West Asian cluster and a portion of it is also L295+ which most people think originated in NW South Asia or South Central Asia.

newtoboard
12-02-2013, 06:10 PM
I am quite surprised that you haven't read the book by David Anthony (where Indo-Iranian and Balto-Slavic are shown as closely related) or that you haven't even looked at the well-known IE tree proposed by Warnow that is shown in the first post of this thread. You have a right to question their views, but neglecting the existence of some commonly accepted linguistic theories is simply not fair.


I did look at the tree on the first post of this thread. It might be good based on chronology and which languages migrated first. But the concept of NW Indo-European is quite accepted these days as well so forgive me if I don't believe in some sort of Proto Satem lanuage spoken by R1a-Z645+ speakers. If Satemization is areal (which I believe it is) then that makes it unlikely. Indo-Iranian is also more satem than Balto-Slavic.

Michał
12-02-2013, 06:10 PM
If I recall the majority of Anatolian R2 belongs to a pan West Asian cluster and a portion of it is also L295+ which most people think originated in NW South Asia or South Central Asia.
I still don't know why this definitely excludes its association with the IE-speakers.

Michał
12-02-2013, 06:30 PM
Yes the Tocharian evidence.
http://penn.museum/documents/publications/expedition/PDFs/52-3/mallory.pdf

I can't help that I find it not convincing enough. For example, in the above-linked paper, Mallory admits that:
"In terms of distribution, the Saka sites tend to lie to the north, east, or south of where most of the mummified remains have been recovered; however, they have also been identified among the later burials at Alwighul. The tall hats of the female mummies from Subeshi might also pass for a Saka trait, and so identification of some of the mummies with the Saka or Iranian speakers in the northeast Tarim is a serious possibility."

alan
12-02-2013, 06:45 PM
You skipped over the problem. You cannot have M73 arising in Mongolia or thereabouts and M269 originating in SW Asia without explaining how those two brother clades ended up at the opposite ends of Asia despite sharing a recent common ancestor at the start of the Neolithic. How do you explain that when there is no evidence of a west to east movement before 3400BC or an east to west movement from either of these extremes that could explain it.

I am not saying the present distribution of M73 is in any way identical to what it was in 5000BC. That would be crazy. It probably been massively distorted. However, there is a general pattern that suggests a location much further west than Altai. It is boxed in by the Hindu Kush and Iran to the south and even at low levels it drops of to practically zero at the Dniester. It is unknown except in one group in northern Altai who retreated west (this is a matter of historical fact - I posted the details before). How far west is a bit vague as the Turkic group had influence from western Siberia to south Russia.

I dont know why you keep bringing up the Indo-Iranians as if they always were a stumbling block. M73 dates from at least 5000BC and M269 from 4000BC. They could have moved about prior to the cultures associated with Indo-Iranian and by different routes and also subsequently been moved about from their old positions by later waves. That may have changed the details but there is still a general pattern for M73 in that it is located north of the Hindu Kush/Iran/the Caucasus and west of Altai and east of the Dnieper.

Also, you yourself have often said that mono-clade clans were involved. Although I doubt that, if it were true then it applies to M73 too and a clan will go where a clan decides to go or is impelled to go. Every clan doesnt move when another clan moves.


And who did they pick it up from then? Turkic expansion was quite late in history so Indo-Iranians are the only candidates. So what the steppe was like very recently is very relevant. Everything in that area was likely heavily if not entirely R1a dominated. Like Generalissimo has said why is R1b always hiding out in ancient samples from the region?

As for the second bolded statement what a weak argument. As if the world revolves around R1b. North of the Caspian and Black sea is also the perfect position for upstream R1a and Z645+. And I love how its somehow accepted as a fact that M269+ occurred in the steppes without the possibility it occurred in Asia. I wasn't aware anybody had narrowed it down. Its distribution is in the Balkans and around the Caspian. You can argue that means it spread from the steppe but that is just speculation since it isn't very strong in the North Caucasus or East Balkans. Also since when does everything had to fit geography. As if the spread of haplogroups is so perfect.

Btw Turkmenistan is in the South and was a part of the farming world. So according to you it hasn't been ruled out as coming out from the South.

alan
12-02-2013, 06:58 PM
R1b is perhaps hiding in ancient DNA because most of the testing focuses on likely or possibly Indo-Iranian cultures of the Bronze Age pot-dating 2000BC usually. Testing is generally too late, and too eastern IMO to remotely rule out R1b on the steppe. There are huge numbers of cultures across the steppes that have not been tested. There could have been any number of languages that have risen and fallen and disappeared in the 3000 years between M73 and those tests. Only once they have tested the tons of cultures across the steppe we can observe R1b's absence as being meaningful.


You skipped over the problem. You cannot have M73 arising in Mongolia or thereabouts and M269 originating in SW Asia without explaining how those two brother clades ended up at the opposite ends of Asia despite sharing a recent common ancestor at the start of the Neolithic. How do you explain that when there is no evidence of a west to east movement before 3400BC or an east to west movement from either of these extremes that could explain it.

I am not saying the present distribution of M73 is in any way identical to what it was in 5000BC. That would be crazy. It probably been massively distorted. However, there is a general pattern that suggests a location much further west than Altai. It is boxed in by the Hindu Kush and Iran to the south and even at low levels it drops of to practically zero at the Dniester. It is unknown except in one group in northern Altai who retreated west (this is a matter of historical fact - I posted the details before). How far west is a bit vague as the Turkic group had influence from western Siberia to south Russia.

I dont know why you keep bringing up the Indo-Iranians as if they always were a stumbling block. M73 dates from at least 5000BC and M269 from 4000BC. They could have moved about prior to the cultures associated with Indo-Iranian and by different routes and also subsequently been moved about from their old positions by later waves. That may have changed the details but there is still a general pattern for M73 in that it is located north of the Hindu Kush/Iran/the Caucasus and west of Altai and east of the Dnieper.

Also, you yourself have often said that mono-clade clans were involved. Although I doubt that, if it were true then it applies to M73 too and a clan will go where a clan decides to go or is impelled to go. Every clan doesnt move when another clan moves.

alan
12-02-2013, 07:19 PM
There is also a serious missing link between Afansievo cultures that cut across the steppes c. 3400BC and ended up in the gates of Altai and mummies from far latter in Tarim, a very different sort of route used by traders. For all we know the Tarim mummies, Afansievo and Tocharian could relate to three different things. I am not sure we can ever prove what they were speaking although the genetics of Afanasievo will emerge when actual burials of that culture are known.

One thing gave me pause for thought about Afanasievo being linked to Tocharian. Now that it has yet again been re-dated to being not such an early split and now dated to around the same time as Yamnaya arose, it could be argued to be less of a perfect match for the linguistic splitting. Another thing Anthony indicated was there was evidence of continuing cultural flow from the Yamnaya world. Neither quite fits the early split/isolate model quite as well as sometimes portrayed. I would tend to think that the redating and continuing contact would imply that they were rather similar to the other Yamnaya groups in terms of linguistic branch.




I can't help that I find it not convincing enough. For example, in the above-linked paper, Mallory admits that:
"In terms of distribution, the Saka sites tend to lie to the north, east, or south of where most of the mummified remains have been recovered; however, they have also been identified among the later burials at Alwighul. The tall hats of the female mummies from Subeshi might also pass for a Saka trait, and so identification of some of the mummies with the Saka or Iranian speakers in the northeast Tarim is a serious possibility."

newtoboard
12-02-2013, 07:32 PM
R1b is perhaps hiding in ancient DNA because most of the testing focuses on likely or possibly Indo-Iranian cultures of the Bronze Age pot-dating 2000BC usually. Testing is generally too late, and too eastern IMO to remotely rule out R1b on the steppe. There are huge numbers of cultures across the steppes that have not been tested. There could have been any number of languages that have risen and fallen and disappeared in the 3000 years between M73 and those tests. Only once they have tested the tons of cultures across the steppe we can observe R1b's absence as being meaningful.

Wrong on both counts if we are talking about how Turks abosrbed it. Turks would have picked it up very late since Turkic expansion is very late. And certainly much later than any Andronovo sample. Also by that period Iranian speakers had back migrated into Europe so they were the people north of the Black and Caspian Seas and well beyond. So where was M73 hiding in between Andronovo and the time it is picked up from Scythians by Turks? And where was it hiding in between Andronovo and 5000 BC?

alan
12-02-2013, 07:32 PM
I think for now we should not use the Afanasievo-Tocharian-Tarim link as a solid plank for any theory. Its intriguing but its not a solid foundation to build on. It is impossible to infer with certainty the exact prior cultural origin of the Tarim mummies and despite already being in possession of DNA we cannot know what they spoke other than being Europeans and R1a it was probably IE. Its not like we can work back from a living Tocharian group and look for similarities. Maybe if we can test the actual graves of known Tocharian speakers from later times and get some DNA from actual Afansievo burials this story will come into focus. However, for now its just an interesting uncertain sideshow.

alan
12-02-2013, 07:59 PM
You tell me. You keep excluding anywhere remotely plausible for M73. I do not remotely believe that all previous populations were swept around with Indo-Iranians anyway. They didnt do a fingertip search for anyone not of their clan and exterminate everyone they could find. Well I hope not - wouldnt want genocidal ancestors. Many older steppe cultures are known to have lingered on very late in the steppes in much reduced pockets millenia after their heyday. So, there was no general wipeout of everything that went before. Movements of Iranians may have been carried out by certain elite lineages of warriors wanting their own patch to rule and they may well not have brought representative selection with them.

That concept works for R1b too. It seems likely this happened with L23 and Armenians for example who had y lines very like R1b in the Balkans but didnt bring the general population they may have conquered with them. In fact no matter where you place L23's expansion point its clear they just spread as an elite and didnt move large chunks of non-L23 people with them. If entire populations had been moved with the R1b expansions there would have been a simultaneous expansion of non R1b lineages from the origin point, be that steppe, Balkans, SW Asia or whatever. That does not seem to have happened. The pattern seems to have been branches of elites transplanted themselves to rule new territories then started to breed heavily at each new conquered location.

You can even see this still at work in the clans in Ireland in the Medieval period. When they expanded, it was the chiefly lineage that expanded. They didnt appear to move populations they had previously conquered with their further expansions. It basically worked as an elite expansion that then transplanted a subset of the elite lineage when they conquered new territories and then this subset expanded in the new locality.

So, I dont think we need to worry about potential earlier steppe substrate populations not being dragged around by the elites who conquered them and therefore all this stuff about M73 or whatever not being dragged around with the Iranians is not an issue at all.


Wrong on both counts if we are talking about how Turks abosrbed it. Turks would have picked it up very late since Turkic expansion is very late. And certainly much later than any Andronovo sample. Also by that period Iranian speakers had back migrated into Europe so they were the people north of the Black and Caspian Seas and well beyond. So where was M73 hiding in between Andronovo and the time it is picked up from Scythians by Turks? And where was it hiding in between Andronovo and 5000 BC?

newtoboard
12-02-2013, 08:08 PM
You tell me. You keep excluding anywhere remotely plausible for M73. I do not remotely believe that all previous populations were swept around with Indo-Iranians anyway. They didnt do a fingertip search for anyone not of their clan and exterminate everyone they could find. Well I hope not - wouldnt want genocidal ancestors. Many older steppe cultures are known to have lingered on very late in the steppes in much reduced pockets millenia after their heyday. So, there was no general wipeout of everything that went before. Movements of Iranians may have been carried out by certain elite lineages of warriors wanting their own patch to rule and they may well not have brought representative selection with them.

That concept works for R1b too. It seems likely this happened with L23 and Armenians for example who had y lines very like R1b in the Balkans but didnt bring the general population they may have conquered with them. In fact no matter where you place L23's expansion point its clear they just spread as an elite and didnt move large chunks of non-L23 people with them. If entire populations had been moved with the R1b expansions there would have been a simultaneous expansion of non R1b lineages from the origin point, be that steppe, Balkans, SW Asia or whatever. That does not seem to have happened. The pattern seems to have been branches of elites transplanted themselves to rule new territories then started to breed heavily at each new conquered location.

You can even see this still at work in the clans in Ireland in the Medieval period. When they expanded, it was the chiefly lineage that expanded. They didnt appear to move populations they had previously conquered with their further expansions. It basically worked as an elite expansion that then transplanted a subset of the elite lineage when they conquered new territories and then this subset expanded in the new locality.

So, I dont think we need to worry about potential earlier steppe substrate populations not being dragged around by the elites who conquered them and therefore all this stuff about M73 or whatever not being dragged around with the Iranians is not an issue at all.

It is the steppe. Tribes move as part of a domino effect. But I don't see the need to make up a story about M73 being present in steppe Indo-Iranians. They apparently absorbed some K and C but the R1b was hiding out? Why did K and C manage to make it to the elites of the society but the closely related R1b-M73 carriers (in terms of autosomal admixture I assume they were similar) didn't? Why is only R1b hiding out?

newtoboard
12-02-2013, 08:26 PM
I am not completley excluding the possibility that R1b-M73 was more Western but it should not be considered a sure thing or even the most likely theory until some M73 pops up in steppe/South Urals samples.

TigerMW
12-02-2013, 08:42 PM
I am not completley excluding the possibility that R1b-M73 was more Western but it should not be considered a sure thing or even the most likely theory until some M73 pops up in steppe/South Urals samples.

What do you think is the most likely explanation for M73?

newtoboard
12-02-2013, 08:57 PM
What do you think is the most likely explanation for M73?


It was somewhere in the Dzungarian Basin or Eastern Siberia. I also wouldn't exclude Gansu, Mongolia or the Russian Far East. Much of the R1b-M73 seems to have been described as having undergone a founder effect in Turks. So much of the the modern R1b-M73 today could be the result of just one or two guys in the past 2000 years. Turko-Mongol groups seem to be very well suited for founder effects (R1a in Kyrgyzstan, C3 among Mongols). We don't have any DNA samples from the regions above for the most part. And just because there is no R1b-M73 there today doesn't mean there wasn't before. Steppe tribes are very mobile. There is no R1a-Z93+ in the Russian forest steppe or north of the Caspian/Black Seas for the most part but it likely was once there.

Michał
12-02-2013, 09:53 PM
Wrong on both counts if we are talking about how Turks abosrbed it. Turks would have picked it up very late since Turkic expansion is very late. And certainly much later than any Andronovo sample.
This is a very good point, but it still doesn't explain why R1b-M73 is missing in most those Eastern pre-Turk sites like Southern Siberia (Tagar) or Tarim basin (Xiaohe) that you suggest as some possible centers of its expansion.



Also by that period Iranian speakers had back migrated into Europe so they were the people north of the Black and Caspian Seas and well beyond.
The same region in Europe was subsequently occupied by different Turkic tribes, so if this is supposed to be a proof for the absence of M73 among the Iranian tribes the same can be said about the Turkic tribes.

Actually, R1b-M73 is present in Ukraine, though at relatiively low frequency. IIRC, the recent paper on Moldavian Y-DNA reports some M73 examples in Western Ukraine. It is of course hard to say whether this is of Iranian or Turkic origin, though I think the latter is more likely.



So where was M73 hiding in between Andronovo and the time it is picked up from Scythians by Turks?
It seems that nobody knows it, but what you suggest in another post is also not confirmed by any aDNA data yet.


And where was it hiding in between Andronovo and 5000 BC?
I think Tersek-Botai seems to be quite likely option, though I agree that with no aDNA data this is just a speculation.

Michał
12-02-2013, 10:00 PM
I am not completley excluding the possibility that R1b-M73 was more Western but it should not be considered a sure thing or even the most likely theory until some M73 pops up in steppe/South Urals samples.
IMO, the very ancient origin of M73 somewhere close to the South Ural region is strongly favored by the fact that in addition to the two potential Asian subclades of M73, we also have a much smaller, though relatively old (!), potential Western European subclade of M73. There is simply no way it was brought there with some Turkic people.

alan
12-02-2013, 10:00 PM
I have already shown in my post that R1 lines appear to have spread as elite lineages on their conquests. They do not appear to have frequently picked up the substrate. You yourself often have posted about mono-lineage populations. That could only have happened in the copper age if these elites basically ruled who they conquered and only rarely encorporated the conquered substrate into their subsequent expansions. It certainly worked that way in Europe for R1b and the Kurgan evidence seems to suggest the R1a military elites did not hugely absorbed their conquered populations into the elite back in the copper and Bronze Age. Citing a couple of exceptions doesnt much change that basic pattern.

The elites probably made clients or otherwise extracted tribute from who they conquered but as a patrilineal society they probably rarely absorbed them. Its even less likely if they were perceived as enemies. So, as far as I am concerned who a particular language speaking elite passed through doesnt give any consistent expectations as to what they should have been absorbed and passed on in their further travels. Its not a safe approach to make inferences from.




It is the steppe. Tribes move as part of a domino effect. But I don't see the need to make up a story about M73 being present in steppe Indo-Iranians. They apparently absorbed some K and C but the R1b was hiding out? Why did K and C manage to make it to the elites of the society but the closely related R1b-M73 carriers (in terms of autosomal admixture I assume they were similar) didn't? Why is only R1b hiding out?

Michał
12-02-2013, 10:03 PM
It was somewhere in the Dzungarian Basin or Eastern Siberia. I also wouldn't exclude Gansu, Mongolia or the Russian Far East. Much of the R1b-M73 seems to have been described as having undergone a founder effect in Turks. So much of the the modern R1b-M73 today could be the result of just one or two guys in the past 2000 years.
You are certainly right regarding those founder effects seen for the major R1b-M73 lineages present in the Eurasian steppe (and, as you note, the same is seen for most R1a lineages dominating in the steppe environment). Nevertheless, it seems obvious that the interclade age for Asian M73 is relatively high, and their common ancestor seems to significantly predate the Turkic expansion (and maybe even the Indo-Iranian expansion, as well). And this is without including the more distantly related Western European cluster (potential subclade) of M73, which would make M73 one of the oldest subclades in both R1b and R1a.



Turko-Mongol groups seem to be very well suited for founder effects (R1a in Kyrgyzstan, C3 among Mongols). We don't have any DNA samples from the regions above for the most part. And just because there is no R1b-M73 there today doesn't mean there wasn't before. Steppe tribes are very mobile. There is no R1a-Z93+ in the Russian forest steppe or north of the Caspian/Black Seas for the most part but it likely was once there.
This is all very true, and I agree with you that the question of ancient origin and some subsequent migrations of R1b-M73 in Asia is still open.

alan
12-02-2013, 10:11 PM
I cannot understand why you would show a preference for an apparently M73-free area around Altai compared to the much stronger representation of M73 in the south Urals and west central Asia. I doubly cannot understand such a conclusion when M73 had a common ancestor with M269 in the Neolithic. It just is so counterintuitive. Any rational guess based on even the broadest look at the distribution would put M73 closer to the Urals or the west end of central Asia with M269 a little further west. If I had to take a final guess at its position of origin I would like Michal place M73 around the south Urals area and M269 to the west in an area with both access to the Balkans and to NW Iran and beyond by passing from the north Caucasus to Iran - probably somewhere like around the Sea of Azov.


It was somewhere in the Dzungarian Basin or Eastern Siberia. I also wouldn't exclude Gansu, Mongolia or the Russian Far East. Much of the R1b-M73 seems to have been described as having undergone a founder effect in Turks. So much of the the modern R1b-M73 today could be the result of just one or two guys in the past 2000 years. Turko-Mongol groups seem to be very well suited for founder effects (R1a in Kyrgyzstan, C3 among Mongols). We don't have any DNA samples from the regions above for the most part. And just because there is no R1b-M73 there today doesn't mean there wasn't before. Steppe tribes are very mobile. There is no R1a-Z93+ in the Russian forest steppe or north of the Caspian/Black Seas for the most part but it likely was once there.

parasar
12-02-2013, 10:59 PM
I cannot understand why you would show a preference for an apparently M73-free area around Altai compared to the much stronger representation of M73 in the south Urals and west central Asia. I doubly cannot understand such a conclusion when M73 had a common ancestor with M269 in the Neolithic ...

I don't think either M73 or M269 had much to do with Asia prior to the European/Anatolian contacts with Asia (abt 330bc).

As far as M73 goes I recall that VV had mentioned that it moved from Europe (incld. Anatolia) eastwards.



A phylogenetic analysis of the R1b1b Project participants' haplotypes reveals three distinct clusters. The clusters are distinct enough that I am identifying them as nesting subclades. In other words, the third cluster is a subclade of the second. And the second cluster, in turn, is a subclade of the first. Here is a NeighborNet network of the participants, which illustrates the clustering.
1st 2nd 3rd
A, B1, and B2, respectively - http://www.familytreedna.com/public/R1b1b1/default.aspx?section=yresults

http://www.vizachero.com/R1b1b/neighbornet.png

Generalissimo
12-02-2013, 11:35 PM
Because there is nothing in the Tocharian language that suggests its close shared ancestry with the Indo-Iranian branch of the IE languages. On the contrary, it is commonly suggested that Indo-Iranian is more closely related to other European IE languages (including Balto-Slavic or Greek) than to Tocharian.

My point was that the Tarim Basin mummies derived from a very early population that moved east, probably from what is now Ukraine (hence the R1a and C4 combo). Indeed, they were in the Tarim Basin earlier than the Andronovo people in South Siberia. So if they weren't the Proto-Tocharians, then they were even more distinct than the Proto-Tocharians, and yet, we know for a fact they belonged to R1a.

It might also be useful to remember that the Tarim Basin mummies were found in a burial complex that was isolated from other archeological sites. That's where they went to bury their dead, and didn't inhabit that area of the Tarim Basin. So they might well have come from a region that eventually became a well known Tocharian site, and I don't think pointy hats are enough to posit them as Indo-Iranians, considering the chronology, and their very different mtDNA structure to those of the Andronovo and Scythian nomads.


However, if those putative M17* and/or M417* Tocharians were so widely distributed in Siberia, Dzungarian Basin and Tarim Basin, why there is not much left of them in Asia today, especially when assuming that the Yuezhi/Kushan people should bring it to Central Asia and India? Also, if those ancient R1a people from Siberia (Afanasevo) and China (Xiaohe) were indeed R1a(xZ93), how would you explain that both modern R1a Chinese samples from the recent study were Z94+ (representing two distantly related subclades).

Population shifts? They happened in Europe, so why not in East Central Asia?

The two Chinese R1a are obviously of recent Turkic origin, and it's very difficult to say where their ancient Indo-European ancestors lived during the Bronze Age. But probably somewhere around the Altai.


Could anyone please give me a link to the study that reports M17* being present in Tibet? Are these SNP data accompanied by any STR haplotypes that would confirm this finding.

http://mbe.oxfordjournals.org/content/early/2013/05/16/molbev.mst093.short

http://2.bp.blogspot.com/-WFcfoMpfMrM/UZZzVmicAlI/AAAAAAAAIz4/IWTnu_ONSyQ/s1600/tibet.jpg


The pre-Neolithic population of Europe is dominated by subclades of mtDNA U. It seems a pretty significant exception that C is found only in peoples in Eastern Europe who can be shown to have had contact with Central Asia. The Dnieper-Donets I people made pointed-based pottery of a type that arrived from Lake Baikal. C and C4a2 are found in east and Central Asia today. C probably expanded from the Ice Age refuges around Lake Baikal and the Yenisei Valley in the Mesolithic. Today 39% of the Tubalar carry C4a2. They are thought to be descendants of hunter-gatherers who found refuge in the Altai-Sayan mountains. See Ancestral Journeys, p. 64 if you want references.

You're confusing the origins of mtDNA C with the origins of the Tarim Basin mummies.

The Tarim Basin mummies didn't have to originate in the same place as the most basal mtDNA C clades tens of thousands of years before them. They could have come from a place where mtDNA C was common much later.

The Ukraine is a good bet, considering the presence of mtDNA C4 in ancient pre-Tarim Basin mummy remains there.

alan
12-03-2013, 12:25 AM
Thanks for posting that. There are few options a movement in that direction from the archaeological record. By definition any movement into Asia by M73 had to happen after 5000BC. It is perhaps rather like some of the apparently early branchings of R1a which have been preserved in western Europe at very low levels far from the likely origin point. I looked with interest at the Moldova study that basically didnt find any except among the Ukrainian minority near the Ukraine border there.



I don't think either M73 or M269 had much to do with Asia prior to the European/Anatolian contacts with Asia (abt 330bc).

As far as M73 goes I recall that VV had mentioned that it moved from Europe (incld. Anatolia) eastwards.



1st 2nd 3rd
A, B1, and B2, respectively - http://www.familytreedna.com/public/R1b1b1/default.aspx?section=yresults

http://www.vizachero.com/R1b1b/neighbornet.png

alan
12-03-2013, 12:40 AM
So, you seem to be distinguishing between the Tarim mummies and the Afanasievo in terms of origin because Afanasievo would normal be traced back to Repin at the Volga-Urals end of the western steppe. If so, I dont disagree because I think its possible that Afanasievo in the Altai, the Tarim mummies and Tocharian may not be one group. I think possibly we have those three items and the temptation to join the dots is strong but not certainly correct. These three things involve such major chronological and geographical missing links that I cannot see this ever being more than an interesting hypothesis. The best that can be done IMO is to test actual Afanasievo burials and compare them with the Tarim ones to see if they are from the same lineages. It still will not prove the link with Tocharian though - that is essentially impossible unless the bones of much later peoples of the historic era actually known to be tocharian speakers exist and can be tested and compared.

The last bit is a big problem. Unless you have modern speakers or burials that are unambiguously known to belong to later speakers recorded in history then the normal method of working back step by step from existing populations of a living language cannot be carried out. Do we have a definate location at a definate time where tocharian speakers lived and buried their dead? I imagine we must be talking about the historic era and the late phases of Tocharian speakers.

My point was that the Tarim Basin mummies derived from a very early population that moved east, probably from what is now Ukraine (hence the R1a and C4 combo). Indeed, they were in the Tarim Basin earlier than the Andronovo people in South Siberia. So if they weren't the Proto-Tocharians, then they were even more distinct than the Proto-Tocharians, and yet, we know for a fact they belonged to R1a.

It might also be useful to remember that the Tarim Basin mummies were found in a burial complex that was isolated from other archeological sites. That's where they went to bury their dead, and didn't inhabit that area of the Tarim Basin. So they might well have come from a region that eventually became a well known Tocharian site, and I don't think pointy hats are enough to posit them as Indo-Iranians, considering the chronology, and their very different mtDNA structure to those of the Andronovo and Scythian nomads.



Population shifts? They happened in Europe, so why not in East Central Asia?

The two Chinese R1a are obviously of recent Turkic origin, and it's very difficult to say where their ancient Indo-European ancestors lived during the Bronze Age. But probably somewhere around the Altai.



http://mbe.oxfordjournals.org/content/early/2013/05/16/molbev.mst093.short

http://2.bp.blogspot.com/-WFcfoMpfMrM/UZZzVmicAlI/AAAAAAAAIz4/IWTnu_ONSyQ/s1600/tibet.jpg



You're confusing the origins of mtDNA C with the origins of the Tarim Basin mummies.

The Tarim Basin mummies didn't have to originate in the same place as the most basal mtDNA C clades tens of thousands of years before them. They could have come from a place where mtDNA C was common much later.

The Ukraine is a good bet, considering the presence of mtDNA C4 in ancient pre-Tarim Basin mummy remains there.

Generalissimo
12-03-2013, 01:25 AM
So, you seem to be distinguishing between the Tarim mummies and the Afanasievo in terms of origin because Afanasievo would normal be traced back to Repin at the Volga-Urals end of the western steppe.

I can definitely see a genetic difference between the Tarim Basin mummies and the Andronovo/Scythian nomads, even though both look like they originally came from west of the Urals. Their cultures also look very different.

Afanesievo will remain a mystery until we see some ancient DNA from their remains.

parasar
12-03-2013, 02:07 AM
Thanks for posting that. There are few options a movement in that direction from the archaeological record. By definition any movement into Asia by M73 had to happen after 5000BC. It is perhaps rather like some of the apparently early branchings of R1a which have been preserved in western Europe at very low levels far from the likely origin point. I looked with interest at the Moldova study that basically didnt find any except among the Ukrainian minority near the Ukraine border there.

I think that it was likely Alexander's (and perhaps even the earlier Trojan episode) movements that disturbed a number of tribes living in Anatolia, Mesopotamia, and the Caucasus. Many just crossed over into the wastelands (scythia) rather than surrender. Alexander did pursue some of them but was for the most part satisfied in shutting them out. Persians had connections to India and they escaped eastward.

I think you are spot on about M73 being akin to early separating branches of R1a, though I expect a more southern distribution for M73.

Some of VV's trees and diagrams are here: http://vizachero.com/R1b1/
http://vizachero.com/R1b1/R-Map.png

alan
12-03-2013, 03:50 AM
I think all the R1b from P297 and above is very dubious on that map. I would put P297 more like the north side of the Caucasus with an arrow from there to M73 near the north Caspian, M269 pretty well at the north side of the Caucasus and L23 arrows going from there both west into the Balkans (and then from there to Anatolia) and south along the Caspian shore of the Caucasus to NW Iran (and from there into northern Mesopotamia). That at least is a pattern I can make archaeological and early historical sense of. The problem moving a copper age clade into Europe by Anatolia is that this does not appear to have happened. Most of the suggested flow was in the opposite direction in the copper and bronze ages. I will only revise that observation if the age of M269 is redated as pre-copper age.


I think that it was likely Alexander's (and perhaps even the earlier Trojan episode) movements that disturbed a number of tribes living in Anatolia, Mesopotamia, and the Caucasus. Many just crossed over into the wastelands (scythia) rather than surrender. Alexander did pursue some of them but was for the most part satisfied in shutting them out. Persians had connections to India and they escaped eastward.

I think you are spot on about M73 being akin to early separating branches of R1a, though I expect a more southern distribution for M73.

Some of VV's trees and diagrams are here: http://vizachero.com/R1b1/
http://vizachero.com/R1b1/R-Map.png

alan
12-03-2013, 04:04 AM
One thing worth noting from Anthony's book is that he noted very strong influence of late Maykop on the Repin culture of the Don-Volga area c. 3500BC. At the same period Maykop was also appearing in NW Iran. So, there is no doubt in my mind that at least some genes of Maykop people were around on both the steppe to the north and east and NW Iran to the south-east c. 3500BC. It still remains tempting to see a link with late Maykop and L23 which is exactly the right age. However, it is also tempting to link it or M269 with Sredny Stog. I know there were hybrid cultures of the two. Pre-Maykop was apparently in contact with and got its metal supply from Sredny Stog related groups and the changes we see as Maykop arose probably mainly were down to looking to NW Iran for copper when the Balkans supply that the Srendny Stog groups had controlled started to fail in the centuries after 4000BC. In fact that made the people of the Maykop zone very rich by being an important link in a new metal supply chain.

Humanist
12-03-2013, 04:20 AM
I think that it was likely Alexander's (and perhaps even the earlier Trojan episode) movements that disturbed a number of tribes living in Anatolia, Mesopotamia, and the Caucasus. Many just crossed over into the wastelands (scythia) rather than surrender. Alexander did pursue some of them but was for the most part satisfied in shutting them out. Persians had connections to India and they escaped eastward.

What Mesopotamian tribes are you referring to? Alexander was welcomed as a liberator by Mesopotamians, when he entered Babylon. There is nothing in the cuneiform record (Sumero-Akkadian) suggesting that Alexander caused great havoc among the native populace.

parasar
12-03-2013, 05:10 AM
What Mesopotamian tribes are you referring to? Alexander was welcomed as a liberator by Mesopotamians, when he entered Babylon. There is nothing in the cuneiform record (Sumero-Akkadian) suggesting that Alexander caused great havoc among the native populace.

I had read the following account:
“After Alexander the King (had) conquered the descendants of the children of Lot and dispelled them into the land Ḳedar ... Alexander conquered all the edges of the land. He started from the west, and went south, entered northwards, transgressed the Caucasus (mountains) and came to Kartli ... And when he saw these fierce pagan tribes, whom we call Bun-Turks and Qypchaqs, who resided along the river Kur, Alexander was astonished, for no (other) tribes would do the (same) ... ."

"martial tribes, an offspring of the Chaldees, Hons, and they asked the ruler of the Bun-Turks for tributed land. And they settled in Zanavi. ... And after some time, Alexander, the king of all the land, came (again) and destroyed these three cities and fortresses and defeated the Hons with the sword.”

"After this, Alexander gained strength and conquered all the land, and he came to the land of Kartli. And he found these strong fortified cities in Inner-Kartli: ... Urbnisi, Ḳasṗi ... Saṛkine, and Zanavi, the quarter of the Jews ...”
http://titus.uni-frankfurt.de/personal/jg/pdf/jg2007c.pdf

The Gates of Alexander to keep out the Jews of Zanavi/Huns are also mentioned. I think they are the same as Darbent (Sanskrit equivalent would be dwar bandh or door closed).
http://en.wikipedia.org/wiki/Gates_of_Alexander

Humanist
12-03-2013, 05:41 AM
I had read the following account:
“After Alexander the King (had) conquered the descendants of the children of Lot and dispelled them into the land Ḳedar ... Alexander conquered all the edges of the land. He started from the west, and went south, entered northwards, transgressed the Caucasus (mountains) and came to Kartli ... And when he saw these fierce pagan tribes, whom we call Bun-Turks and Qypchaqs, who resided along the river Kur, Alexander was astonished, for no (other) tribes would do the (same) ... ."

"martial tribes, an offspring of the Chaldees, Hons, and they asked the ruler of the Bun-Turks for tributed land. And they settled in Zanavi. ... And after some time, Alexander, the king of all the land, came (again) and destroyed these three cities and fortresses and defeated the Hons with the sword.”

"After this, Alexander gained strength and conquered all the land, and he came to the land of Kartli. And he found these strong fortified cities in Inner-Kartli: ... Urbnisi, Ḳasṗi ... Saṛkine, and Zanavi, the quarter of the Jews ...”
http://titus.uni-frankfurt.de/personal/jg/pdf/jg2007c.pdf

The Gates of Alexander to keep out the Jews of Zanavi/Huns are also mentioned. I think they are the same as Darbent (Sanskrit equivalent would be dwar bandh or door closed).
http://en.wikipedia.org/wiki/Gates_of_Alexander

Livius.org


The consequences of the battle of Gaugamela can hardly be overestimated. The great Persian army was no longer there to protect Babylonia, the wealthiest satrapy of the empire, or its capital Babylon, one of the largest cities in the world. Its population was not expected to resist the Macedonians; the Babylonians had revolted as recently as 336/335 under Nidin-Bêl.

The Macedonians looted the Persian camp, plundered Arbela, advanced quickly to the south and on 18 October, they were at Sippar, where Alexander announced that he would spare the houses of the Babylonians. After this declaration, the Persian commander Mazaeus, who had gone from Gaugamela to Babylon, formally surrendered the city and on the twenty-second, Alexander's army entered the city through the famous Ištar Gate and the Procession Street, the victorious king riding in the royal chariot. An eyewitness account survives in the history of Quintus Curtius Rufus.

The Babylonians recognized the new ruler as 'king of the world', probably a translation of the Greek title Alexander had adopted after the battle near Issus, 'king of Asia'. Alexander had promised that his men would not enter the houses of the ancient city, but this did not mean that the Babylonian women were safe. A century before, the Greek researcher Herodotus of Halicarnassus had written that in Babylon, the women had to serve the goddess of love as prostitutes. This was not true, but Alexander's men believed that the custom existed and behaved accordingly. In a brief digression on Babylonian customs, Curtius Rufus describes how, in his view, the women all behaved like whores - it is easy to see that their behavior was far from voluntary.


Hollywood's take on Alexander's Babylonian entrance:


http://www.youtube.com/watch?v=VsnVi1Gid_I

Michał
12-03-2013, 12:28 PM
My point was that the Tarim Basin mummies derived from a very early population that moved east, probably from what is now Ukraine (hence the R1a and C4 combo). Indeed, they were in the Tarim Basin earlier than the Andronovo people in South Siberia. So if they weren't the Proto-Tocharians, then they were even more distinct than the Proto-Tocharians, and yet, we know for a fact they belonged to R1a.
I agree that it would make some sense but only if those very ancient Afanasevo and/or Tarim Basin people represented some early separated lineage(s) of R1a (or at least some very early separated sublineage of Z93), which is not supported by any data, so far. Let’s hope that some ancient SNP-tested Afanasevo or Tarim basin samples will soon be available, so we can compare it with our predictions.

Also, the above scenario is based on the assumption that the Late PIE dialect should be dated to about 2500-3000 BC (or to 2500 BC , as suggested by Anthony), which I personally consider highly unlikely. Such late dates for Late PIE would imply that the Corded Ware people became IE speakers only after they have contacted some Late Yamna people in the Eastern Carpathian region (as proposed by Anthony). Frankly speaking, it is very hard to imagine that such relatively late local contacts at the upper Dniester would have resulted in a sudden spread of the Late PIE dialect to some previously expanded Corded Ware populations in different regions of Central, Western and Eastern Europe (not to mention Scandinavia). Also, this scenario would imply that we should expect to find more traces of the Pre-Anatolian people in the Balkans and in the Middle Danubian region than in Anatolia, and this is because the early spread of the Kurgan people that started at about 4000 BC (Suvorovo, Cernavoda, etc.) was directed mostly west (including Cotofani, Baden and Vucedol) rather than south (Ezero, Troy) or south-east to Anatolia (not much evidences for that).

Based on the available data, I would date the branching points for the Late PIE and Early PIE to about 4500-4000 BC and 6000-5500 BC, respectively. This is of course much closer to the very early dates suggested by Gray and Atkinson, and although I don’t agree with their Out-of-Anatolia model, I find their age estimates more likely than those suggested by Anthony.



It might also be useful to remember that the Tarim Basin mummies were found in a burial complex that was isolated from other archeological sites. That's where they went to bury their dead, and didn't inhabit that area of the Tarim Basin. So they might well have come from a region that eventually became a well known Tocharian site, and I don't think pointy hats are enough to posit them as Indo-Iranians, considering the chronology, and their very different mtDNA structure to those of the Andronovo and Scythian nomads.
I think we should also consider an alternative scenario in which the frequently suggested time-frame for Proto-Indo-Iranian and its strong association with the Andronovo culture would be revised a bit. Assuming that I am right about the much earlier dates for the Late PIE (4500-4000 BC, as suggested above), it seems possible that Afansevo and Andronovo could have represented just two early separated branches of Indo-Iranians (roughly corresponding to Iranian and Indo-Aryan, respectively), which would place the Proto-Indo-Iranians in the Eastern Yamnaya cultural horizon at about 3500-3000 BC. It is worth noting that Mallory dates the common ancestor of the Iranian and Indo-Aryan languages to about 2500 BC, while Gray and Atkinson suggest a nearly identical date (2600 BC). This is of course significantly later than the 3500-3000 time frame suggested in my scenario, but still significantly earlier than the final phase of Andronovo, a cultural horizon dated to 1900-1500 BC only. The fact that the linguistics-based dates are somehow younger than my own estimates could be quite easily explained by the much later intensive contacts between the late Andronovo and post-Afanasevo populations, which is well documented by some occasional Andronovo-related findings in the East (including China), and some apparent population movements of the Eastern “true nomads” towards Central Asia and the Southern Ural region in the Late Bronze-Age period (roughly 1500-1000 BC).



Population shifts? They happened in Europe, so why not in East Central Asia?
I don’t think we have any archaeological data that would support such massive population replacement associated with the movement of the Andronovo-derived people to China. There are some evidences of Andronovo artefacts in China, but nothing that would suggest a massive movement (please anybody correct me if I am wrong about it). On the other hand, there are some data suggesting a significant population movement in the opposite direction, as mentioned above.



The two Chinese R1a are obviously of recent Turkic origin, and it's very difficult to say where their ancient Indo-European ancestors lived during the Bronze Age. But probably somewhere around the Altai.
Agreed. However, since the Altai region was strongly associated with the suggested ancestors of those putative Tocharians from Xiaohe, i.e. with Afanasevo, this doesn’t change that much. Most of the Turkic R1a is likely to be derived from the ancient Afanasevo and post-Afanasevo people from the Southern Siberia, Dzungaria and Tarim regions.



http://mbe.oxfordjournals.org/content/early/2013/05/16/molbev.mst093.short

Thanks! So this was actually R1a(xM17) and not M17*. While M417* (or even M17*, though with some objections) would make the potential association with the Tocharian speakers quite likely, I doubt this can be said about such early separated lineage (R1a* or R1a1*), as this would be rather more consistent with some early R1a* people from Central Asia or Siberia who were not included into the Paleolithic (or Early Mesolithic) group migrating to Eastern Europe and giving rise to the major M417 clade of R1a.

BTW, it is worth noting that there are actually more examples of Tibetan R1b-M73 and R1b(xM73) than R1a(xM17).

Michał
12-03-2013, 12:32 PM
I had read the following account:
“Alexander conquered all the edges of the land. He started from the west, and went south, entered northwards, transgressed the Caucasus (mountains) and came to Kartli ... And when he saw these fierce pagan tribes, whom we call Bun-Turks and Qypchaqs, who resided along the river Kur, Alexander was astonished, for no (other) tribes would do the (same) ... ."
This is not a contemporary relation but something written much more recently by someone who was not fully aware of the very ancient history of that region. There is no way Alexander could have met any Turkic tribes (like Kipchaks) in the Caucasus region, since the Alexander’s military conquest predated the Turkic expansion from the Altai region by at least several centuries.

This reminds me of a Polish Medieval chronicler Wincenty Kadłubek who in his “Chronica Polonurum” described how Alexander the Great and Julius Caesar played an important role in the history of the ancient Poles… :)

Ral
12-03-2013, 12:39 PM
I think you divide such concepts as " linguistic ancestors " of European nations and the " physical ancestors " ( their own ). I believe that the concept of " R1a - Indo-European subclade " based on the assumption that such remote ethnic groups like Europeans and Indians should have common haplogroup . And this really exists-r1a. However, ignoring the fact that many European nations have haplogroup r1a negligible amount. For example, the Ossetians have r1a ,as I know, less than 1%.
This means that the original haplogroup of speakers can fade over time, or language may be submitted without physical contact. And both of these statements are not in favor of R1a as Indo-European haplogroup.
Especially evident in the altaic-speakers.
What would you any subclade of haplogroup not taken as the starting point for the Altaic, there a significant amount of this haplogroup altaic ethnicities do not have, or have a negligible percentage, which also can be explained by later migrations.
I think it is time to consider alternative concepts, such as I haplogroup is pre-indoeuropian.
This immediately solve linguistic inconsistencies (Satem-centum isoglosses).
And I think that research ADNA only clarify the routes of ancient migrations R1a carriers. But the concept of "Proto-Indoeuropean R1a" will stay in conflict with the linguistics.

Michał
12-03-2013, 12:43 PM
I think we should also consider an alternative scenario in which the frequently suggested time-frame for Proto-Indo-Iranian and its strong association with the Andronovo culture would be revised a bit. Assuming that I am right about the much earlier dates for the Late PIE (4500-4000 BC, as suggested above), it seems possible that Afansevo and Andronovo could have represented just two early separated branches of Indo-Iranians (roughly corresponding to Iranian and Indo-Aryan, respectively), which would place the Proto-Indo-Iranians in the Eastern Yamnaya cultural horizon at about 3500-3000 BC. It is worth noting that Mallory dates the common ancestor of the Iranian and Indo-Aryan languages to about 2500 BC, while Gray and Atkinson suggest a nearly identical date (2600 BC). This is of course significantly later than the 3500-3000 time frame suggested in my scenario, but still significantly earlier than the final phase of Andronovo, a cultural horizon dated to 1900-1500 BC only. The fact that the linguistics-based dates are somehow younger than my own estimates could be quite easily explained by the much later intensive contacts between the late Andronovo and post-Afanasevo populations, which is well documented by some occasional Andronovo-related findings in the East (including Southern Siberia China), and some apparent population movements of the Eastern “true nomads” towards Central Asia and the Southern Ural region in the Late Bronze-Age period (roughly 1500-1000 BC).

I should also mention that the very fact that the Mitanni-Indo-Aryan language is attested in West Asia at about 1500 BC makes it practically impossible to date the Proto-Indo-Iranians to the Late Adronovo phase.

Generalissimo
12-03-2013, 01:37 PM
BTW, it is worth noting that there are actually more examples of Tibetan R1b-M73 and R1b(xM73) than R1a(xM17).

Yes, I saw that, but I don't think it means much unless we have ancient DNA to put it into context.

My opinion is that there must be an important reason why R1b isn't showing up in any of the ancient steppe or Chinese remains, while R1a is. That reason, I think, is the fact that R1a moved along with the early Indo-Europeans in several waves from the Copper Age onwards into Asia, while R1b didn't.

In fact, I think R1b arrived on the steppe and near the Tarim Basin during and after the Silk Road. That's why it'll never show up in any Asian remains that can be classified as potentially early Indo-European. I'm also of the opinion that most of its major expansions in Europe were well after anything that can be considered the early Indo-European timeframe, and so the only prehistoric remains it'll show up in are those belonging to the Bell Beakers.

Generalissimo
12-03-2013, 01:43 PM
And I think that research ADNA only clarify the routes of ancient migrations R1a carriers. But the concept of "Proto-Indoeuropean R1a" will stay in conflict with the linguistics.

I don't see any conflict between R1a phylogeny, all the R1a found in ancient Kurgans to date, and linguistics.

http://img845.imageshack.us/img845/1748/r1a1clades.jpg

http://img268.imageshack.us/img268/2769/hxzk.png

newtoboard
12-03-2013, 02:09 PM
I think that it was likely Alexander's (and perhaps even the earlier Trojan episode) movements that disturbed a number of tribes living in Anatolia, Mesopotamia, and the Caucasus. Many just crossed over into the wastelands (scythia) rather than surrender. Alexander did pursue some of them but was for the most part satisfied in shutting them out. Persians had connections to India and they escaped eastward.

I think you are spot on about M73 being akin to early separating branches of R1a, though I expect a more southern distribution for M73.

Some of VV's trees and diagrams are here: http://vizachero.com/R1b1/
http://vizachero.com/R1b1/R-Map.png

Why is that?

newtoboard
12-03-2013, 02:15 PM
I think we should also consider an alternative scenario in which the frequently suggested time-frame for Proto-Indo-Iranian and its strong association with the Andronovo culture would be revised a bit. Assuming that I am right about the much earlier dates for the Late PIE (4500-4000 BC, as suggested above), it seems possible that Afansevo and Andronovo could have represented just two early separated branches of Indo-Iranians (roughly corresponding to Iranian and Indo-Aryan, respectively), which would place the Proto-Indo-Iranians in the Eastern Yamnaya cultural horizon at about 3500-3000 BC. I




Do you really believe in these fringe views? So Andronovo corresponds to Proto Indo-Aryan yet all its descendant cultures in Central Asia, South Siberia and Europe (Timber Grave) magically end up speaking Iranian? And we are also supposed to also believe there were migrations from Afanasevo into West Asia that are completely undocumented? And somehow we magically have Iranian speaking cultures like Yaz surrounded by an Indo-Aryan culture? I have never been convinced the split occurred in Europe either. Somewhere in Uzbekistan sounds more likely to me.

newtoboard
12-03-2013, 02:36 PM
I have already shown in my post that R1 lines appear to have spread as elite lineages on their conquests. They do not appear to have frequently picked up the substrate. You yourself often have posted about mono-lineage populations. That could only have happened in the copper age if these elites basically ruled who they conquered and only rarely encorporated the conquered substrate into their subsequent expansions. It certainly worked that way in Europe for R1b and the Kurgan evidence seems to suggest the R1a military elites did not hugely absorbed their conquered populations into the elite back in the copper and Bronze Age. Citing a couple of exceptions doesnt much change that basic pattern.

The elites probably made clients or otherwise extracted tribute from who they conquered but as a patrilineal society they probably rarely absorbed them. Its even less likely if they were perceived as enemies. So, as far as I am concerned who a particular language speaking elite passed through doesnt give any consistent expectations as to what they should have been absorbed and passed on in their further travels. Its not a safe approach to make inferences from.

I2 and J2 among Slavs and Indo-Aryan groups disagrees that they did pick up the substrate. I have only argued about monoclade populations on the steppe. It might be a couple of exceptions but there aren't many overall samples (the one Andronovo C might be one samples but only 10 were tested so it is not insignificant imo).

Michał
12-03-2013, 03:05 PM
& newtoboard,
I know you are able to logically analyze all available data, and though I frequently disagree with you about some details, I also have much respect for you because I know you are able to change your mind after confronted with some convincing argumentation. However, it is a bit annoying when you attack your “opponent” with some unprepared (and illogical) arguments instead of trying to reevaluate all the data in the light of your opponent’s view.


Do you really believe in these fringe views? So Andronovo corresponds to Proto Indo-Aryan yet all its descendant cultures in Central Asia, South Siberia and Europe (Timber Grave) magically end up speaking Iranian?
Srubna (Timber Grave) does not descend from Andronovo! These were two roughly contemporary cultures, both likely derived from the Late Yamna horizon. Also, how on Earth can you know that Srubna people spoke Iranian?



And we are also supposed to also believe there were migrations from Afanasevo into West Asia that are completely undocumented?
Where did I mention any movement of Afanasevo westward? According to my scenario, that initial westward expansion was supposed to take place in the 1500-1000 BC period (when Afanasevo was long gone) and I haven’t even suggested that it reached West Asia that early.



And somehow we magically have Iranian speaking cultures like Yaz surrounded by an Indo-Aryan culture?
Does the Andronovo=Proto-Indo-Iranian hypothesis explain this better than my alternative scenario?



I have never been convinced the split occurred in Europe either. Somewhere in Uzbekistan sounds more likely to me.
When did it take place in your opinion and which particular Andronovo-descending cultures would you associate with the Proto-Iranians and Proto-Indo-Aryans, respectively?

alan
12-03-2013, 03:09 PM
One last thing struck me again last night and I had a think about it before rejecting it. The German paper about Maykop makes a very strong case for strong similarities in material culture between Maykop, northern Iran and central Asia at about the latitude of northern Iran. That is kind of interesting and I have previously overlooked the central Asia aspect of this because Iran was more immediate to Maykop. The paper being in German and using google translator is hard to follow so I am not sure of the details. Obviously its not clear how much gene flow was involved and how much emulation by natives was involved. However, it does create an interesting linkage. However, I have got to be honest I doubt it has much to do with R1b. A move from south central Asia to the north Caucasus via north Iran doesnt work for M73 because its absent in Iran and the south Caucasus and rare even in the north Caucasus. For M269/L23 its not quite so bad in the sense that there is a reasonable amount in the north Caucasus, the Armenian area and NW Iran. However, its very hard to see a gene flow commencing at the south central Asia end given the Hindu Kush results. There could be a case for M269 as a middleman to this network being located in NW Iran. In general, I briefly got interested but it seems to fall down on most counts.

I think now I have never been so convinced that only ancient DNA across a broad spectrum of Neoltithic and copper age cultures is going to clarify the R1b story in that period. I am convinced the story belong somewhere on any side of the circum pontic caspian area (minus early farming areas) but that is terrible vague including as it does the western steppe, the south Urals, the Bakans, NW Anatolia, the Caucasus and west central Asia. I think its pretty well a brick wall trying to tease this story out without a lot more ancient DNA. There are dozens if not hundreds of cultures in that zone c. 6000-3000BC and you dont want just one or two samples to represent whole populations so its going to potentially take 100s of ancient y DNA samples to sort this.

newtoboard
12-03-2013, 03:19 PM
& newtoboard,
I know you are able to logically analyze all available data, and though I frequently disagree with you about some details, I also have much respect for you because I know you are able to change your mind after confronted with some convincing argumentation. However, it is a bit annoying when you attack your “opponent” with some unprepared (and illogical) arguments instead of trying to reevaluate all the data in the light of your opponent’s view.

My apologies.



Srubna (Timber Grave) does not descend from Andronovo! These were two roughly contemporary cultures, both likely derived from the Late Yamna horizon. Also, how on Earth can you know that Srubna people spoke Iranian?

I should have phrased that better. The two are very connected. And Timber Grave is usually assumed to be the ancestor of European Scythians or Cimmerians (who might be Iranian speakers but are almost certainly not Indo-Aryan speakers). So it is hard to picture An Indo-Aryan culture in between two cultures which ultimately both end up giving rise to NE Iranian speakers (if Afanasevo leads to the steppe Saka or Scythians as I assume it does in your scenario)


Where did I mention any movement of Afanasevo westward? According to my scenario, that initial westward expansion was supposed to take place in the 1500-1000 BC period (when Afanasevo was long gone) and I haven’t even suggested that it reached West Asia that early.

You would still need a movement from the former territory of this Proto Iranian culture (Afanasevo) or from the Tarim to go through Andronovo to get to the final destinations of the Persians and Medes.



Does the Andronovo=Proto-Indo-Iranian hypothesis explain this better than my alternative scenario?

Yes. I would say most of this horizon ended up speaking Iranian with Indo-Aryans in the SE. This would explain how the Yaz cultures and other Central Asian cultures ended up speaking Iranian.




When did it take place in your opinion and which particular Andronovo-descending cultures would you associate with the Proto-Iranians and Proto-Indo-Aryans, respectively?

It took place when one group (which I think was still speaking Indo-Dardic-Nuristani) ended up absorbing elements of the BMAC. That is why I think a split in Uzbekistan is more likely than one on the steppe. They migrated SE and are represented by the SWAT and Gandhara cultures.

parasar
12-03-2013, 03:24 PM
This is not a contemporary relation but something written much more recently by someone who was not fully aware of the very ancient history of that region. There is no way Alexander could have met any Turkic tribes (like Kipchaks) in the Caucasus region, since the Alexander’s military conquest predated the Turkic expansion from the Altai region by at least several centuries.

This reminds me of a Polish Medieval chronicler Wincenty Kadłubek who in his “Chronica Polonurum” described how Alexander the Great and Julius Caesar played an important role in the history of the ancient Poles… :)

That is the reason I had just mentioned "wastelands (scythia)" in my initial post as the names of the barbarian characters changes but the story remained pretty consistent.

The Turks and Huns are mentioned in the Conversion of Kartli

The Conversion of Kartli (Georgian: მოქცევაჲ ქართლისაჲ, transliterated as mok’ts’evay k’art’lisay) is the earliest surviving medieval Georgian historical compendium, independent from The Georgian Chronicles, the major corpus historicum of medieval Georgia. Written in the 7th and 9th centuries, this chronicle follows the history of Kartli (a core Georgian region known to the Classical authors as Iberia; sometimes archaically referred to all of Georgia) from the earliest times to the 7th century, making a particular focus on Christianization of Georgians by Saint Nino early in the 4th century.[1] http://en.wikipedia.org/wiki/Conversion_of_Kartli_(chronicle)

Prior to that we had the Huns (Jerome 347–420 AD) and before that the Alans (Josephus 37–100 AD http://books.google.com/books?id=4tsKAQAAIAAJ&pg=PA151)

Besides the Caucasus one, Alexander built a number of other Iron Gates. There are a couple in Uzbekistan. One of them also has the name Darbend.
Some confirmation comes from this gate, as found near this Iron Gate was a ball - “the only weapon from Alexander’s army ever recovered” http://books.google.com/books?id=Im8Ujg_XAmIC&pg=PA83

alan
12-03-2013, 03:45 PM
This all raises the question of how R1b operated, predominantly expanding as they did into an already developed farming world yet somehow being able to penetrate almost everywhere at lightning speed. That is really mind boggling. R1b (well I really mean M269 and downstream) seems to have had a modus operandi of spreading, locally expanding at the expense of the locals and slowly diminishing them, in some cases to small minorities. That implies hogging of resources without encorporating others into the party on any significant scale. It almost makes we think they operated like a caste but more likely that just operated on very strictly patrilineal lines and access to resources was almost entirely by a persons lineage. Something along those lines was still practices by Gaelic clans in Medieval Ireland where the impression given is that conquest of new territories was done by a fission off the lineage and that lineage went on to breed out the locals. The small fission of a lineage could safely hold onto new lands in a sea of locals mainly by continuing support of the rest of the lineage they had branched off from. I have always had the impression that what we see in Medieval Gaelic Ireland looks like it was the system practiced by R1b in the prehistoric era too.

It is interesting to speculate about the origins of this system. It appears to have been a very unstable system, causing constant interand intra-lineage strife, it discouraged major investment in sophisticated settlements, was heavily pastoral with a much smaller agricultural unit, frequently practiced seasonal migration to upland pastures and really is was not suited to a stable farming world at all. It is a system that only really makes sense when a lineage can freely expand into new territories and the mobility that ephemeral settlements give makes it hard for a stronger enemy to actually conquer them, but it is poorly suited to life once that is no longer a possibility. It meant that if you were not the king or chief you would slip down into the ordinary farmer class within a couple of generations if you did not conquer new land. That created a motive for constant low level warfare as people sought to avoid downward mobility.

To me the system owes something to a steppe-like state of existence and mentality and is a huge contrast to the stable villages of farmers we see in part of SE Europe and SW Asia in the Neolithic (which was probably a much nicer society in many ways).


I2 and J2 among Slavs and Indo-Aryan groups disagrees that they did pick up the substrate. I have only argued about monoclade populations on the steppe. It might be a couple of exceptions but there aren't many overall samples (the one Andronovo C might be one samples but only 10 were tested so it is not insignificant imo).

nuadha
12-03-2013, 03:50 PM
I can understand how appealing the idea was that R1a = satem and R1b = centum. The idea swept the genetic genealogy forums years ago, but just like the early idea among linguists that satem = eastern IE and centum = western IE, it falls apart on Tocharian. People thought that Tocharian simply must be R1b. I did not. The centum-satem isogloss is chronological, not geographical, while the reverse is true of the R1a/R1b division.

The centum-satem isogloss was geographical at the early stages of IE dispersals. We have that the earlier branchings of IE - Anatolian, Tocharian, and Italo-Celtic - are all connected with r1b and one of them (italo-celtic) was totally dominated by r1b.

A while back I saw a good amount of neolithic influence (genetic) in CW which made me doubt that CW would resemble PIE peoples. Now I see that even David Anthony derives the IE in Corded Ware, and possibly pre Balto-Slavic too, from Cucuteni-Trypillian mixed with some yamnaya while he derives pre italo-celtic directly from yamnaya people.

Additionally, silja dillenberger, has already stated that there was a big genetic shift in the PC steppe from the yamnaya and eneolithic peoples to the Catacomb peoples. The yamnaya are seen as being closer to modern Europeans than the catacomb people which is good at explaining how neolithic Europe became more like modern Europe.

In summary, we have r1b tied to the earliest IE branchings and whose dispersals are directly from the PIE peoples themselves. Then we have the later Satem branchings of IE tied to r1a, with at least pre Balto-slavic probably not leaving the steppes with PIE peoples, and after a huge genetic shift on the PC steppes by Catacomb we see a steppe that is dominated by r1a.

r1b (m269 and m73) in PIE and r1a (m417 or a bit lower) in eastern-central Europe to the PC steppes with catacomb looks like a very reasonable story.


The first movement east which could carry it is therefore the Sintashta culture, starting 2100 BC. There is nothing else moving east between the trek to Afanasevo c. 3300 BC and Sintashta.

There was enough time for Catacomb type r1a to reach the Tarim basin by the time of those important mummies.

Also, notice how those mummies lack mtdna U and are generally dominated by west Eurasian ydna and east Eurasian mtdna. It looks like an early, forefront, male migration that just arrived - ahead of their counterpart women.

newtoboard
12-03-2013, 03:51 PM
This all raises the question of how R1b operated, predominantly expanding as they did into an already developed farming world yet somehow being able to penetrate almost everywhere at lightning speed. That is really mind boggling. R1b (well I really mean M269 and downstream) seems to have had a modus operandi of spreading, locally expanding at the expense of the locals and slowly diminishing them, in some cases to small minorities. That implies hogging of resources without encorporating others into the party on any significant scale. It almost makes we think they operated like a caste but more likely that just operated on very strictly patrilineal lines and access to resources was almost entirely by a persons lineage. Something along those lines was still practices by Gaelic clans in Medieval Ireland where the impression given is that conquest of new territories was done by a fission off the lineage and that lineage went on to breed out the locals. The small fission of a lineage could safely hold onto new lands in a sea of locals mainly by continuing support of the rest of the lineage they had branched off from. I have always had the impression that what we see in Medieval Gaelic Ireland looks like it was the system practiced by R1b in the prehistoric era too.

It is interesting to speculate about the origins of this system. It appears to have been a very unstable system, causing constant interand intra-lineage strife, it discouraged major investment in sophisticated settlements, was heavily pastoral with a much smaller agricultural unit, frequently practiced seasonal migration to upland pastures and really is was not suited to a stable farming world at all. It is a system that only really makes sense when a lineage can freely expand into new territories and the mobility that ephemeral settlements give makes it hard for a stronger enemy to actually conquer them, but it is poorly suited to life once that is no longer a possibility. It meant that if you were not the king or chief you would slip down into the ordinary farmer class within a couple of generations if you did not conquer new land. That created a motive for constant low level warfare as people sought to avoid downward mobility.

To me the system owes something to a steppe-like state of existence and mentality and is a huge contrast to the stable villages of farmers we see in part of SE Europe and SW Asia in the Neolithic (which was probably a much nicer society in many ways).

Where did this happen besides NW Europe? Not in West Asia, Southern Europe, the Balkans, Eastern Europe or Northern Europe (Scandavia +Germany where plenty of R1a, I1 and I2b still exists).

newtoboard
12-03-2013, 03:57 PM
The centum-satem isogloss was geographical at the early stages of IE dispersals. We have that the earlier branchings of IE - Anatolian, Tocharian, and Italo-Celtic - are all connected with r1b and one of them (italo-celtic) was totally dominated by r1b.

A while back I saw a good amount of neolithic influence (genetic) in CW which made me doubt that CW would resemble PIE peoples. Now I see that even David Anthony derives the IE in Corded Ware, and possibly pre Balto-Slavic too, from Cucuteni-Trypillian mixed with some yamnaya while he derives per italo-celtic directly from yamnaya people.

Additionally, silja dillenberger, has already stated that there was a big genetic shift in the PC steppe from the yamnaya and eneolithic peoples to the Catacomb peoples. The yamnaya are seen as being closer to modern Europeans than the catacomb people which is good at explaining how neolithic Europe became more like modern Europe.

In summary, we have r1b tied to the earliest IE branchings and whose dispersals are directly from the PIE peoples themselves. Then we have the later Satem branchings of IE tied to r1a, with at least pre Balto-slavic probably not leaving the steppes with PIE peoples, and after a huge genetic shift on the PC steppes by Catacomb we see a steppe that is dominated by r1a.

r1b (m269 and m73) in PIE and r1a (m417 or a bit lower) in eastern-central Europe to the PC steppes with catacomb looks like a very reasonable story.


There was enough time for Catacomb type r1a to reach the Tarim basin by the time of those important mummies.

Also, notice how those mummies lack mtdna U and are generally dominated by west Eurasian ydna and east Eurasian mtdna. It looks like an early, forefront, male migration that just arrived - ahead of their counterpart women.

There is nothing indicating R1b Tocharians is a fact. Nor is there any Yamnaya DNA out yet so R1b in PIE is far from confirmed or even a plausable theory at his point. There is also no evidence the shift was an R1b to R1a ones. This seems to be repeated as if Silja Dillenberger actually said there was an R1b to R1a shift.

Those mummies have mtdna H. They also had C and only C. Which was found in Neolithic Ukraine anyways.

newtoboard
12-03-2013, 04:00 PM
And trying to derive Indo-Iranian out of anything but Yamnaya is a lost cause.

Humanist
12-03-2013, 04:06 PM
I think that it was likely Alexander's (and perhaps even the earlier Trojan episode) movements that disturbed a number of tribes living in Anatolia, Mesopotamia, and the Caucasus. Many just crossed over into the wastelands (scythia) rather than surrender. Alexander did pursue some of them but was for the most part satisfied in shutting them out. Persians had connections to India and they escaped eastward.


What Mesopotamian tribes are you referring to? Alexander was welcomed as a liberator by Mesopotamians, when he entered Babylon. There is nothing in the cuneiform record (Sumero-Akkadian) suggesting that Alexander caused great havoc among the native populace.

parasar. Perhaps this past post of mine is somewhat related to what you are speaking of?


I did a search for "Germanicia," and I eventually came across this bit. It appears to be linked to what is called "British Israelism."

Even though I believe that the connections between the Near East and Europe have been underestimated, particularly in recent decades, much of what is contained below strikes me as a bit too fanciful. However, the writings of Pliny the Elder are interesting, I must admit.


The Assyrian Empire developed from the city-state of Asur (named for Asshur, a son of Shem—one of Noah's three sons—see Genesis 10:1, 22). Asshur was a brother of Arphaxad—an ancestor of Abraham, who was the father of the Hebrews (Genesis 11:10–26). Thus, true Assyrians and the descendants of Abraham (the Israelites) are kindred peoples. The name Assur means "leader" or "successful." Josephus, writing in the first century ad, writes that the Assyrians "became the most fortunate of nations, beyond others" (Antiquities of the Jews, 1:6:6). In light of their abilities and contributions to Western civilization, this is also true for the Germans. Assur was worshiped as the chief god of Assyria—the god of war—and was portrayed as a solar deity with a winged disc. The Hittites also used both the winged disc and the swastika. The swastika is a symbol for the sun, power, energy, Thor's hammer and the god of weather and storms. An ancient swastika has been found on a limestone slab in front of a temple of Assur (In Search of… The Origin of Nations, White, p. 311).

The Hittites and Assyrians also used a double-headed eagle to symbolize the sky gods—storm, thunder and the sun. These symbols reappear in the culture of Germany, Prussia and Austria, and especially the Third Reich. The Hittites (whom Assyria eventually conquered and absorbed) show linguistic and cultural links with two of the German tribes—Hessians and Prussians. Even more interesting, as historian Josef Bihl noted, are legends that Germany's oldest city, Trier, was founded in about 2000bc by Trebeta, the son of an Assyrian king named Ninus (In Deutschen Landen, p. 69). Visitors can still read an inscription on a historic house in Trier's marketplace, stating that this Assyrian colony was founded 1,300 years before Rome.

Some historians have described the Assyrians as the "Prussians of the Ancient World" (The Ideal and Destiny, McCulloch, p. 224), submissive to centralized authority, with a "deep rooted feeling of superiority"—the idea of being a "master race" (Mass Deportations and Deportees in the Neo-Assyrian Empire, Obed, p. 89). The Assyrians were extremely nationalistic, with "a strong sense of participating in a common and native way of life" (ibid., p. 66)—similar to the German idea of a volk or a people. After the fall of the Assyrian Empire, the Roman historian Pliny mentioned a tribe of the "Assyriani" among the Scythian peoples in the Crimea north of the Black Sea (Natural History, Bk IV. XII. 81). The historian Jerome, writing in the 4th century ad, said that the "descendents of Assur" were among the Celto-Scythian-Hun hordes then invading Europe (Nicene and Post-Nicene Fathers, Jerome Letter 123, section 16). Researcher Leon Poliakov notes the ancient Bavarian account that the people of Bavaria came into central Europe from the region of Armenia by the Black Sea (The Aryan Myth, p. 76). Considering this information, it is not surprising to find medieval Arab writers describing the Germans as "Assyrians" (Israelites und Hyksos, Germol, pp. 89–90). The links between Germany and Assyria can be found, and are neither far-fetched nor imagined.

Douglas S. Winnail

Of course, they may not have a solid basis in fact. But, I thought it was interesting enough to share, given the discussion.

Note "Assyrii" and "Chalybes."

Wikipedia

http://i1096.photobucket.com/albums/g326/dok101/Caucasus_.jpg




The Assyrians sometimes moved and were in different areas. Around 530 B.C.:

Scylax of Caryanda names the coast of the Black Sea, from the Chalybians to Armene, westward of the promontory of Syrias, Assyria. Strabo states that these Syrians, who extended from the Taurus northwards as far as the Pontus; were named Leuko-Syrians, i.e. white Syrians. (Dunker M. The history of antiquity, Volume 1. Evelyn Abbott Publisher R. Bentley & son, 1877. Original from the University of Michigan, Digitized Sep 12, 2007, p. 540)

Wikipedia


Chalybes

The Chalybes or Chaldoi (Χάλυβες, Χάλυβοι, Χάλδοι. Georgian ხალიბები) were a Georgian tribe of Classical Antiquity who inhabited the area of Pontus and Cappadocia, located in modern Turkey.

They lived in the northern area of Anatolia known as Chaldia, near the shores of the Black Sea, from the Halys to Pharnakeia and Trabzon in the east and as far south as eastern Asia Minor.

The main sources for the history of the Chaldoi are accounts from classical authors, including Homer, Strabo, and Xenophon. In Roman times, the Chaldaei (homonymous but unrelated to the Semitic Chaldeans) and Chalybes are mentioned by Plutarch (Lucull. c. 14) as settled in Pontus and Cappadocia, or the Pontus Cappadocicus section of the Roman province of Pontus.

Silesian
12-03-2013, 04:09 PM
The centum-satem isogloss was geographical at the early stages of IE dispersals. We have that the earlier branchings of IE - Anatolian, Tocharian, and Italo-Celtic - are all connected with r1b and one of them (italo-celtic) was totally dominated by r1b.

I came to that conclusion a long time ago. Just one look at the placement of the branches on the language tree. It's no wonder there is no attempt to show a chronological placement other than R1b.

newtoboard
12-03-2013, 04:11 PM
I am curious on this Neolithic genetic influence in Corded Ware. Probably due to low resolutions of various subclades. There is almost certainly more Mesolithic ancestry in R1a European groups than R1b ones.

newtoboard
12-03-2013, 04:12 PM
I came to that conclusion a long time ago. Just one look at the placement of the branches on the language tree. It's no wonder there is no attempt to show a chronological placement other than R1b.

There is no evidence Satemization even existed at the time of those early dispersal much less there being a geographical cline.

Silesian
12-03-2013, 04:20 PM
There is no evidence Satemization even existed at the time of those early dispersal much less there being a geographical cline.
Nonsense, post your language branched tree complete from start to finish showing chronological R1a satem/centum excluding R1b. Everything is in your favor you have all the archeological answers and ancient samples R1a.

newtoboard
12-03-2013, 04:25 PM
Nonsense, post your language branched tree complete from start to finish showing chronological R1a satem/centum excluding R1b. Everything is in your favor you have all the archeological answers and ancient samples R1a.

Satemization is likely an areal feature. There are various degrees of Satemization. A language tree has nothing to do with that.

alan
12-03-2013, 04:28 PM
This is not a serious point I am making but its always puzzled me how the mummies with a lot of non-European DNA apparently looked European. That wouldnt be the normal outcome if that was constantly going on today. It might be evidence that the Asian mt DNA was absorbed long before the time of the mummies and although the mtDNA lineages were retained in-group and grew, the autosomal impact was diluted by this mixing event not being repeated in more recent times. That could be indirect evidence that the mummies were descended from people who had had a mixing event or phase in east Eurasia many generations before the mummies. Then again this is 'off the cuff' musing so dont all attack at once please!


The centum-satem isogloss was geographical at the early stages of IE dispersals. We have that the earlier branchings of IE - Anatolian, Tocharian, and Italo-Celtic - are all connected with r1b and one of them (italo-celtic) was totally dominated by r1b.

A while back I saw a good amount of neolithic influence (genetic) in CW which made me doubt that CW would resemble PIE peoples. Now I see that even David Anthony derives the IE in Corded Ware, and possibly pre Balto-Slavic too, from Cucuteni-Trypillian mixed with some yamnaya while he derives pre italo-celtic directly from yamnaya people.

Additionally, silja dillenberger, has already stated that there was a big genetic shift in the PC steppe from the yamnaya and eneolithic peoples to the Catacomb peoples. The yamnaya are seen as being closer to modern Europeans than the catacomb people which is good at explaining how neolithic Europe became more like modern Europe.

In summary, we have r1b tied to the earliest IE branchings and whose dispersals are directly from the PIE peoples themselves. Then we have the later Satem branchings of IE tied to r1a, with at least pre Balto-slavic probably not leaving the steppes with PIE peoples, and after a huge genetic shift on the PC steppes by Catacomb we see a steppe that is dominated by r1a.

r1b (m269 and m73) in PIE and r1a (m417 or a bit lower) in eastern-central Europe to the PC steppes with catacomb looks like a very reasonable story.



There was enough time for Catacomb type r1a to reach the Tarim basin by the time of those important mummies.

Also, notice how those mummies lack mtdna U and are generally dominated by west Eurasian ydna and east Eurasian mtdna. It looks like an early, forefront, male migration that just arrived - ahead of their counterpart women.

newtoboard
12-03-2013, 04:33 PM
This is not a serious point I am making but its always puzzled me how the mummies with a lot of non-European DNA apparently looked European. That wouldnt be the normal outcome if that was constantly going on today. It might be evidence that the Asian mt DNA was absorbed long before the time of the mummies and although the mtDNA lineages were retained in-group and grew, the autosomal impact was diluted by this mixing event not being repeated in more recent times. That could be indirect evidence that the mummies were descended from people who had had a mixing event or phase in east Eurasia many generations before the mummies. Then again this is 'off the cuff' musing so dont all attack at once please!

That Tarim group is the exception. Most migrations eats involved the movement of women and Bronze Age groups of Kazakhstan and South Siberia were very West Eurasian on the mtDNA side as well carrying a very varied mix of West Eurasian mtdna such as U1,U2e, U3, U4, U5, U7, K, HV, H, W and T lineages.

That could also be evidence for the mtDna C coming from Ukraine. I think it is strnage how mixed Andronovo groups carried a variety of East Eurasian lineages (including A, C and Z) while this mixed Tarim group only had mtdna C (much like Ukraine).

alan
12-03-2013, 04:48 PM
Certain non-R has been found in one of the two sites tested. Archaeologists generally have always had the feeling that Corded Ware is some sort of farmer-steppe hybrid. Fashions change and go to extremes but I think a great deal of people have always taken this middle ground. Same with beakers. A few influential people pushed the idea of no such thing as the beaker people but I know that most archaeologists (including the ones that taught me) remained in the camp that there must have been at least a limited movement mainly because its about the best evidence of any migration into the isles between the Neolithic and the centuries on either side of 0. Most people did not 'go all Francis Pryor' in the 70s-0s but unfortunately pushing novel extreme ideas like total anti-migationism is the way some people went to set themselves apart from the crowd or appear progressive and attract interest. There is a lot more common sense in the archaeological world than you would think from tracking the decades of wild swings in ideology among what were actually a tiny group of high profile academics.


I am curious on this Neolithic genetic influence in Corded Ware. Probably due to low resolutions of various subclades. There is almost certainly more Mesolithic ancestry in R1a European groups than R1b ones.

newtoboard
12-03-2013, 04:54 PM
Certain non-R has been found in one of the two sites tested. Archaeologists generally have always had the feeling that Corded Ware is some sort of farmer-steppe hybrid. Fashions change and go to extremes but I think a great deal of people have always taken this middle ground. Same with beakers. A few influential people pushed the idea of no such thing as the beaker people but I know that most archaeologists (including the ones that taught me) remained in the camp that there must have been at least a limited movement mainly because its about the best evidence of any migration into the isles between the Neolithic and the centuries on either side of 0. Most people did not 'go all Francis Pryor' in the 70s-0s but unfortunately pushing novel extreme ideas like total anti-migationism is the way some people went to set themselves apart from the crowd or appear progressive and attract interest. There is a lot more common sense in the archaeological world than you would think from tracking the decades of wild swings in ideology among what were actually a tiny group of high profile academics.

I agree with all this. I responded to naudha 's point that R1a couldn't have been among PIE speakers because Corded Ware had too much Neolithic influence )what that had to do with PIE makes no sense to me as Corded Ware is the fusion of steppe, local Mesolithic and Neolithic elements just like most parts of Europe were) which means he was alluding to the fact that R1b groups had more Mesolithic ancestry and that is not the case (for some reason this myth of NW Europeans being more Mesolithic like and being lighter pigmented than NE Europeans persists). He was referring to mtdna not y-dna I believe. Which ties back into your previous point about exotic mtdnas expanding within a population but not having much of an autosomal impact.

parasar
12-03-2013, 05:06 PM
parasar. Perhaps this past post of mine is somewhat related to what you are speaking of?
...


Possible. Bakhtnassar is often implicated as one who transported or removed peoples.

Then there is the story told in the Icelandic sagas of Turkland and Troy. By the time of the sagas were compiled to preserve them the story of the Turks may have reached Snorri Sturlson or maybe there indeed was some movement from Anatolia to the Steppes preserved in those sagas.


The Aesir are "Asia-men"; and they sprang from Troy. According to Heimskringla they dwelt in Asaland, east of the river Tanakvisl [Don], which divides Europe and Asia. The chief city of Asaland was Asgard, a great town of sacrifices, and here dwelt Odin, who owned possessions across the mountains in Turkland. At the time of the Roman invasion he migrated with his people and eventually settled in Sweden the Great [Russia] ... Where had Snorri discovered the Trojan origin of his race? The theory was not current in the North, for it does not appear until his time. http://books.google.com/books?pg=PA152

After all Virgil does call the residents of Troy - Teucri (Teukros, Tarku), thereby perhaps the confusion.
Teucri http://books.google.com/books?id=c29gaBdKtlcC&pg=PA200
Teukros, Tarku http://books.google.com/books?id=MWkKAAAAIAAJ&pg=PA156

Jean M
12-03-2013, 05:10 PM
The centum-satem isogloss was geographical at the early stages of IE dispersals. We have that the earlier branchings of IE - Anatolian, Tocharian, and Italo-Celtic - are all connected with r1b ....

Let me try to explain more clearly. The centum-satem isogloss is linguistic. We may be trying to link haplogroups to it, but the isogloss itself - the sound change - is linguistic. Now looking just at the linguistic evidence i.e. where we find written evidence of people speaking this or that language, Tocharian was east of the European steppe, while Italic and Celtic were (and remain) west of the European steppe. See map: https://en.wikipedia.org/wiki/Centum-satem_isogloss

That ruined the idea that Von Bradke had in 1890. He thought that the first branching of PIE was into an eastern Satem group, and a western Centum group. In short he saw the isogloss as a geographical split. We now realise that it is chronological:


c. 4200 BC - The earliest form of PIE had neither the centum or satem sound. It formed the basis for the Anatolian branch.
c. 3500 BC - The Centum sound had developed. Centumization removed the palatovelars from the language. It did not matter which direction people went from the steppe. They took with them this form of PIE. It formed the basis for the Tocharian, Celtic, Italic, Germanic and Greek branches.
c. 2500 BC - The Satem sound developed. This formed the basis for the Baltic, Slavic, Indic and Iranian branches. Again it did not matter in which direction people went. They took this sound.

Ral
12-03-2013, 05:29 PM
I don't see any conflict between R1a phylogeny, all the R1a found in ancient Kurgans to date, and linguistics.

http://img845.imageshack.us/img845/1748/r1a1clades.jpg


Where are (for example) Romanian,Greek,Albanian, Armenian branches?

newtoboard
12-03-2013, 05:30 PM
Let me try to explain more clearly. The centum-satem isogloss is linguistic. We may be trying to link haplogroups to it, but the isogloss itself - the sound change - is linguistic. Now looking just at the linguistic evidence i.e. where we find written evidence of people speaking this or that language, Tocharian was east of the European steppe, while Italic and Celtic were (and remain) west of the European steppe. See map: https://en.wikipedia.org/wiki/Centum-satem_isogloss

That ruined the idea that Von Bradke had in 1890. He thought that the first branching of PIE was into an eastern Satem group, and a western Centum group. In short he saw the isogloss as a geographical split. We now realise that it is chronological:


c. 4200 BC - The earliest form of PIE had neither the centum or satem sound. It formed the basis for the Anatolian branch.
c. 3500 BC - The Centum sound had developed. Centumization removed the palatovelars from the language. It did not matter which direction people went from the steppe. They took with them this form of PIE. It formed the basis for the Tocharian, Celtic, Italic, Germanic and Greek branches.
c. 2500 BC - The Satem sound developed. This formed the basis for the Baltic, Slavic, Indic and Iranian branches. Again it did not matter in which direction people went. They took this sound.


Thank you for this post. Too often people confuse Anatolian as Centum. And forget Centumization is a process as well. PIE was not a Centum language.

newtoboard
12-03-2013, 05:32 PM
Jean do you think Satemization is a shared process between all thoe languages or did it occur in an Indo-Iranian group who then transmitted it west. The full Satemization of Indo-Iranian vs the partial of Balto-Slavic supports the second imo.

Jean M
12-03-2013, 05:36 PM
c. 4200 BC - The earliest form of PIE had neither the centum or satem sound. It formed the basis for the Anatolian branch.
c. 3500 BC - The Centum sound had developed. It formed the basis for the Tocharian, Celtic, Italic, Germanic and Greek branches.
c. 2500 BC - The Satem sound developed. This formed the basis for the Baltic, Slavic, Indic and Iranian branches.


OK. Now with these facts before us, it absolutely leaps to the eye that


R1a is a strong signature among the descendants of the c. 2500 BC PIE rump, taken overall.
R1b is predominant among at least the Celtic and Italic group of departures prior to Satemisation. Germanic is mixed R1a and R1b. Greek and Armenian speakers have R1b-L23 in their mixture of haplogroups. And because no R1a mummies have been found with tattoos in Tocharian, people feel that they can still argue the toss over Tocharian-speakers.


Several positions can be taken.


Michal has chosen to see R1b as the genetic signature of the Centum speakers and therefore the earliest speakers of PIE.
My position is that R1a and R1b were closely associated on the steppe and intermarrying, but because of patrilocality, they tended to cluster in groups strong in either R1b or R1a. R1a was at the east end of the steppe, while R1b was more west and centre. So where people left from would explain their haplogroup mix, while the time they left dictated the dialect they carried. I therefore have no problem with Tocharian R1a.

Silesian
12-03-2013, 05:38 PM
..... In short he saw the isogloss as a geographical split. We now realise that it is chronological:


c. 4200 BC - The earliest form of PIE had neither the centum or satem sound. It formed the basis for the Anatolian branch.
c. 3500 BC - The Centum sound had developed. Centumization removed the palatovelars from the language. It did not matter which direction people went from the steppe. They took with them this form of PIE. It formed the basis for the Tocharian, Celtic, Italic, Germanic and Greek branches.
c. 2500 BC - The Satem sound developed. This formed the basis for the Baltic, Slavic, Indic and Iranian branches. Again it did not matter in which direction people went. They took this sound.


https://upload.wikimedia.org/wikipedia/commons/a/ae/Centum_Satem_map.png

So the oldest PIE language with neither centum-satem, is flanked by Greek on one side,

is an independent branch of the Indo-European family of languages. Native to the southern Balkans, western Asia Minor, Greece, and the Aegean Islands, it has the longest documented history of any Indo-European language, spanning 34 centuries
and Armenian on the other? Two independent language branches?

Linguists classify Armenian as an independent branch of the Indo-European language family.[2]

Ral
12-03-2013, 05:46 PM
Let me make one thing linguists will not go on about the geneticists.
Linguistics has its own developed methodology. Some linguistic concepts themselves are controversial and not well recognized (glottochronology and so on). But never linguists would say something like: language A and language B descended from language С according to genetics and DNA. Linguists can only cede part controversial concepts.
Otherwise it will be the end of linguistics as a science.

Jean M
12-03-2013, 05:49 PM
So the oldest PIE language with neither centum-satem, is flanked by Greek on one side, and Armenian on the other?

Yes, but that is just the product of the accidents of history. As far as we can work out, Armenian springs from the same source as Greek - a group of people who moved into the Balkans. The ancestors of the Armenians did not arrive in what is now Armenia until the 6th century BC.

Jean M
12-03-2013, 06:00 PM
Where are (for example) Romanian,Greek,Albanian, Armenian branches?

Exactly. We cannot explain the spread of IE languages with R1a alone. But it is exciting to see the links appearing between the different branches of Balto-Slavic and Indo-Iranian (plus some Germanic.)

Jean M
12-03-2013, 06:05 PM
Let me make one thing linguists will not go on about the geneticists. Linguistics has its own developed methodology.

This needs to be clearly understood by all engaged in these debates. For example we cannot use estimated dates from DNA to date a language. Languages are dated on their vocabulary. We cannot use DNA to contradict evidence from linguistics. We can only use genetic evidence to get a clearer idea of how languages spread i.e. by mass movement or female movement or elites or whatever.

Silesian
12-03-2013, 06:13 PM
Yes, but that is just the product of the accidents of history. As far as we can work out, Armenian springs from the same source as Greek - a group of people who moved into the Balkans. The ancestors of the Armenians did not arrive in what is now Armenia until the 6th century BC.
Is it correct to say that around the region of the oldest written attested non centum-satem we have the oldest attested written centum branch[Greek Linear A/B]. The oldest attestation of ZoroasterAvestan , Behistun Inscription.
Possibly the oldest attestation to Kurds by Sumerians. In the symbol of mountains.

��, a pictograph of a mountain.[1......t is likely that this name coincides with the modern day Kurds who are the predominant ethnicity inhabiting much of the Zagros mountain range.Hennerbichler believes the term Kurd and similar ethnic labels to have been derived from the Sumerian word stem “kur”, meaning mountain. [5]

https://en.wikipedia.org/wiki/Kur

Also found in Albanian as PIE


From Proto-Albanian *kur, from Proto-Indo-European *kʷur (compare Lithuanian kur̃, Armenian ur).

http://en.wiktionary.org/wiki/kur

The oldest attested peace treaty between Hittite-Mitanni, using ocean god Aruna, Varuna?

In a treaty between the Hittites and the Mitanni (between Suppiluliuma and Shattiwaza, ca. 1380 BC), the deities Mitra, Varuna, Indra, and Nasatya (Ashvins) are invoked

Humanist
12-03-2013, 06:38 PM
Possibly the oldest attestation to Kurds by Sumerians. In the symbol of mountains.

So, you mean to suggest as a possibility that the modern Kurdish ethnonym has been around for 4000+ years? I find that extremely unlikely. I sincerely doubt that any group of people on this planet has any such continuity. At least as far as what is today Europe, ME, and West Asia are concerned.

parasar
12-03-2013, 06:53 PM
I do not have an issue with the the basic structure, but I want to reconcile the timelines with the genetics. If we keep trying to fit genetics to preconceived notions then we are not using genetics as independent evidence.

As I have noted a number of times, it is not a coincidence that Y-R including R-M479 correlates so strongly with IE languages. Therefore the chronology must follow the rate of change of isolated languages such as Icelandic (geography) or Sanskrit (caste and convention) that show very little change, and this must have been the case with much higher isolation in the past.

Therefore I think timelines should be expanded perhaps five-tenfold.

~40000 ybp - The earliest form of PIE Y-P(xQ) moves west
~24000 ybp - PIE Y-R move east with LGM. R-M479 separates out and stays east.
~12000 ybp - The present distribution of R1b and R1a develops, and for the most part has stayed like that with Z93 in the east and Z283 in the middle, and M269 in the west.

I do not see the Satam and Kentum as much of a split since 'non-refined' linguistic groups in the east (i.e. those few not Persianized/Sankritized) retain the Kentum forms.

Silesian
12-03-2013, 06:57 PM
So, you mean to suggest that the modern Kurdish ethnonym has been around for 4000+ years? I find that extremely unlikely. I sincerely doubt that any group of people on this planet has any such continuity. At least as far as what is today Europe, ME, and West Asia are concerned.
Cllassical Syriac.

Mount Judi (Arabic: الجودي‎ al-Ǧūdī, Aramaic: קרדו‎ Qardū,[1] Kurdish Cûdî,
Classical Syriac: ܩܪܕܘ Qardū,[1]

parasar
12-03-2013, 07:00 PM
So, you mean to suggest as a possibility that the modern Kurdish ethnonym has been around for 4000+ years? I find that extremely unlikely. I sincerely doubt that any group of people on this planet has any such continuity. At least as far as what is today Europe, ME, and West Asia are concerned.

The query is addressed to Silesian, right?

jamesdowallen
12-03-2013, 07:04 PM
Do others find it curious that Basque speakers have very high concentration of R1b-L11? Is it plausible that the group of West Europeans who best retained their language would be the same group that most thoroughly replaced its males?

Instead, perhaps Basque and R1b were brought together to Western Europe before the Indo-European expansion, perhaps by the Cardial Ware farmers who leapfrogged along the Mediterranean coast ca 5500 BC. (Bengt and Ruhlen think Basque and Hurrian had a common early Neolithic ancestor.) West Europe subsequently adopted Celtic by language shift, not immigration.

I'm sure this seems far-fetched. But how do you explain the nearly pure Basque R1b?

parasar
12-03-2013, 07:09 PM
Do others find it curious that Basque speakers have very high concentration of R1b-L11? Is it plausible that the group of West Europeans who best retained their language would be the same group that most thoroughly replaced its males?

Instead, perhaps Basque and R1b were brought together to Western Europe before the Indo-European expansion, perhaps by the Cardial Ware farmers who leapfrogged along the Mediterranean coast ca 5500 BC. (Bengt and Ruhlen think Basque and Hurrian had a common early Neolithic ancestor.) West Europe subsequently adopted Celtic by language shift, not immigration.

I'm sure this seems far-fetched. But how do you explain the nearly pure Basque R1b?

Or perhaps Basque was the language of the Y-G (more likely G, but others too) whose Y got replaced or supplanted, but the due to isolation the prior language was still retained.

Humanist
12-03-2013, 07:20 PM
Cllassical Syriac.

Silesian. I have posted about this before (see my post, from another forum, below). However, this certainly does not mean that the same people have carried the same name for 4000+ years.


Corduene is a possible corruption of the Akkadian (and Syriac) name for that general location, Mat/Bet Qardu. Translated, it means something along the lines of "home/land of the brave/heroic ones."


An example of the significance of the term "qardu" in Neo-Assyrian society:

The Mythology of Kingship in Neo-Assyrian Art (http://www.amazon.com/Mythology-Kingship-Neo-Assyrian-Art/dp/0521517907/ref=sr_1_1?s=books&ie=UTF8&qid=1384043336&sr=1-1&keywords=0521517907)

Mehmet-Ali Ataç (2010)

http://i1096.photobucket.com/albums/g326/dok101/Faces/qardu__.jpg


EDIT

Wikipedia


Carduchoi in Xenophon

A people called the Carduchoi are mentioned in Xenophon's Anabasis. They inhabited the mountains north of the Tigris in 401 BC, living in well-provisioned villages. They were enemies to the king (of Persia), as were the Greek mercenaries with Xenophon, but their response to thousands of armed and desperate strangers was hostile. They had no heavy troops who could face the battle-hardened hoplites, but they used longbows and slings effectively, and for the Greeks the "seven days spent in traversing the country of the Carduchians had been one long continuous battle, which had cost them more suffering than the whole of their troubles at the hands of the king and Tissaphernes put together."[9]


The query is addressed to Silesian, right?

That is correct.

Silesian
12-03-2013, 07:21 PM
Do others find it curious that Basque speakers have very high concentration of R1b-L11? Is it plausible that the group of West Europeans who best retained their language would be the same group that most thoroughly replaced its males?
The same query can made by substituting, Brahui for Basque and R1a for R1b.

Do others find it curious that Brahui speakers have very high concentration of R1a? Is it plausible that the group of Dravidians who best retained their language would be the same group that most thoroughly replaced its males?

newtoboard
12-03-2013, 07:24 PM
The same query can made by substituting, Brahui for Basque and R1a for R1b.

Do others find it curious that Brahui speakers have very high concentration of R1a? Is it plausible that the group of Dravidians who best retained their language would be the same group that most thoroughly replaced its males?

Like you have been told before by Jean and other the Brahui didn't retain their language. The consensus is that they are migrants from Central India.

newtoboard
12-03-2013, 07:25 PM
Brahui actually replaced the IE speakers.

Silesian
12-03-2013, 07:30 PM
Silesian. I have posted about this before (see my post, from another forum, below). However, this certainly does not mean that the same people have carried the same name for 4000+ years.

Kur in PIE and is found in Sumerian-Akkadian/Bablyonian

Akkadian (lišānum akkadītum, �������� ak.kADû) (also Accadian, Assyro-Babylonian)[1]...... Akkadian Empire (ca. 2334–2154 BC), although the language predates the founding of Akkad......The mutual influence between Sumerian and Akkadian had led scholars to describe the languages as a sprachbund.[3] Akkadian proper names were first attested in Sumerian texts from ca. the late 29th century BC...... In later Babylonian myth Kur is possibly an Anunnaki,....According to The Oxford Companion to World Mythology, the Anunnaki "are the Sumerian deities of the old primordial line; they are chthonic deities of fertility, associated eventually with the underworld, where they became judges. [4]


Sumerian:


A second, popular meaning of Kur was "underworld", or the world under the earth.[3]



Kurdalægon (Ossetian: Куырдалӕгон[1]) is the heavenly god of the blacksmiths in Ossetian mythology. His epithet is the heavenly one; he shows the dead man's horse, thus helping him on his journey to the other side. He is a close friend of the Narts.

Silesian
12-03-2013, 07:33 PM
Brahui actually replaced the IE speakers.

So we have evidence that R1a did not spread the IE language to the Dravidian speaking Brahui and Sanskrit has Dravidian content? Perhaps PIE speaking R1a from destination X could not have spread PIE with no centum-satem to Anatolia?

Ral
12-03-2013, 08:26 PM
I do not have an issue with the the basic structure, but I want to reconcile the timelines with the genetics. If we keep trying to fit genetics to preconceived notions then we are not using genetics as independent evidence.

As I have noted a number of times, it is not a coincidence that Y-R including R-M479 correlates so strongly with IE languages. Therefore the chronology must follow the rate of change of isolated languages such as Icelandic (geography) or Sanskrit (caste and convention) that show very little change, and this must have been the case with much higher isolation in the past.

Therefore I think timelines should be expanded perhaps five-tenfold.

~40000 ybp - The earliest form of PIE Y-P(xQ) moves west
~24000 ybp - PIE Y-R move east with LGM. R-M479 separates out and stays east.
~12000 ybp - The present distribution of R1b and R1a develops, and for the most part has stayed like that with Z93 in the east and Z283 in the middle, and M269 in the west.

I do not see the Satam and Kentum as much of a split since 'non-refined' linguistic groups in the east (i.e. those few not Persianized/Sankritized) retain the Kentum forms.

You must understand that the limiting factor here , do not let too ancientize division of Indo-European language is part of the overall specific vocabulary related to relatively new concepts - wheel and so on.
But this is not so simple.
1. many terms associated with technical innovations can be "nomadic" .
2 . In linguistics , as far as I know , in order to recognize the term " all-Indo-European " is not necessarily to this term were present in all IE languages ​​, but only in a certain number of languages. (It is assumed that he was in the other , but disappeared ) .
Generally , linguistics is not an exact science and there errors may be .
But personally, I admit that IE lang could start to disintegrate 8,000 years ago (rather than 5-6 thousand years). Such concepts exist .
I guess that linguistic diversity in Europe due to the fact that the Indo-Europeans lived in large parts of Europe for a long time in small isolated groups ( caucasian model with its linguistic diversity ) . Moreover pre-IE were relatively genetically uniform .Was this situation in Europe and when-I do not know.
*This model does not agree with the concept of a large mass invasion IE -speakers outside with a relatively uniform lang.
In this case, I believe the European languages ​​would be separated from each other not more than Iranian (or Indo-Iranian).

newtoboard
12-03-2013, 08:47 PM
I understood perfectly. I just specified the location because the Indian subcontinent is a big place! The pattern of migrations into it and within it seems very complex, and I can't claim to have a real grip on it. But if we are looking for remnants of the Kushans, the north-west is where we would look I think.

The lack of M73 there is indicative that Kushans and Tocharians did not have M73 imo. Even the R1b-M73 often mentioned in Northern Pakistan belongs to the Hazara and the Pakistani Hazaras are recent imigrants from Central Afghanistan. There is also no reason why a group new to the region would preserve Kushan elements better than the natives of NW South Asia and South Central Asia (at least native in comparison to the Hazaras).

newtoboard
12-03-2013, 08:49 PM
I should add that deduced Tocharian R1a + Iranian R1a does not rule out R1b of some type moving eastwards at some time. The possibility remains of


R1b in trace amounts among Iranian speakers
R1b-M73 having its own trajectory unconnected to IE, e.g. moving east along the Silk Road from some site in the Caucasus


This should be considered as a strong possibility.

parasar
12-03-2013, 09:14 PM
You must understand that the limiting factor here , do not let too ancientize division of Indo-European language is part of the overall specific vocabulary related to relatively new concepts - wheel and so on.
But this is not so simple.
1. many terms associated with technical innovations can be "nomadic" .
2 . In linguistics , as far as I know , in order to recognize the term " all-Indo-European " is not necessarily to this term were present in all IE languages ​​, but only in a certain number of languages. (It is assumed that he was in the other , but disappeared ) .
Generally , linguistics is not an exact science and there errors may be .
But personally, I admit that IE lang could start to disintegrate 8,000 years ago (rather than 5-6 thousand years). Such concepts exist .
I guess that linguistic diversity in Europe due to the fact that the Indo-Europeans lived in large parts of Europe for a long time in small isolated groups ( caucasian model with its linguistic diversity ) . Moreover pre-IE were relatively genetically uniform .Was this situation in Europe and when-I do not know.
*This model does not agree with the concept of a large mass invasion IE -speakers outside with a relatively uniform lang.
In this case, I believe the European languages ​​would be separated from each other not more than Iranian (or Indo-Iranian).

I have taken into account the wheel argument. They could well be later shared innovations. Or, I had mentioned in another thread (about vehicle and axle) that we are looking at components that have more basic meaning that are also applicable to the wheel.
http://www.anthrogenica.com/showthread.php?1517-Afghan-Hindu-Kush-Where-Eurasian-Sub-Continent-Gene-Flows-Converge&p=17940&viewfull=1#post17940
http://www.anthrogenica.com/showthread.php?1517-Afghan-Hindu-Kush-Where-Eurasian-Sub-Continent-Gene-Flows-Converge&p=17943&viewfull=1#post17943

My timelines are just a proposal based on current Y distribution.

Jean M
12-03-2013, 09:31 PM
My timelines are just a proposal based on current Y distribution.

This is the problem. Current Y-distribution does not tell us the date for PIE.


I have taken into account the wheel argument. They could well be later shared innovations. Or, I had mentioned in another thread (about vehicle and axle) that we are looking at components that have more basic meaning that are also applicable to the wheel.

Wriggling around with vehicle vocabulary won't change anything. That vocabulary is just part of a set relating to post-Neolithic innovations. The first farmers used digging sticks rather than ploughs. They had no wheels or wagons, no gold or silver. They kept cattle for beef, not milk and cheese. They did not make wine. They did not spin wool. Yet PIE had words for all these things.

Does it not occur to you that linguists, specialists in PIE, have studied this topic for decades. If it was easy to dismiss the Copper Age dating for PIE, it would have been done and accepted by linguists long ago.

parasar
12-03-2013, 10:04 PM
This is the problem. Current Y-distribution does not tell us the date for PIE.

To me it often does, otherwise sharp correlations would not exist. Languages did not go around picking Y dna and say - oh ok - you are O so no IE for you. But such boundaries are (or were) the norm.



Wriggling around with vehicle vocabulary won't change anything. That vocabulary is just part of a set relating to post-Neolithic innovations. The first farmers used digging sticks rather than ploughs. They had no wheels or wagons, no gold or silver. They kept cattle for beef, not milk and cheese. They did not make wine. They did not spin wool. Yet PIE had words for all these things.

Does it not occur to you that linguists, specialists in PIE, have studied this topic for decades. If it was easy to dismiss the dating for PIE, it would have been done and accepted by linguists long ago.

It does.

I don't know what PIE had or did not have, it is a construct of those same linguists. DNA correlations give us independent techniques. If they match what the linguists are positing, then yes their theories are confirmed. Otherwise they need to be re-examined in light of the new evidence.

Jean M
12-03-2013, 10:26 PM
To me it often does, otherwise sharp correlations would not exist.

Correlations exist because a child usually learns his or her first language from his or her biological parents. That is not always the case, but true often enough to produce a statistically significant correlation. But what we must understand is that:


Languages change constantly, and not necessarily at a fixed rate.
Meanwhile mutations accrue along both mtDNA and Y-DNA lineages, but not necessarily in exact synchronicity with linguistic change.
A man can marry a different language speaker, who brings up their children in her language. Or a man can change language because of adoption into a different community, so that a Y-lineage suddenly belongs to a different language group after that change.


There is no way to use DNA to confirm or refute the dates ascribed to a particular language on linguistic grounds. Ral is absolutely right about that.

alan
12-03-2013, 10:46 PM
Shame there is no M73 in the south Caucasus and just a tiny bit among a Turkic group in the north Caucasus. I think M73 is very unlikely to have originated in the Caucasus. I would think it possible that the P297* could have passed through there but I am only saying that because there is none anywhere and it could have originated anywhere really.

If you really are looking for a location and a non-IE angle on M73 then somewhere like Uralic groups in western Siberia or the Urals or the north of west central Asian is more plausible.

I personally think though that we have to accept that M73 is scattered in such a way that it was not massive anywhere and that would make it susceptible to language shifts. In all likelihood its language or dialect at the time it arose c. 5000BC does not exist today. It is vastly older than the Turkic expansion and its phylogeny rules out some far eastern source. I personally think loads of dialects and languages that no longer exist probably were present in 5000BC and the present Slavic/Iranian/Turkic/Uralic mix probably is due to dominance of those groups many millenia later. There could have been a myriad of dialects before any that survive today. Slavic in relatively new to much of modern Russia and Ukraine. Turkic is a Medieval spread. proto-Uralic is thought to not even have existed in the early PIE period by some credible scholar. Iranian, well I dont know the details but its also millenia later than M73. Even PIE itself is usually put somewhat later than the origin date of M73. So, IMO its irrational to think we can know what an early clade like M73 spoke as it existed at a time before any known dialects of IE and even pre-dates PIE and proto-Uralic. If the majority of scholars ideas on language dating is correct first M73 men would not have spoken any language or even proto-language that we know about or have even reconstructed. So, I think this is a pointless exercise. Who knows what sort of odd pre-Anatolian or pre-proto Uralic, pre-proto Caucasian or even extinct language family they would have spoken.

The only hint we can get for M73's language is that it could have been a very distant cousin of whatever M269 spoke, but with a separation time depth of anything up to 5000 years from their common P297 ancestor. That is the equivalent time depth between modern languages and Yamnaya or pre-proto Germanic and modern English. M73 could have spoken a very distant cousin language to IE albeit if a linguist could be transported back in time to 4000BC they probably would be able to spot the distant cousin relationship.

Another reason I suspect that the steppe may have been riddled with very divergent languages is that it was not until Yamanaya or perhaps Suvorovo period that they were fully mobile pastoralists. Before that they were probably mainly separated and isolated by vast areas of dry steppe. The myriad of cultures in the pre-Yamanaya period and even the remnants that existed through the Yamnaya period as separate entities is perhaps the material expression of language diversity, possibly not just dialect diversity.

There were probably some unifying groups capable of expanding their language or dialect in a wider way and perhaps may have caused convergence. The only pre-Yamnaya group capable of that would be Stedny Stog IMO. They were the first wider networkers in the steppe and their influence ran the length of the western steppes, even to the north Caucasus. Prior to that I think there could have been languages across the western steppe with thousands of years of divergence.

So, my basic message is its pointless to try and interpret something like M73 or what its carriers speak today when its simply too ancient to relate to any existing languages. It was apparently a small clade and it may have changed its language identity many times. The later history of the steppe from 1000BC to today alone should warn us against this. Noone on the western steppe today speaks a language of any massive great in-situ time depth.


This should be considered as a strong possibility.

alan
12-03-2013, 10:57 PM
I would also add that unless a language has existing survivors or cast-iron cases of burials to dna test where we know what they spoke at a given time in a given place because of historical sources or inscriptions then there is no basis for working backwards to infer language of a lineage. Even in the exceptions we cannot be sure if language shift had previously happened to that lineage. I think this is a whole lot more difficult and perhaps impossible than we like to admit. I think unless they can dig up some certain historically attested late Tocharian speakers then we will never know if the Tarim mummies relate to that group or not.

parasar
12-03-2013, 11:01 PM
Correlations exist because a child usually learns his or her first language from his or her biological parents. That is not always the case, but true often enough to produce a statistically significant correlation. But what we must understand is that:


Languages change constantly, and not necessarily at a fixed rate.
Meanwhile mutations accrue along both mtDNA and Y-DNA lineages, but not necessarily in exact synchronicity with linguistic change.
A man can marry a different language speaker, who brings up their children in her language. Or a man can change language because of adoption into a different community, so that a Y-lineage suddenly belongs to a different language group after that change.


There is no way to use DNA to confirm or refute the dates ascribed to a particular language on linguistic grounds. Ral is absolutely right about that.

This line of discussion, started with Michał's post, to which responded - "Golly Michał! That's is very brave of you. The early stages of your model are completely different from mine" -
why even go with such exercises if not to confirm or refute? Isn't that the purpose?

And why would these correlations exist in far flung locations such as Iceland and Lanka. It is because, IMO, language is quite persistent, and even moving to far off locations, folk at least in the past stuck to their languages. Obviously we will never be able to prove that the MA-1's folk were archaic IE speaking, but just going by his R and U, I would think that that is good bet. Plus autosomally MA-1 was part Indo (IE speaking), part European (IE speaking), 0% West Asian (not IE speaking), 0% East Asian (not IE speaking), and 0% Siberian (not IE speaking), just provides intuitive support.

Jean M
12-03-2013, 11:43 PM
This line of discussion, started with Michał's post, to which responded - "Golly Michał! That's is very brave of you. The early stages of your model are completely different from mine" -
why even go with such exercises if not to confirm or refute? Isn't that the purpose?


I cannot speak for Michał, but what I'm interested in doing is looking for correlations between languages and haplogroups that may help me to understand migrations. I am not looking for genetic confirmation of linguistic dating. On the contrary, deductions from linguistics, history and archaeology may be used upon occasions to help calibrate mutation rates used by geneticists.


And why would these correlations exist in far flung locations such as Iceland and Lanka. It is because, IMO, language is quite persistent, and even moving to far off locations, folk at least in the past stuck to their languages.

PIE is spoken in India and Iceland? It manifestly isn't. You just declared a moment ago that you don't know the vocabulary of PIE because it is no longer spoken. PIE is long gone. Language changes. I think I understand the romantic impulse that moves you to periodically leap towards Palaeolithic Continuity Theory. But only a few days ago, you were willing to accept from Michał's estimates of the dates for various R1a1 lineages that the


age of IE is at a minimum about 5500 years, though an age of >6500 years is more likely.
http://www.anthrogenica.com/showthread.php?1507-Some-provisional-calculations-for-haplogroup-R1a-based-on-the-first-FGC-result&p=20741&viewfull=1#post20741

I feel that you were mistaken in thinking that you could use these estimates to fix the date of PIE, but you were actually accepting a date close to that estimated by linguists. ;)

parasar
12-03-2013, 11:54 PM
I cannot speak for Michał, but what I'm interested in doing is looking for correlations between languages and haplogroups that may help me to understand migrations. I am not looking for genetic confirmation of linguistic dating. On the contrary, deductions from linguistics, history and archaeology may be used upon occasions to help calibrate mutation rates used by geneticists.



PIE is spoken in India and Iceland? It manifestly isn't. You just declared a moment ago that you don't know the vocabulary of PIE because it is no longer spoken. PIE is long gone. Language changes. I think I understand the romantic impulse that moves you to periodically leap towards Palaeolithic Continuity Theory. But only a few days ago, you were willing to accept from Michał's estimates of the dates for various R1a1 lineages that the


http://www.anthrogenica.com/showthread.php?1507-Some-provisional-calculations-for-haplogroup-R1a-based-on-the-first-FGC-result&p=20741&viewfull=1#post20741

I feel that you were mistaken in thinking that you could use these estimates to fix the date of PIE, but you were actually accepting a date close to that estimated by linguists. ;)
Oops did I say that? As that is my lowest possible estimate if R1a goes with IE which I don't think possible.

alan
12-04-2013, 12:06 AM
Although I agree with that, I also think that modest dispersal of IEs could have occurred say 500 or 1000 years before the inventions because I dont think it would be easy or even possible to distinguise between an inherited word from the PIE source and a word that was passed on between late PIEs who had modestly dispersed already. Its one thing to be able to spot borrowings between branches but its another to spot a loan passed after dispersal between very similar later PIE languages. So I would always feel that use of material objects to date the PIE initial dispersal should have a 500-1000 year wriggle room. To give an example, I dont think the wheel being invented and the word being coined perhaps c. 3500BC could be spotted as a loan taken on by early PIE groups who dispersed a little earlier (but not completely out of contact) in say 4000BC or even perhaps 4500BC. That is why I doubt the Suvorovo groups were in general Anatolians as Anthony argues. They remained closeby and made return journeys back to the steppe and the wheel word would and almost certainly did reach them a few centuries later.

So, I agree that the vocab is that of the copper age although I would accept a date for PIE at least as far back as 4500BC.

I also think that the possible location of the rise of PIE is a little broader than the Volga-Urals. I think the PIE into proto-Uralic borrowing concept is far to unsafe now to even use it as an arguement and I think that widens the possibilities for PIE at least the length of the western steppes. Many of the aspects of PIE vocab can also be placed slightly off the steppe as well as the steppe including gold, copper, dairying, the wheel etc are all earlier outside the steppe albeit relatively closeby in the Balkans and Caucasus etc.

So, while I broadly agree with the copper age date and general zone where PIE must have emerged I think a Yamanaya model is too prescriptive and I would include the whole western steppe and even its fringes beyond and even in a steppe model anytime in the period after 4500BC seems possible to me as a date when PIE had basically evolved. I think the Suvorovo horizon was responsible for the roots of many of the palaeo-Balkan IE languages as well as Celtic and Italic. I do not find the idea that they were on the whole Anatolian branches convincing at all. More likely they were full PIEs albeit that they learned of the wheel and the word for it several centuries later. To me Anatolians had to have been separated well off from the main groups and certainly they couldnt have passed through the Caucasus given that the wheel is earlier there than on the steppes. IMO to have avoided the standard PIE wheel vocab reaching them they must have palced themselves in a spot away from the steppe, Caucasus or any other area where the knowledge of the wheel wouldnt have quickly reached them

Then again maybe that is entirely wrong and the Anatolians had different wheel vocab because they were located in the Caucasus and recieved the wheel before their steppes neighbours did and had no need to join in with the Yamnaya wheel vocab.


This is the problem. Current Y-distribution does not tell us the date for PIE.



Wriggling around with vehicle vocabulary won't change anything. That vocabulary is just part of a set relating to post-Neolithic innovations. The first farmers used digging sticks rather than ploughs. They had no wheels or wagons, no gold or silver. They kept cattle for beef, not milk and cheese. They did not make wine. They did not spin wool. Yet PIE had words for all these things.

Does it not occur to you that linguists, specialists in PIE, have studied this topic for decades. If it was easy to dismiss the Copper Age dating for PIE, it would have been done and accepted by linguists long ago.

nuadha
12-04-2013, 12:47 AM
Let me try to explain more clearly. The centum-satem isogloss is linguistic. We may be trying to link haplogroups to it, but the isogloss itself - the sound change - is linguistic. Now looking just at the linguistic evidence i.e. where we find written evidence of people speaking this or that language, Tocharian was east of the European steppe, while Italic and Celtic were (and remain) west of the European steppe. See map: https://en.wikipedia.org/wiki/Centum-satem_isogloss

That ruined the idea that Von Bradke had in 1890. He thought that the first branching of PIE was into an eastern Satem group, and a western Centum group. In short he saw the isogloss as a geographical split. We now realise that it is chronological:


c. 4200 BC - The earliest form of PIE had neither the centum or satem sound. It formed the basis for the Anatolian branch.
c. 3500 BC - The Centum sound had developed. Centumization removed the palatovelars from the language. It did not matter which direction people went from the steppe. They took with them this form of PIE. It formed the basis for the Tocharian, Celtic, Italic, Germanic and Greek branches.
c. 2500 BC - The Satem sound developed. This formed the basis for the Baltic, Slavic, Indic and Iranian branches. Again it did not matter in which direction people went. They took this sound.



OK, I see what you are saying. when I said it was a geographic split I meant that satem speaking people had to have physical distance from centum speaking people for the two to diverge, whereby the eventual satum people picked up a change and other centum people didn't.

Clearly IE people are associated with r1a and r1b. The point I make is that the linguistic connection r1b people have to each other, in the IE world, goes back to the early history of IE which is before the system shift. The linguistic connection r1a people have with each other in the IE world goes back to the time and place where satem evolved.

The idea here is that it is much more likely that pre satem people, who happened to be carrying r1a, shifted their language to IE rather than pre Anatolian, pre Tocharian, and pre Italo-celtic (all pre satem) people all carrying r1b by chance and then each of them shifted their language to IE.

So when people associate r1a and r1b with IE, in essence, they should also associate r1b with early IE and r1a with later (Satem) IE.

nuadha
12-04-2013, 12:53 AM
BTW, the division between r1a and r1b and their relation to past migrations isn't some absolute line; its just a difference in overall patterns which is reason enough to talk about them separately.

I know things aren't black and white with one migrations transferring one new haplogroup.

nuadha
12-04-2013, 01:23 AM
Do others find it curious that Basque speakers have very high concentration of R1b-L11? Is it plausible that the group of West Europeans who best retained their language would be the same group that most thoroughly replaced its males?

Instead, perhaps Basque and R1b were brought together to Western Europe before the Indo-European expansion, perhaps by the Cardial Ware farmers who leapfrogged along the Mediterranean coast ca 5500 BC. (Bengt and Ruhlen think Basque and Hurrian had a common early Neolithic ancestor.) West Europe subsequently adopted Celtic by language shift, not immigration.

I'm sure this seems far-fetched. But how do you explain the nearly pure Basque R1b?

this question doesn't make any sense to me. you would rather have all non basque western r1b men change their language rather than a small group of r1b men shifting their language to basque. but this contradicts the value you give to correlations...

alan
12-04-2013, 03:44 AM
Yes the Basque-R1b-non-IE thing has had its day. Its one small exception to the basic rule that European R1b is nearly all IE speaking. One exception in a small group should not lead to the tail wagging the dog. I dont know what the big deal is. All the rest of us have accepted that our Palaeolithic and early Neolithic ancestor's male lines were massively elipsed after the Neolithic. That doesnt mean that we dont have any heritage of that period, we probably have a significant amount of autosmal DNA from them. Just not on the y line. Autosomal DNA seems to have a strongly north-south cline while Y dna cuts across that pattern. I suspect the north European element must have a lot of Mesolithic origins. Ireland for example is very high in north European, higher than England, despite missing out on most of the Germanic input so its hard to see how it got so high there unless there is a large Mesolithic component that got absorbed into the Neolithic population.


this question doesn't make any sense to me. you would rather have all non basque western r1b men change their language rather than a small group of r1b men shifting their language to basque. but this contradicts the value you give to correlations...

parasar
12-04-2013, 03:46 AM
Oops did I say that? As that is my lowest possible estimate if R1a goes with IE which I don't think possible.

This is what I had posted:

So with a few hundred years here and there, it can be said that:
1.Age of Z93=Age of Z283=Age of CTS4385 - About 6500 years
2. Age of L657=Age of M458 - About 5500 years

1 does give us a split, but not a clear sharp split between India and Europe.
2 gives us a mutually exclusive split between India and Europe.

Therefore, if R1a correlates with IE, then age of IE is at a minimum about 5500 years, though an age of >6500 years is more likely.

The above were the minimum dates based on the Full Genome SNP based TMRCAs for M458 and L657.
The reason these dates, IMO, establish the minimums is because of their mutually exclusive status in Europe and India.

To clarify, these dates are not what I would estimate as the age of IE taking into account R2 and R1b, especially the latter which I associate as much with archaic IE as I do R1a.

parasar
12-04-2013, 03:55 AM
Yes the Basque-R1b-non-IE thing has had its day. Its one small exception to the basic rule that European R1b is nearly all IE speaking. One exception in a small group should not lead to the tail wagging the dog. I dont know what the big deal is. All the rest of us have accepted that our Palaeolithic and early Neolithic ancestor's male lines were massively elipsed after the Neolithic. That doesnt mean that we dont have any heritage of that period, we probably have a significant amount of autosmal DNA from them. Just not on the y line. Autosomal DNA seems to have a strongly north-south cline while Y dna cuts across that pattern. I suspect the north European element must have a lot of Mesolithic origins. Ireland for example is very high in north European, higher than England, despite missing out on most of the Germanic input so its hard to see how it got so high there unless there is a large Mesolithic component that got absorbed into the Neolithic population.

Couldn't have said it better! And that same logic applies to folk who keep bringing up the isolate Brahui's Dravidian status in connection with R1a.

alan
12-04-2013, 04:00 AM
Generally speaking I believe the absence of R1b to date in any pre-beaker ancient DNA is partly because it simply wasnt around in most of Europe much before then and partly because of the absence of testing in so many of the cultures in the areas of Europe where it is most likely to have been hiding c. 5000-3000BC judging by the modern distribution of earlier clades - places like the western end of the western steppes, the Caucasus, the Balkans etc. It really hasnt stood a chance of being detected. A lot of discussion about R1b in Asia is all very well but noone is saying it was a big player in the eastern steppes, central Asia etc. It clearly was the big player further west and whatever its oldest origins, it must have passed through somewhere like the Balkans or the Danube in pre-beaker times on its way further west. Its practically impossible to explain R1b in central and western Europe without that sort of movement. The common idea that R1b magically jumped from the east to Iberia and then spread back from Iberia with beaker pots is IMO baloney and a far too literal interpretation of the beaker phenomenon. Its far more likely that beaker and R1b were separate things until they met up somewhere like the Alps or central Europe. I also really doubt that beaker pot in its initial phase had anything whatsoever to do with IE or R1b. My money is on G or E. I doubt beaker and R1b came together much before 2600BC. The earliest beaker culture in Iberia and adjacent completely lacks IE cultural characteristics IMO.

Michał
12-04-2013, 01:16 PM
My apologies.
Accepted. :)


I should have phrased that better. The two are very connected. And Timber Grave is usually assumed to be the ancestor of European Scythians or Cimmerians (who might be Iranian speakers but are almost certainly not Indo-Aryan speakers).
I agree that it is very likely that Srubna was associated with the ancestors of the Cimmerians. However, if you associate Andronovo (1900-1500 BC) with the Proto-Indo-Iranians, there is no way that Cimmerians could have spoken any Iranian language (not to mention any Scythian language as you seem to suggest, as well).



So it is hard to picture An Indo-Aryan culture in between two cultures which ultimately both end up giving rise to NE Iranian speakers (if Afanasevo leads to the steppe Saka or Scythians as I assume it does in your scenario)
Since I expect you’ll agree that there is no chance that Srubna was associated with any Iranian language (see above), I must admit that I don’ t see any major problem here. BTW, I have noticed that you don’t see a similar problem when some early Indo-Aryan-speaking populations (like the Mitanni Aryans and the Early Indo-Aryans from India) are separated by the region that should be supposedly associated with some Early Iranians (or even Proto-Iranians) according to your scenario.



You would still need a movement from the former territory of this Proto Iranian culture (Afanasevo) or from the Tarim to go through Andronovo to get to the final destinations of the Persians and Medes.
Right, and it seems quite likely that following the decline of the Andronovo culture at about 1500 BC, its former territory (occupied mostly by the so-called Sargary culture) has experienced some substantial influx of people from the East which started with the expansion of the Karasuk culture and some of the closely related (or derived) cultural groupings (like the descending Irmen culture). This relatively early westward expansion (strongly associated with the introduction of the “true nomadism”) is for example suggested by Hanks, Epimakhov and Renfrew in their paper “Towards a refined chronology for the Bronze Age of the southern Urals, Russia”:

“The results of our study for the Final Bronze Age have revealed some significant characteristics. As noted above, the cultural attribution of sites of this period is not clearly understood and they reflect an extended time interval from approximately the late fourteenth to the tenth centuries BC. This coincides with changes in sedentary settlement patterning and what appears to be an increase in the scale of mobile pastoralism (Kostyukov et al. 1995a; 1995b; Kostyukov 1999). It is important to note that the Final Bronze Age has not received the same level of archaeological investigation as other periods. However, our results confirm that the sites we sampled were contemporary. As a result of calibration, the youngest limit of the Bronze Age phase was adjusted back by one century, and the oldest limit was adjusted to the second half of the fourteenth century BC. As noted above, there has been a persistent problem with coupling absolute dates with theorised cultural stages. Here we will just note that the material patterns of these periods reflect very little in the way of common characteristics and one cannot disregard the possibility of the movement of new populations and/or cultural traditions into the Urals region from West Siberia. At the present moment, much more additional research will be required to address this problem.
[…]
A West Siberian cultural influence, possibly the Irmen', is detectable within the funerary rites, which consist of human remains placed in a crouched position with a southern orientation of the head, relatively small tomb constructions, and similarities in the arrangement of included burial goods. A number of chance finds of Karasuk shaped daggers in the Urals region also appears to support an important connection with West Siberia. Whether these similarities reflect actual population migration into the Urals or the diffusion of cultural practices and inter-regional trade cannot be proven at this time with available archaeological evidence.”



Yes. I would say most of this horizon ended up speaking Iranian with Indo-Aryans in the SE. This would explain how the Yaz cultures and other Central Asian cultures ended up speaking Iranian.
I am not aware of any data strongly indicating that the Yaz culture was associated with the Proto-Iranians. It is sometimes suggested that it could have been associated with a subgroup of early Eastern-Iranians but this is based solely on the fact that this culture lacks any burials which might indicate a Zoroastrian-like funeral practice. However, this is definitely not enough to derive all Iranian speakers (or even all Eastern-Iranian speakers for that matter) from this single culture.



It took place when one group (which I think was still speaking Indo-Dardic-Nuristani) ended up absorbing elements of the BMAC. That is why I think a split in Uzbekistan is more likely than one on the steppe. They migrated SE and are represented by the SWAT and Gandhara cultures.
This would indicate that you date Proto-Indo-Iranian to about 1700 BC (or 1900-1500 BC), which is far too late based on the linguistic data. Where (and when) would you place the common ancestors of the Mitanni-Aryans and the Indo-Aryans from India?

I am of course not absolutely sure that my alternative scenario regarding the Proto-Indo-Iranians is correct, but it seems that it fits both the linguistic and genetic data much better than the traditional Andronovo hypothesis. Importantly, this new hypothesis can be quite easily verified, as it implies that Afanasevo should include mostly R1a-Z94* and/or R1a-Z2124 people, while Sintashta-Petrovka and some core Andronovo sites should be dominated by another subclade of Z94 known as R1a-L657 (though not necessarily L657+, as this branch is characterized by at least eight mutations that are common to all known sublineages). It is worth noting that the apparent difference in the expansion age of Z2124 and L657 corresponds quite well to the age of Afanasevo and Andronovo, respectively.

Michał
12-04-2013, 01:19 PM
The Turks and Huns are mentioned in the Conversion of Kartli

But this is related to a period that postdates the Alexander’s times by more than 600 years, so suggesting that he could have met the Turks in the Caucasus region seems to be a pure confabulation. :)

newtoboard
12-04-2013, 01:25 PM
They speak a Dravidian language, but genetically they are very Western Pakistani. Maltese speak a Semitic language, but genetically they are very South Italian. Hungarians speak an Uralic language, but genetically they are very Central European. Language is not race.

And this conflicts with what I posted how?

newtoboard
12-04-2013, 01:31 PM
OK, I see what you are saying. when I said it was a geographic split I meant that satem speaking people had to have physical distance from centum speaking people for the two to diverge, whereby the eventual satum people picked up a change and other centum people didn't.

Clearly IE people are associated with r1a and r1b. The point I make is that the linguistic connection r1b people have to each other, in the IE world, goes back to the early history of IE which is before the system shift. The linguistic connection r1a people have with each other in the IE world goes back to the time and place where satem evolved.

The idea here is that it is much more likely that pre satem people, who happened to be carrying r1a, shifted their language to IE rather than pre Anatolian, pre Tocharian, and pre Italo-celtic (all pre satem) people all carrying r1b by chance and then each of them shifted their language to IE.

So when people associate r1a and r1b with IE, in essence, they should also associate r1b with early IE and r1a with later (Satem) IE.

Satem dialects didn't evolve out of some sort of western Centum dialects. PIE is not Centum or Satem. Pre Anatolian is pre Centum as well. And Centumization is a change as well. This is just one of many isoglasses anyway. And there are Satem languages whose carriers don't have R1a. But this is just one of those posts which is all about only one haplogroup being related to PIE and Satem dialects evolving out of a vaccum. As if those dialects only came into existence with Satemization.

Michał
12-04-2013, 01:43 PM
R1a is a strong signature among the descendants of the c. 2500 BC PIE rump, taken overall.
I assume you still correlate Corded Ware with R1a, though they expanded about 3000-2800 BC, thus probably earlier than some hypothetical R1b-rich Centum-speaking ppopulations (like the ancestors of Greeks and Armenians). If so, what language was spoken by the early Corded Ware people? Is there any evidence of significant infiltration of the Late Yamna people into the nearly entire CW-occupied territory, or at least an evidence for some significant changes in the archaeological culture that would have spread from this particular location to Lithuania, Northern Belarus and Russia, thus correlating with this hypothetical language shift? Do you consider Germanic as Satem or Centum language? Do you connect pre-Germanic with the process of IE-ization (or “Late PIE-ization”) of Corded Ware, similarly to what David Anthony seems to suggest?

Jean M
12-04-2013, 01:48 PM
Satem dialects didn't evolve out of some sort of western Centum dialects. PIE is not Centum or Satem.

Let us be very clear. PIE evolved. It went through three stages. To use David Anthony's terminology:


Archaic PIE from which the Anatolian branch sprang. Neither Centum nor Satem.
Early PIE evolved from Archaic PIE. Centum. Celtic, Italic etc sprang from this form.
Late PIE evolved from Early PIE. Satem. Baltic, Slavic, Indic, Iranian sprang from this form.


The Centum-Satem isogloss was a sound change within PIE. Centum changed to Satem. It is not that Satem evolved from some particular Centum language. The change happened within the PIE rump left behind in the PIE homeland region after groups departed which would one day develop the Antolian and Centum language branches.

Michał
12-04-2013, 01:48 PM
The lack of M73 there is indicative that Kushans and Tocharians did not have M73 imo.
Please do not confuse the evidently Iranian(Saka)-speaking Tocharians/Kushans (who were mentioned by the ancient Greek sources) with the unfortunately chosen name for the Agnean/Turfanian and Kuchean languages (commonly/”wrongly” reported as Tocharian A and B ). Based on the analysis of different ancient sources (mostly written by some Greek and Chinese authors), it is quite evident that those ancient “Tochari” from Bactria were just one of many Saka tribes, probably derived from the large Yuezhi tribe that originated in the Chinese provinces Gansu and Xinjiang (including the territories where the Tarim mummies were found). All this clearly indicates that the ancient R1a people occupying the Tarim and Dzungarian basins before 100 BC (thus including all those European-like Tarim mummies) were simply some Iranian speaking Sakas (or their direct ancestors, if some very early dates are considered). This is of course fully consistent with what Mallory wrote about the possibility that the Tarim mummies are associated with the evident Iranian presence in the Tarim basin.

The only subclade of R1a that is relatively common in India and could have been attributed to the Kushans (i.e. to the early R1a people from China) is Z2123, a subclade of Z2124, which is of course perfectly consistent with my alternative scenario presented above.

newtoboard
12-04-2013, 01:55 PM
How do you know they did not pick up the language of the Bactrians on the journey west? That seems very likely to me. Z2123+ has such a wide range I doubt we can say it is associated with only R1a people from China. I still disagree with L657+ representing Andronvo when there is not much L657+ in South Asia. If anything it would be more appropriate to represent L657+ with the first migrations out of Central Asia into West Asia and South Asia and Z2124+ and downstream clades with what was left in Central Asia.

Also I still thick Timber Grave spoke Iranian. Abashevo was supposed to have played a large part in Timber Grave. And Cimmerians are often associated with speaking Iranian languages.

newtoboard
12-04-2013, 02:00 PM
Let us be very clear. PIE evolved. It went through three stages. To use David Anthony's terminology:


Archaic PIE from which the Anatolian branch sprang. Neither Centum nor Satem.
Early PIE evolved from Archaic PIE. Centum. Celtic, Italic etc sprang from this form.
Late PIE evolved from Early PIE. Satem. Baltic, Slavic, Indic, Iranian sprang from this form.


The Centrum-Satem isogloss was a sound change within PIE. Centum changed to Satem. It is not that Satem evolved from some particular Centum language. The change happened within the PIE rump left behind in the PIE homeland region after groups departed which would one day develop the Antolian and Centum language branches.

Exactly. Although I would also consider the idea that satemization occurred in Indo-Iranian and spread to Balto-Slavic rather than them being descended from some common late PIE dialect. This makes more sense than saying R1a had no part in the earliest PIE because some Satem languages are correlated to R1a today [(despite Satemization being an areal feature) and many disagreements on how close Balto-Slavic even is to Indo-Iranian] only became IE speaking because of some population replacement naudha thinking is related to a replacement of R1b by R1a despite no one having ever having said this or no support for this. We are going back to the old forest steppe R1a , steppe R1b scenario which is wrong.

Jean M
12-04-2013, 02:10 PM
I assume you still correlate Corded Ware with R1a, though they expanded about 3000-2800 BC, ... If so, what language was spoken by the early Corded Ware people? Is there any evidence of significant infiltration of the Late Yamna people into the nearly entire CW-occupied territory, or at least an evidence for some significant changes in the archaeological culture that would have spread from this particular location to Lithuania, Northern Belarus and Russia, thus correlating with this hypothetical language shift? Do you consider Germanic as Satem or Centum language? Do you connect pre-Germanic with the process of IE-ization (or “Late PIE-ization”) of Corded Ware, similarly to what David Anthony seems to suggest?


.. it absolutely leaps to the eye that


R1a is a strong signature among the descendants of the c. 2500 BC PIE rump, taken overall.
R1b is predominant among at least the Celtic and Italic group of departures prior to Satemisation. Germanic is mixed R1a and R1b. ...


I do indeed feel that R1a spread via Corded Ware. CW has long been seen as "Kurganised" i.e. influenced by Yamnaya elites, rather than representing a complete population replacement. Certainly it looks from genetic and linguistic evidence (Y-DNA I1 and substrate in Germanic) that there was no wipe-out in the north, but a substantial inward Copper Age flow nonetheless. Here is how I phrased the problem in AJ, pp. 131-2:


The new lifestyle travelled along multiple routes. The next movement visible in the archaeology flowed to the western end of the steppes, integrating the lowland steppe and upland farming communities of Late Cucuteni-Tripolye origin in the Usatovo culture around the mouth of the river Dniester. David Anthony has argued the case for this culture as the first link in a long chain leading to the Pre-Proto-Germanic dialect splitting away, the next link being migration up the Dniester through Late Cucuteni-Tripolye territory into the widespread Corded Ware culture (2750–2400 BC). Yet Proto-Germanic did not develop until about 500 BC. So it might be more helpful to visualise migration up the Dniester as part of the spread of a dialect which would have been intelligible to Indo-Europeans across a broad expanse of Bronze Age Europe.*

The orthodox view of the Corded Ware culture as native to the North European Plain once seemed supported by a local pottery sequence. The first pottery was the pointed-bottom and everted lip type of the hunter-gatherers. Then we have the Funnel Beaker (TRB ) type with everted lip, followed by Corded Ware with everted lip. This vision of continuity has been overturned. Ancient DNA shows clearly that the TRB people did not descend from the hunter-gatherers who made the previous type of pottery.

Nor is the Corded Ware culture, with its influences from Yamnaya, a straightforward cultural descendant of the TRB. Archaeologically it seems to be the result of people moving up the rivers Prut, Dnieper and Dniester from the steppe and blending with previous peoples of the North European Plain. What does that mean in human terms? The TRB had already adopted much of the technology of the Secondary Products Revolution and appeared to be thriving on it c. 3400 BC. There was a diversity of burial rites, but the creation of large causewayed enclosures speaks of a society well able to cooperate. Then there are signs of a population decrease from about 3350 BC. New types of burial custom appear in Funnel Beaker (TRB ) sites after about 3000 BC. The dead are mainly buried collectively, but individual burials with weapons appear. These typical warrior burials suggest conflict appearing in this society. That could be the result of internal pressure over scarce resources, given the climate change mentioned earlier. Then the population rose again after about 2900 BC, which probably indicates the arrival of the people later identified by their characteristic Corded Ware pottery, with its own warrior burials.

This complex picture presents an interesting challenge for genetics. So far we have mtDNA from one TRB site in Sweden (H, J and T2b). We have Y-DNA R1a1 from one Corded Ware site in Germany, together with a wider mixture of mtDNA haplogroups (H, I, K1a2, K1b, U5b and X2). These mtDNA samples are too small for meaningful comparison. So what is needed is more aDNA from a range of sites of these cultures, including Y-DNA. One scrap of evidence though is significant. Scientists managed to extract a large part of the nuclear genome of the TRB farmer carrying mtDNA H. It was more similar to modern-day southern Europeans and Anatolians than Scandinavians. If this pattern holds good for other samples, modern Scandinavians are not the undiluted descendants of TRB farmers.

* The former CW regions have some of the "Old European hydronymy" thought to be from PIE or close to it, before the development of specific languages branches.

Michał
12-04-2013, 02:45 PM
How do you know they did not pick up the language of the Bactrians on the journey west?
I consider it very unlikely because there were too many of them and they simply had not enough time to switch to a new language (it all happened relatively quickly).


Z2123+ has such a wide range I doubt we can say it is associated with only R1a people from China.
Exactly! So how it was possible that the original Tocharian-speaking people were Z2123+ (just like the Sardinians and Karachays from Europe)?

This can be relatively easily explained when assuming that Z2123 was associated with a significant proportion of the post-Afansevo Iranian-speaking people, including at least the Indo-Scythians and some North-Eastern Iranian tribes that migrated to Europe (like Sarmatians).



I still disagree with L657+ representing Andronvo when there is not much L657+ in South Asia.
Not much L657 in South Asia??? There is more L657 in India and Pakistan than Z280 or M458 in the entire world. Can you show me another subclade of R1a that is more frequent in South Asia?



Also I still thick Timber Grave spoke Iranian.
How can you be so inconsistent? If the Timber Grave culture in Europe is dated to about 1800-1200 BC, while you believe that the Proto-Indo-Iranian was spoken by the Andronovo people (1900-1500 BC) how can you claim that the Timber Grave people spoke Iranian?

EDIT (some personal comments removed)

parasar
12-04-2013, 03:00 PM
But this is related to a period that postdates the Alexander’s times by more than 600 years, so suggesting that he could have met the Turks in the Caucasus region seems to be a pure confabulation. :)

Josephus in the first century called them Alans, so my guess is the names changed as the 'barbarians' changed. I think the generic name would be Scythian, ie, Alexander built (or reinforced if they were there before) to keep out the the Scythians.




... Alans (Josephus 37–100 AD http://books.google.com/books?id=4tsKAQAAIAAJ&pg=PA151)


Josephus:

Now the nation of the Alani who (as I said before somewhere) are Scythians, and dwell near the Tanais ' and the lake Mseotis,2 formed about this time the design of invading Media and the country beyond it, in order to plunder them. Accordingly they negotiated with the king of the Hyrcani,3 for he was master of the pass that king Alexander had shut up with iron gates. And as he gave them leave to come through them, they came en masse, and fell upon the Medes who expected nothing of the kind, and plundered their country, which was populous and rich in cattle, and nobody durst make any resistance against them. For Pacorus, the king of the country, had fled away for fear into places where they could not easily get at him, and had yielded up every thing he had to them, and had only with difficulty saved his wife and concubines from them, after they had been made captives, by paying a hundred talents for their ransom. These Alani therefore plundered the country with great ease and without opposition, and proceeded as far as Armenia, laying all waste before them. Now Tiridates was king of that country, and he met them and fought them, but was nearly taken alive in the battle; for a certain man threw a net over him from some distance, and would soon have drawn him to him, if he had not quickly cut the cord with his sword, and fled first. And the Alani, being still more provoked by this fight, laid waste the country, and carried off with them a great multitude of men, and a great quantity of other spoil they had got from both kingdoms, and then retired back to their own country.

parasar
12-04-2013, 03:15 PM
Not much L657 in South Asia??? There is more L657 in India and Pakistan than Z280 or M458 in the entire world. Can you show me another subclade of R1a that is more frequent in South Asia?
..
:)
True. There are probably 20 million L657+ males in my state of origin. L657 is also the most frequent R1a1 clade in Arabia and likely in (eastern) Iran.

http://upload.wikimedia.org/wikipedia/commons/thumb/e/e0/India_Bihar_locator_map.svg/543px-India_Bihar_locator_map.svg.png

Michał
12-04-2013, 04:00 PM
I do indeed feel that R1a spread via Corded Ware. CW has long been seen as "Kurganised" i.e. influenced by Yamnaya elites, rather than representing a complete population replacement. Certainly it looks from genetic and linguistic evidence (Y-DNA I1 and substrate in Germanic) that there was no wipe-out in the north, but a substantial inward Copper Age flow nonetheless. Here is how I phrased the problem in AJ, pp. 131-2:
I am not sure if I correctly interpret your answers to my specific questions. For example, it is still not clear to me what language was spoken by the original Corded Ware population in your opinion. Since they have expanded over much of Europe before your assumed Centum to Satem transformation within PIE took place, I suspect you think they originally spoke a Centum language. Since you have also listed Germanic as a Centum language, the Satemization of the post-CW R1a people could not have been related to some secondary influence from the migrating pre-Germanic population. Thus, I still find your scenario somehow inconsistent (or at least not clear enough).

Additionally, I don’t know your answer to the Graeco-Armenian question in relation to Late PIE. How would you date the separation of this branch from the PIE dialect continuum and where would you exactly place its origin?

Jean M
12-04-2013, 05:52 PM
I am not sure if I correctly interpret your answers to my specific questions. For example, it is still not clear to me what language was spoken by the original Corded Ware population in your opinion.

We can't be any more precise than assuming that it was a early IE dialect, or range of dialects. The only clue is the one I gave you in a footnote - some of the hydronyms which appear to pre-date any of the known IE branches. CW is a puzzle every bit as complex as the Bell Beaker story, and with far less to go on in the way of written/inscribed sources.

Germanic, which is accounted a Centum language, looks like it could be a very late descendant of some CW dialect or mixture of same. It does not necessarily follow that every part of the CW territory was occupied by people speaking a Centum version of PIE for the whole CW period. As Andersen sees it, the forests of north-eastern Europe were penetrated time after time by small groups of Indo-Europeans, whose descendants were absorbed linguistically by the succeeding wave. In other words, Centum arrivals could be succeeded by Satem PIE arrivals, succeeded by speakers closer to Proto-Balto-Slavic, etc. Then we have people pushing down from Scandinavia as it got too cold to farm there in the Late Bronze Age, bringing with them a descendant of Centum CW, which developed into Proto-Germanic c. 500 BC in contact with Celtic and Baltic.

newtoboard
12-04-2013, 05:58 PM
I consider it very unlikely because there were too many of them and they simply had not enough time to switch to a new language (it all happened relatively quickly).


Exactly! So how it was possible that the original Tocharian-speaking people were Z2123+ (just like the Sardinians and Karachays from Europe)?

This can be relatively easily explained when assuming that Z2123 was associated with a significant proportion of the post-Afansevo Iranian-speaking people, including at least the Indo-Scythians and some North-Eastern Iranian tribes that migrated to Europe (like Sarmatians).



Not much L657 in South Asia??? There is more L657 in India and Pakistan than Z280 or M458 in the entire world. Can you show me another subclade of R1a that is more frequent in South Asia?



How can you be so inconsistent? If the Timber Grave culture in Europe is dated to about 1800-1200 BC, while you believe that the Proto-Indo-Iranian was spoken by the Andronovo people (1900-1500 BC) how can you claim that the Timber Grave people spoke Iranian?

I am sorry to say it but you seem to show a really deadly mixture of” R1b-phobia” and “Iranophilia”. How can you live with that? :)


I meant to say there is not much L657+ in Central Asia so it doesn't make sense to see Andronovo as being represented by L657

How is this inconsistent? I said originally Andronovo is originally represented by Proto Indo-Iranian. That doesn't mean it was speaking Proto Indo-Iranian throughout its entire lifespan. How do you know it didn't break up into Indo-Aryan and Indo-Iranian between 1900 BC and 1800 BC? I am still not convinced the descendants of Timber Grave are Cimmerians. If I recall there were other cultures in Western Ukraine that were mentioned in Mallory's writings of them and I believe they were West of the Dnieper. Either way if Mallory among others entertains the theory that they were possibly speaking NE Iranian what makes you so sure they weren't? Either way there is no reason to connect Timber Grave with R1b if that is your point when it is the descendant culture of Eastern Yamanaya and Poltavka. That and its connections with both Abashevo and Andronovo are a good bet it was ancestral to some sort of Indo-Iranian population.

I exhibit none of those things you mentioned. That is a pretty serious bias to accuse someone of. You are the one who cringes when anybody even mentions R1a and PIE. I could easily accuse you of being biased because your nationality might make you want R1a native to Corded Ware and not from Yamnaya. Funny you accuse me of Iranophilia (even when I have been one of the few on these forums who has never been convinced of inflated Northern European scores in NW South Asia relating to Scythians where others have argued for that) when the theory of a proto Satem R1a group is probably endorsed only by modern believers of Sarmatism. But I didn't go down that route. I am sorry that you did.

Jean M
12-04-2013, 06:06 PM
Additionally, I don’t know your answer to the Graeco-Armenian question in relation to Late PIE. How would you date the separation of this branch from the PIE dialect continuum and where would you exactly place its origin?

As shown on the diagram from Warnow et al 2003 in the first post on this thread, linguistically the departure of people whose descendants would eventually speak Greek and Armenian can be placed very roughly around 3000 BC. Greek is Centum. Armenian has absorbed so much vocabulary from Iranian that it confused linguists, but they now seem happy to recognise that it sprang from the same branch-line as Greek, along with the related (now defunct) Phrygian and Thracian. A homeland for what is often called the Balkan group of languages can be dimly perceived in Thrace and Greek Macedonia. So the parent of Greek and Armenian could have arrived in Thrace as part of the same southern movement that created the Coţofeni culture, sometimes seen as generating the ancestor of Thracian. http://en.wikipedia.org/wiki/Co%C8%9Bofeni_culture

newtoboard
12-04-2013, 06:11 PM
As shown on the diagram from Warnow et al 2003 in the first post on this thread, linguistically the departure of people whose descendants would eventually speak Greek and Armenian can be placed very roughly around 3000 BC. Greek is Centum. Armenian has absorbed so much vocabulary from Iranian that it confused linguists, but they now seem happy to recognise that it sprang from the same branch-line as Greek, along with the related (now defunct) Phrygian and Thracian. A homeland for what is often called the Balkan group of languages can be dimly perceived in Thrace and Greek Macedonia. So the parent of Greek and Armenian could have arrived in Thrace as part of the same southern movement that created the Coţofeni culture, sometimes seen as generating the ancestor of Thracian. http://en.wikipedia.org/wiki/Co%C8%9Bofeni_culture

This Dienekes post seems relevant:

http://dienekes.blogspot.com/2013/08/the-place-of-armenian-in-indo-european.html

parasar
12-04-2013, 06:26 PM
I meant to say there is not much L657+ in Central Asia so it doesn't make sense to see Andronovo as being represented by L657


...

Now what you are saying does make sense. It is based on the L657 origin timelines (and some sketchy STRs we have from Andronovo) that I believe that Andronovo has nothing to do with L657.

Michał tried to make a distinction between true L657+ and those in the umbrella L657 category based on the eight SNPs at that level, but we have to wait and see if that distinction has any validity.

Silesian
12-04-2013, 06:31 PM
As shown on the diagram from Warnow et al 2003 in the first post on this thread, linguistically the departure of people whose descendants would eventually speak Greek and Armenian can be placed very roughly around 3000 BC. Greek is Centum. Armenian has absorbed so much vocabulary from Iranian that it confused linguists, but they now seem happy to recognise that it sprang from the same branch-line as Greek, along with the related (now defunct) Phrygian and Thracian. A homeland for what is often called the Balkan group of languages can be dimly perceived in Thrace and Greek Macedonia. So the parent of Greek and Armenian could have arrived in Thrace as part of the same southern movement that created the Coţofeni culture, sometimes seen as generating the ancestor of Thracian. http://en.wikipedia.org/wiki/Co%C8%9Bofeni_culture

Areas where R1a is so low; single digit, to non-existent, like in Anatolian/Hittite/Luwian speaking areas, ancient Armenian speaking areas, and Greek speaking areas ,ancient Albanian speaking areas. However they are all connected with Coţofeni culture and PIE? It makes sense to shy away from any association with a species of ydna and language given the low to non existent numbers.

Silesian
12-04-2013, 06:40 PM
This Dienekes post seems relevant:

http://dienekes.blogspot.com/2013/08/the-place-of-armenian-in-indo-european.html
If no chronological dated ydna/PiE tree can be provided; with such relevant information maybe you can nail the species of R1a involved in this movement/exchange?
L657 / 584/ 93_94/ 283?

newtoboard
12-04-2013, 06:54 PM
If no chronological dated ydna/PiE tree can be provided; with such relevant information maybe you can nail the species of R1a involved in this movement/exchange?
L657 / 584/ 93_94/ 283?

If R1a was involved then obviously Z93*/Z93(xZ94). There is some Z93+ among Greeks and Armenians and much of it is Z93(xZ94) which is not very common among modern Indo-Iranians.

newtoboard
12-04-2013, 06:57 PM
Now what you are saying does make sense. It is based on the L657 origin timelines (and some sketchy STRs we have from Andronovo) that I believe that Andronovo has nothing to do with L657.

Michał tried to make a distinction between true L657+ and those in the umbrella L657 category based on the eight SNPs at that level, but we have to wait and see if that distinction has any validity.

Doesn't L657+ distribution within Iran, Afghanistan and South Asia seem very coastal and Southern? It is quite possible L657+ originated in NW South Asia or regions directly to the west.

Jean M
12-04-2013, 06:57 PM
Areas where R1a is so low; single digit, to non-existent, like in Anatolian/Hittite/Luwian speaking areas, ancient Armenian speaking areas, and Greek speaking areas ,ancient Albanian speaking areas. However they are all connected with Coţofeni culture and PIE?

The Anatolian branch (Hittite, Luwian etc) has not been associated with the Coţofeni culture. David Anthony sees the departure of the ancestor of Anatolian expressed archaeologically by herder settlements of the Suvorovo group appearing in the Danube valley about 4200 BC. Anthony suggests that groups from this culture entered Anatolia around 3000 BC, founding Troy.

As I keep on saying, both R1b and R1a appear to be involved in the spread of IE languages. So why is specifically low R1a a problem? Greek and Armenian speakers are linked by R1b-L23, but also carry a range of other haplogroups probably absorbed from descendants of Neolithic farmers in Thrace and then in the places they entered - Greece, Anatolia, the Armenian Plain. We only see overwhelming dominance of R1b/R1a in IE speaking regions which offered greater opportunities for expansion of the incoming population, mainly because of low levels of existing population. Or so I deduce.

newtoboard
12-04-2013, 06:58 PM
Areas where R1a is so low; single digit, to non-existent, like in Anatolian/Hittite/Luwian speaking areas, ancient Armenian speaking areas, and Greek speaking areas ,ancient Albanian speaking areas. However they are all connected with Coţofeni culture and PIE? It makes sense to shy away from any association with a species of ydna and language given the low to non existent numbers.

Jean never said any of that. Why are you putting words in her mouth and then attacking straw man arguments as if they somehow prove only R1b can be associated with PIE?

Silesian
12-04-2013, 07:18 PM
The Anatolian branch (Hittite, Luwian etc) has not been associated with the Coţofeni culture. David Anthony sees the departure of the ancestor of Anatolian expressed archaeologically by herder settlements of the Suvorovo group appearing in the Danube valley about 4200 BC. Anthony suggests that groups from this culture entered Anatolia around 3000 BC, founding Troy.

As I keep on saying, both R1b and R1a appear to be involved in the spread of IE languages. So why is low R1a a problem specifically? Greek and Armenian speakers are linked by R1b-L23, but also carry a range of other haplogroups probably absorbed from descendants of Neolithic farmers in Thrace and then in the places they entered - Greece, Anatolia, the Armenian Plain. We only see overwhelming dominance of R1b/R1a in IE speaking regions which offered greater opportunities for expansion of the incoming population, mainly because of low levels of existing population.

Armenian, Greek, Albanian,Mede are some of the closest languages surrounding Hittite and Luwian[centum_satem]. 3 of these languages are still spoken, one is considered a substrate. Of the three languages still used they are classified as separate branches. It would be reasonable to conclude that of the three remaining languages in use should have a common denominator if not with Hittite/Luwian than at lest with each other if they have a common source. If that source is R1a is the species Z93?

Silesian
12-04-2013, 07:22 PM
If R1a was involved then obviously Z93*/Z93(xZ94). There is some Z93+ among Greeks and Armenians and much of it is Z93(xZ94) which is not very common among modern Indo-Iranians.
Where is the source of Z93 found in Greeks and Armenians and is it found in Albanians?

newtoboard
12-04-2013, 07:24 PM
The Anatolian branch (Hittite, Luwian etc) has not been associated with the Coţofeni culture. David Anthony sees the departure of the ancestor of Anatolian expressed archaeologically by herder settlements of the Suvorovo group appearing in the Danube valley about 4200 BC. Anthony suggests that groups from this culture entered Anatolia around 3000 BC, founding Troy.

As I keep on saying, both R1b and R1a appear to be involved in the spread of IE languages. So why is specifically low R1a a problem? Greek and Armenian speakers are linked by R1b-L23, but also carry a range of other haplogroups probably absorbed from descendants of Neolithic farmers in Thrace and then in the places they entered - Greece, Anatolia, the Armenian Plain. We only see overwhelming dominance of R1b/R1a in IE speaking regions which offered greater opportunities for expansion of the incoming population, mainly because of low levels of existing population. Or so I deduce.

Armenian Plain? I assume you meant to type Highland. :)

newtoboard
12-04-2013, 07:26 PM
Where is the source of Z93 found in Greeks and Armenians and is it found in Albanians?

Why are Albanians relevant? Most linguists notice some common features between Greek, Armenian and Indo-Iranian. Albanian is not mentioned. I simply suggested is is possible that Z93+ (as a minority lineage since most Greeks and Armenians can be represented by R1b-L23+) is the lineage responsible for those similarities. Much like Z284+ might account for similarities between Balto-Slavic and Germanic even though Germanic is better represented by R1b-U106+.

Silesian
12-04-2013, 07:28 PM
Jean never said any of that. Why are you putting words in her mouth and then attacking straw man arguments as if they somehow prove only R1b can be associated with PIE?

I did not mean to attack with a straw man, sorry if it came across that way. I want to better understand the spread of this group of languages including Albanian. I have put forth R1a was just as important to the spread of PIE and R1b, R1a is the second most common paternal marker in my [email protected] cluster. Michael has a great knowledge and put forth his ideas, I would like to get your take on the spread of PIE together with R1a/R1b, even if you do not want to make the same branching map.

Silesian
12-04-2013, 07:28 PM
Why are Albanians relevant? Most linguists notice some common features between Greek, Armenian and Indo-Iranian. Albanian is not mentioned. I simply suggested is is possible that Z93+ (as a minority lineage since most Greeks and Armenians can be represented by R1b-L23+) is the lineage responsible for those similarities. Much like Z284+ might account for similarities between Balto-Slavic and Germanic even though Germanic is better represented by R1b-U106+.
If you do not want to use them that is fine.

Jean M
12-04-2013, 07:33 PM
It would be reasonable to conclude that of the three remaining languages in use should have a common denominator.

The common denominator is descent from PIE. The position of Albanian in relation to other IE branches is uncertain.

It does not matter if a language today (say Spanish) is spoken right next door to another language (say Quechua) in the same town. That does not make them descended from the same language family, let alone the same branch. The relationships between languages are worked out by linguists from (surprise!) linguistic evidence. ;) Not where people live today. And definitely not from what haplogroups they carry.

alan
12-04-2013, 08:29 PM
I have often wonder if anyone has speculated about the cause of the initial satem speakers. Was it a substrate effect? If so then that might give hints as to where it might have occurred.


Satem dialects didn't evolve out of some sort of western Centum dialects. PIE is not Centum or Satem. Pre Anatolian is pre Centum as well. And Centumization is a change as well. This is just one of many isoglasses anyway. And there are Satem languages whose carriers don't have R1a. But this is just one of those posts which is all about only one haplogroup being related to PIE and Satem dialects evolving out of a vaccum. As if those dialects only came into existence with Satemization.

newtoboard
12-04-2013, 08:34 PM
I have often wonder if anyone has speculated about the cause of the initial satem speakers. Was it a substrate effect? If so then that might give hints as to where it might have occurred.


Maybe but I don't see the value in this sort of speculation since almost nothing can ever confirm it. Even some modern Western European languages have undergone partial Satemization.

We could ask the same thing for Centum. That is a process as well.

TigerMW
12-04-2013, 10:06 PM
... We could ask the same thing for Centum. That is a process as well.

What do you mean? I thought that there was no Centumization process. Centum is like a paragroup in the IE family. They are just languages that haven't been Satemized. Is there something more that I don't understand?


[[[Mikeww/moderator: BTW, as a moderator, I'm going to start deleting devoid of content replies in this thread. This is nothing more than implementation of forum policies. I've received a lot of reports related to this thread. This will probably make some folks unhappy, but someone has to make some judgements. I've also got multiple warnings out by PMs on personal criticism, devoid of content posting, etc. Compliments are great and that is what the "thanks" button is for. Please try to focus on the content and logic of the topic. Please try to add new evidence or logic with postings. Disagreement and debate is great! ... however, one-line commentaries and ad hominen innuendo are not helpful. ]]]

Jean M
12-04-2013, 10:36 PM
What do you mean? I thought that there was no Centumization process. Centum is like a paragroup in the IE family. They are just languages that haven't been Satemized. Is there something more that I don't understand?


I made a couple of posts on this. Here's the first: http://www.anthrogenica.com/showthread.php?1519-Languages-and-Y-DNA-lineages&p=21836&viewfull=1#post21836



c. 4200 BC - The earliest form of PIE had neither the centum or satem sound. It formed the basis for the Anatolian branch.
c. 3500 BC - The Centum sound had developed. Centumization removed the palatovelars from the language. It did not matter which direction people went from the steppe. They took with them this form of PIE. It formed the basis for the Tocharian, Celtic, Italic, Germanic and Greek branches.
c. 2500 BC - The Satem sound developed. This formed the basis for the Baltic, Slavic, Indic and Iranian branches. Again it did not matter in which direction people went. They took this sound.



I clarified in this post http://www.anthrogenica.com/showthread.php?1519-Languages-and-Y-DNA-lineages&p=21945&viewfull=1#post21945


Let us be very clear. PIE evolved. It went through three stages. To use David Anthony's terminology:


Archaic PIE from which the Anatolian branch sprang. Neither Centum nor Satem.
Early PIE evolved from Archaic PIE. Centum. Celtic, Italic etc sprang from this form.
Late PIE evolved from Early PIE. Satem. Baltic, Slavic, Indic, Iranian sprang from this form.

parasar
12-04-2013, 10:56 PM
Doesn't L657+ distribution within Iran, Afghanistan and South Asia seem very coastal and Southern? It is quite possible L657+ originated in NW South Asia or regions directly to the west.

Possible, but I am not sure. Z93, Z94, Z2124, and Z2123 perhaps, in the Northwest. These may be Sanskritic Indus lines.

I can't make out L657 at all - it is perhaps a Prakrit line. My closest matches are from Andhra, Lanka, and Gujarat in that order. Once we get more sampling form UP and Bihar, I may get closer matches from those areas. A coastal spread from India looks likely, though I would not be surprised of sporadic finds in eastern Asia along Buddhist trails.

Many years back I had tracked a rajput brahman line from Lanka - Gautam if I recall correctly. We are often called Gautam Babhan. While Gautam is indeed a Brahman line, some say the moniker may be due to adopting the Buddha's precepts. These Gautams had come over from Lanka to Kannauj and were related to the Gaharwar rajput, the Vijayanagar folk, as well had some Gangaikonda Chola connections. I will need to revisit my notes.

bolek
12-04-2013, 11:26 PM
What do you mean? I thought that there was no Centumization process. Centum is like a paragroup in the IE family. They are just languages that haven't been Satemized. Is there something more that I don't understand?



Centum and satem division has no role in the history of PIE language development. There were never satem and centum languages, one of them older and the other younger. It is nonsense from XIX century. In all languages ‘s’ and ‘k’ interchange all the time and it was so in PIE as well. In one dialect of the same language you can find ‘s‘ and in another dialect of the same language it is ‘k’ in the same word. You can see it not only in Slavic, Baltic, Indo-Iranian, but also in Romance, Greek, Germanic and others.

Satem centum issoglos is now no longer taken to reflect a fundamental division in descent. There was no satem and centum split in PIE. PIE was neither centum nor satem.

Greek, a "centum" language, is more closely related to Sanskrit, a “satem” language, than to the others so called “centum” languages.
Tocharian has more in common with Slavic languages than with Celtic or Germanic.

Centumisation is a process where original ‘s/z’ –>’h/k/g’, for example original Sanskrit ‘Sarasvati’ becomes ‘Harasvati’ in Iranian. Centumisation and satemisation processes happen all the time.

The terms “centum” and “satem” remain a useful descriptive shorthand and nothing more.

nuadha
12-05-2013, 01:06 AM
Satem dialects didn't evolve out of some sort of western Centum dialects. PIE is not Centum or Satem. Pre Anatolian is pre Centum as well. And Centumization is a change as well. This is just one of many isoglasses anyway. And there are Satem languages whose carriers don't have R1a. But this is just one of those posts which is all about only one haplogroup being related to PIE and Satem dialects evolving out of a vaccum. As if those dialects only came into existence with Satemization.

when did I say any of these things?!? and you've clearly missed the point.

what matters is the chronological order of the branchings. I don't care if you want to call satem arf, centum borg, and pre centum PIE grrrrr. You could even call pre Tocharian grrrreeennnborg for all I care. The point is that the early branchings of IE - which includes pre Anatolian, pre Tocharian, and pre italo Celtic - imply that at least 3 groups whose last IE connection was early IE. So it is reasonable to guess that their r1b commonality, whereby r1b (m269 + m73) is itself IE related, goes back to early IE.

In the case of satem, we have the various descends who share a connection to the first speakers of satem which was later in IE history. Their descendants also tend to share r1a. The correlation probably means that the early speakers of satem probably had a good amount of r1a. That population, also has a connection to (early) PIE but they are separated from the early PIE population by about 1000 years and probably by some geographic distance too. The temporal and spacial distance between the two makes a genetic difference very plausible. But to say anything similar about r1b amounts to saying that three early IE groups leaving each other in opposite directions somehow all absorbed r1b (m269 and m73) from non IE people even though r1b (m269 + m73) tends to be more rare outside the historical IE world. Then, on top of that you would expect that this pre satem group was the (only?) remaining group which avoided r1b absorbtion.

Far far easier to say early IE was r1b based on the ringe et al phylogeny.

nuadha
12-05-2013, 01:25 AM
The Anatolian branch (Hittite, Luwian etc) has not been associated with the Coţofeni culture. David Anthony sees the departure of the ancestor of Anatolian expressed archaeologically by herder settlements of the Suvorovo group appearing in the Danube valley about 4200 BC. Anthony suggests that groups from this culture entered Anatolia around 3000 BC, founding Troy.

have you made use of the fact that David thinks pre italo celtic spread to central Europe by the yamnaya while he suggest that pre germanic and pre (balto) - Slavic might have spread primarily with the descendants of Cucuteni-Trypillian culture.

could you remind me of what you expect the yamnaya to be in relation to your theory?

Michał
12-05-2013, 02:11 AM
It does not necessarily follow that every part of the CW territory was occupied by people speaking a Centum version of PIE for the whole CW period. As Andersen sees it, the forests of north-eastern Europe were penetrated time after time by small groups of Indo-Europeans, whose descendants were absorbed linguistically by the succeeding wave. In other words, Centum arrivals could be succeeded by Satem PIE arrivals, succeeded by speakers closer to Proto-Balto-Slavic, etc.
Do you mean that all those new waves of Indo-Europeans entering the CW territory were derived from Yamna? Can you say anything more about their identity, time span and geographical distribution? Which particular wave would you associate with the Proto-Balto-Slavs? Honestly, I find this scenario highly unlikely. Based on what we know about CW, its initial expansion was relatively quick and the typical CW package was nearly the same in distantly located regions. Most importantly, it was evidently different from the typical Yamna package (otherwise we would be unable to distinguish these two cultural horizons). Therefore, I don't think it is reasonable to assume that every new wave of Yamna-derived newcomers reaching the CW territory was automatically transformed into the CW-like entity (while producing some substantial language shifts at the same time).


Then we have people pushing down from Scandinavia as it got too cold to farm there in the Late Bronze Age, bringing with them a descendant of Centum CW, which developed into Proto-Germanic c. 500 BC in contact with Celtic and Baltic.
Is this related to your hypothesis that the R1a-Z284 CW-derived people significantly contributed to the development of Proto-Germanic? If so, we have discussed this before and I don't recall having seen any convincing evidence that would support your scenario.
http://www.anthrogenica.com/showthread.php?1027-Correlations-of-various-R1a-and-I-subclades-that-might-link-with-R1b-subclades&p=8038#post8038

Michał
12-05-2013, 02:19 AM
I meant to say there is not much L657+ in Central Asia so it doesn't make sense to see Andronovo as being represented by L657
I guess you are not suggesting that L657 was not derived from Andronovo (as this would contradict your previous assumption that Andronovo was associated with the Proto-Indo-Iranians) but rather that L657 did not expand before reaching South Asia. I think you may be right about it, as it seems still possible that L657 is only 3500 years old, although based on the recent full sequencing results, I would consider it slightly more likely that L657 is significantly older (at least 4000-4500 years old), which would be of course quite consistent with the age of Andronovo. The only way to verify this is to get some SNP-tested ancient samples from Andronovo.

There is indeed not much L657 left within the former Andronovo territory (though there is at least one interesting subcluster of L657 seen in Kazakhstan). On the other hand, when assuming that to completely dominate the very populous Indian subcontinent, the Indo-Aryans needed to be represented by a relatively large number of the Andronovo-derived people, this process could have indeed lead to some significant population movement southward. Also, this is actually not much different from any other ancient steppe culture that was replaced by some newcomers. For example, we have not much Z93 left in Ukraine, although we suspect that this clade was very common there in the past.



How is this inconsistent? I said originally Andronovo is originally represented by Proto Indo-Iranian. That doesn't mean it was speaking Proto Indo-Iranian throughout its entire lifespan. How do you know it didn't break up into Indo-Aryan and Indo-Iranian between 1900 BC and 1800 BC?
I consider it a bit unrealistic to assume that some two relatively well-defined archeological cultures that show parallel development (with no apparent population exchange) and evident continuity over a couple of centuries could be suspected of representing two populations that were involved in some language shift that is not suggested by any specific archeological, linguistic or genetic data.

Most importantly, you have just suggested that the most likely Proto-Iranian culture was Yaz (1300-1000 BC), so how is this consistent with the much older Timber Grave culture (1800-1200 BC) being associated with their descendants who spoke a specific Iranian language?



I am still not convinced the descendants of Timber Grave are Cimmerians. If I recall there were other cultures in Western Ukraine that were mentioned in Mallory's writings of them and I believe they were West of the Dnieper. Either way if Mallory among others entertains the theory that they were possibly speaking NE Iranian what makes you so sure they weren't?
I can perfectly understand if someone suggests that Timber Grave were Iranian-speakers if this is consistent with all the remaining elements of his/her hypothesis. But in your case, you have tried to combine a series of elements that just do not fit each other.


Either way there is no reason to connect Timber Grave with R1b if that is your point when it is the descendant culture of Eastern Yamanaya and Poltavka.
If the major reason to identify Timber Grave with the Iranian-speakers was to prevent other people from associating this culture with R1b, then I rest my case...


That and its connections with both Abashevo and Andronovo are a good bet it was ancestral to some sort of Indo-Iranian population.
You are changing your view regarding the most likely origin of Proto-Indo-Iranians and Proto-Iranians in nearly every post, so I better wait until you rethink this again and come up with some logical and consistent model that can be further discussed.



I exhibit none of those things you mentioned. That is a pretty serious bias to accuse someone of.
This was supposed to be a joke, but indeed a very stupid one, so please accept my apology.

Michał
12-05-2013, 02:22 AM
As shown on the diagram from Warnow et al 2003 in the first post on this thread, linguistically the departure of people whose descendants would eventually speak Greek and Armenian can be placed very roughly around 3000 BC. Greek is Centum. Armenian has absorbed so much vocabulary from Iranian that it confused linguists, but they now seem happy to recognise that it sprang from the same branch-line as Greek, along with the related (now defunct) Phrygian and Thracian. A homeland for what is often called the Balkan group of languages can be dimly perceived in Thrace and Greek Macedonia. So the parent of Greek and Armenian could have arrived in Thrace as part of the same southern movement that created the Coţofeni culture, sometimes seen as generating the ancestor of Thracian. http://en.wikipedia.org/wiki/Co%C8%9Bofeni_culture
The earliest stages of Cotofeni are dated to about 3500-3200, which is about the same age as in the case of Afanasevo (or even slightly earlier). This means that although Afanasevo and Cotofeni represent archaeological cultures that are nearly equally distantly related to some Eastern Yamna-derived cultures like Andronovo, you suggest that the Andronovo and Cotofeni people were speaking some much more closely related languages than the language spoken by the Afanasevo people, and this is despite the fact that Andronovo and Afanasevo shared the same Y-DNA haplogroup (R1a) and were derived from the same Eastern part of the relatively large Yamna horizon. I hope you will understand that I cannot consider this a reasonable hypothesis.

parasar
12-05-2013, 03:52 AM
There is indeed not much L657 left within the former Andronovo territory (though there is at least one interesting subcluster of L657 seen in Kazakhstan) ...

This is a Y7+ line that looks to derive from Arab missionaries (Khojas and Mirzas) sent from Bokhara in the Kuchum Khan period. Babasan yurt (on the Tobol river near Kazakh-Russia border), was an establishment named after a Mirza Babasan.

Jean M
12-05-2013, 11:01 AM
have you made use of the fact that David thinks pre italo celtic spread to central Europe by the yamnaya while he suggest that pre germanic and pre (balto) - Slavic might have spread primarily with the descendants of Cucuteni-Trypillian culture. could you remind me of what you expect the yamnaya to be in relation to your theory?

I don't think that David Anthony specifies the percentages of Cucuteni-Trypillian v Yamnaya input in the Usatovo mix or the communities on the Middle Dnieper, but simply views them as mixed. We can't be specific. I certainly follow that. We can't expect the Cucuteni-Trypillian Y-DNA input to be exactly the same in the various riverine pockets in which it survived, even to start with, and genetic drift could make a difference as time went on. Y-DNA I2a2b and R1b-U106 seem to have moved north from Usatovo with R1a. Y-DNA I2a1b1 (M359.2/P41.2) seems to have moved with R1a1a (notably M458) in the explosion of the Slavs. (I also noted recently that a teeny bit of R1b-L23 could have been a fellow traveller.) However some calculations indicate a Copper or Bronze Age spread of Y-DNA J and E from the Balkans, and I am still unsure whether some of this could have been travelling with some of the more Cucuteni-influenced IE departures, such as the Greek/Armenian predecessors.

We can't expect Yamnaya to be all one Y-DNA haplogroup either. Archaeologists view it as an "horizon" i.e. a cultural influence, that spread westwards over a lot of pre-existing cultures on the steppe. One of those cultures was Kemi Oba on and near the Crimea. To judge from end results, the people who left the Kemi Oba/Yamnaya part of the steppe to move up the Danube were heavily R1b-P312, whereas the people at the eastern end of the steppe were heavily R1a1a.

Jean M
12-05-2013, 11:13 AM
The earliest stages of Cotofeni are dated to about 3500-3200, which is about the same age as in the case of Afanasevo (or even slightly earlier). This means that although Afanasevo and Cotofeni represent archaeological cultures that are nearly equally distantly related to some Eastern Yamna-derived cultures like Andronovo, you suggest that the Andronovo and Cotofeni people were speaking some much more closely related languages than the language spoken by the Afanasevo people, and this is despite the fact that Andronovo and Afanasevo shared the same Y-DNA haplogroup (R1a) and were derived from the same Eastern part of the relatively large Yamna horizon. I hope you will understand that I cannot consider this a reasonable hypothesis.

You have lost me. The language that we deduce spread westwards over the steppe with the Yamnaya Horizon c. 3300 BC was early PIE. The departures westward into Cotofeni and eastwards into Afanasevo took place at around the same time and carried the same form of PIE, as we deduce from the end result i.e. Centum languages.

Andronovo represents a later departure eastwards of people speaking late PIE. The end result as we deduce was a Satem language.

All perfectly logical. Y-DNA haplogroup does not dictate the form of a language. Boys carrying a wide variety of haplogroups go to school today in New York and will end up speaking English with a New York accent. ;) Part of the vector of language shift across the steppe would have been patrilocal marriages between the newly mobile Yamnaya of various steppe origins.

Jean M
12-05-2013, 11:36 AM
Do you mean that all those new waves of Indo-Europeans entering the CW territory were derived from Yamna? Can you say anything more about their identity, time span and geographical distribution?

Yamnaya is seen as the cultural influence within which PIE spread. This influence was routed up various rivers into the North European Plain, taking with it the distinctive single graves and pottery decorated with cords, both of which have steppe antecedents. Along some of the rivers it travelled, it passed through Late Cucuteni. To the east though that influence was missing. CW is a huge horizon rather than a single culture. It is unified by the traits derived from Yamnaya, but should not be regarded as a uniform entity.


Which particular wave would you associate with the Proto-Balto-Slavs?

Fatyanovo. That is the one generally preferred by academics, as you can see from David Anthony's book.


Honestly, I find this scenario highly unlikely

You are fully entitled to your views. I understand perfectly the logic that you were following. Archaeologists and linguists have had a long struggle with the messy reality of CW, which does not present any neat and easy answers. We need a lot more ancient DNA to really sort out the migrations involved, as I say in the book.

Jean M
12-05-2013, 12:44 PM
Centum and satem division has no role in the history of PIE language development. There were never satem and centum languages, one of them older and the other younger. It is nonsense from XIX century.

You appear to be mixing up two different things. The 19th-century idea was that the first split in PIE was between:


a group moving west which developed centum
a group moving east which developed satem


In this model, there was no idea that one was older than the other. The model was geographical.

The chronological model is a 20th-century understanding of the series of sound shifts that explain the development of the various branches. Satem followed centum in PIE. That basic understanding is not falsified by some later shifts between the two in individual branches/languages.

Andrew Lancaster
12-05-2013, 12:52 PM
You appear to be mixing up two different things. The 19th-century idea was that the first split in PIE was between:

The chronological model is a 20th-century understanding of the series of sound shifts that explain the development of the various branches. Satem followed centum in PIE. That basic understanding is not falsified by some later shifts between the two in individual branches/languages.

I think there is still some doubt about the idea that there was one single satem shift in a language or dialect ancestral to all the supposed satem languages. The shift itself is of a common type. For example it happened later in imperial Latin (a centum language).

Andrew