Saturday, October 8, 2016
Dead cat bounce
Max Planck's Johannes Krause featured this curious map at a recent talk in Moscow on the Proto-Indo-European homeland debate (two hours into the clip here).
It appears to be an attempt to merge several of the main competing PIE homeland hypotheses into a single "hybrid model" under the umbrella of a modified version of the recently much maligned and now practically dead Neolithic Anatolian PIE hypothesis.
Make no mistake, it is an exceedingly strange effort that will be torn apart on several levels if it ever gets published.
See also...
Big deal of 2018: Yamnaya not related to Maykop
Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...
Subscribe to:
Post Comments (Atom)
275 comments:
«Oldest ‹Older 201 – 275 of 275Dave & Atrior
From what I recall, BMAC dates from 2200 BC; whilst southern Andronovo is c. 2000 BC, and Sintashta somewhat later
Why I brought R1b(V88) into the discussion is because of the VSO Sentence Structure in Celtic and Afro Asiatic Languages. Does this show any connection between the two ? The fact that the Basque people are Genetically the nearest to Neolithic people but their Sentence Structure is not VSO makes me think that Neolithic people did not have a link to Afro Asiatic. On the other hand there is a possibility that the Mesolithic people of Europe spoke a VSO language and distributed this into Proto-Afro-Asiatic via R1b(V88)....
@Open Genomes
Stop hijacking threads. We are not your personal army. We are a diverse bunch with very diverse political opinions, but here we discuss genetics, not politics. If that entices some political enemy of you, I couldn't care less.
Why does Celtic have a VSO Sentence Structure just like Afro Asiatic Languages but Basque do not? After all Basque DNA shows a closer affinity to Neolithic Farmers. Was Celtic VSO an isolated invention? Was the original Mesolithic languages in Europe VSO and spread from there into Afro Asiatic like R1b(V88) ? If Afro Asiatic did influence Celtic then where would the contact zone have been?
@ Ric Hern
I spoke a lot about that in another thread with many letters to Nirjhar 007. This hypothesis is inconsistent very likely. Unfortunately I have to go to Florence to-day, otherwise I'd have considered the question of R-V88 in Lebanon again. Now I say what I said: we need a SNP test of those haplotypes and an YFull exam: so everything will be clear. We have many R-V88 from Sardinia tested from Francalacci and now are in the YFull tree and in the smal's tree at FTDNA.
@Davidski
The genetic structure of the world’s first farmers (Sup. Info p.123)
While the Early/Middle Bronze Age ‘Yamnaya’-related group (Steppe_EMBA) is a good genetic match (together with Neolithic Iran) for ANI, the later Middle/Late Bronze Age steppe population (Steppe_MLBA) is not.
The map you sent shows I-I leaving Andronovo (MLBA) to enter BMAC in 1800-1700 B.C.
Hence why I said about the map: "the dates don't match with key variables - Lazaridis Steppe EMBA being one of them."
Maybe new data will change this. But I don't think will. Anyhow, makes more sense to me this way.
@Rob Yes. I'm fine with adjusting those dates too, but that's what they are right now.
@Atriðr
I just told you that Lazaridis' Steppe_EMBA grouping including Potapovka and Srubnaya individuals and you totally ignored that.
Please stop coddling your pet theory and deal with the facts.
Btw, Lazaridis also left out one of the Andronovo samples probably because it didn't fit into either the Steppe_EMBA or Steppe_MLBA groupings. It was more Central Asian than either of these groupings, and probably a hint of what's to come from the vast Andronovo horizon.
Gioiello is there a link between the VSO Sentence Structure of Celtic and Afro Asiatic ? Basque people are the closest genetically to the Neolithic Farmers but they do not have a VSO Sentence Structure in their language. This makes me think that Mesolithic people of Europe could have had a VSO like language that spread via R1b(V88) into Proto-Afro-Asiatic....
@Davidski
I'm in your house, but no straw men I insist.
I commented on your link and the trail it theorized. I brought up an issue, which you've yet to address; I did not ignore Potapovka or Srubnaya as you claim.
As for what I think, interesting choice of words, as I've never mentioned it. I had no issues with Andronovo, till Andronovo started having issues. Same with Yamnaya.
If arrows point in a direction, I'll keep looking till it doesn't look good anymore. Right now, arrows I'm seeing are looking interesting. The moment they don't, I'll look elsewhere.
Yes, I saw that Lazaridis left out one sample and probably should have left in.
@rob wrt The exact dates for the Steppe cultures and the BMAC, there is a discordance between the dates based on traditional stratigraphy and those obtained by carbon dating (the latter providing dates 200 to 300 yrs earlier)
They should define what is meant by PIE language in their hybrid model. Surely fully developed PIE with wheel and wagon vocabulary etc. could not have been spoken 8000 years ago in Neolithic Iran.
If they suggest that Neolithic Iran could contribute something to future IE languages then I can accept it. Also Eastern Hunter Gatherers could contribute something to future IE languages as there are links between IE and Uralic languages. And Anatolian Farmers could contribute something.
It is obvious that single tree linguistic model for PIE cannot hold for hybrid model. It must be a mixed model. Some words can be explained by tree model but others by wave model. It should be a much more complex model.
There is no doubt that after 3000 BC there was an expansion from Central-Eastern Europe of R1a people who had wagons, horses and houses. Their expansion influenced India, Iran, Greece and Western Europe. Their linguistic influence can be explained by the wave model:
http://i1076.photobucket.com/albums/w443/priwas/fala-1.jpg
@ Jijnasu, Atrior
Yes I understand, but my question is: how does the "Aryan trail" get from Srubna or Andronovo to BMAC, when the latter is equally old, if not older, and looks culturally wholly different ?
@Rob
The Mainstream view is not that the urban cultures of south central Asia themselves were a product of the Indo-Iranians but that they were an autochthonous civilization taken over by Pastoralists from the steppe who also borrowed some of their cultural innovations on their journey southwards. A section of the vocabulary common to IA & Iranic associated with urban technologies or religious themes are attributed to this substrate. However very little can be said conclusively about the nature of this interaction IMO
@ Davidski
"Whether the authors know this or not, their model hangs on two assumptions.
- that R-M417 and R-M269 are not EHG markers but arrived on the steppe from Iran
- that steppe admixture in South Asia dates to the historic period.
Both of these assumptions are wrong, and everything else is background noise."
I agree that R ultimately should come from northern Eurasia, because it belongs to K node and Q is found in America and Malta1 was R. However, the time scale is very big and people can cover long distances in a 1000-year time, let alone in 20 000 year-time.
I do not think that this new model implies that R-M417 and R-M269 must have come from Iran, but it is still possible that both haplos had reached South Caucasus/ Northwest Iran c. 2500 BC where proto-IE language developed. R1b surely was there early.
You are very stuck on yDNA and genetic transmission of languages in some cases (Yamnaya) but not in other cases (Uralic Finnics). I understand that this model is not based on an yDNA based model of expansion.
As for that steppe admixture in South Asia, you will be in trouble if the new paper detects R1a1 in India c. 2500 without any steppe admixture. Sure, this "nightmare" may never come true for you.
I do not think that R1a1 must have spread to India through Central Asia. It is possible that it happened through Caucasus. IMO, Yamnaya theory is more attractive if R1a1 went to Iran through Caucasus. In any case, I agree that there are linguistic reasons for presuming that proto-IE was spoken north of Caucasus.
@Kristiina
I said R-M417 and R-M269.
These are very specific subclades with rather specific histories.
@Kristiina
Current mainstream theory doesn't allow for a PIE development in 2500 BC, it's older. Development in BA Iran is out even in this Max Planck model.
@ Jijnasu
Thanks, yes Im aware of the popular paradigm. My remark was direct specifically at the link Dave posted, and Atrior was discussing, which linked BMAC as part of the 'Aryan trail'.
Indeed, BMAC looks nothing like the steppe, but is a central Asian culture, for lack of better descriptive.
If so, steppe admixture must have arrived quite late, c. 17- 1500 BC, which means the earliest Veda's were possibly still being composed on the steppe (going by the steppe scenario). Of course, there are 3 or 4 other caveats around this....
The model does not suggest any pro-IE c. 2500 BC - and nor do I - but apparently an Indo-Iranian protolanguage c. 2500 BC in Pakistan. Can't you grasp what I mean when I say that this model is probably not based on yDNA. R1a-Z93 may have been incorporated into IE speaking groups c. 3000 BC in Southern Russia and been taken to Iran and India. We know that also I2 was incorporated into IE speaking groups in Central Europe. This Max Plank model probably implies that proto-IEs were predominantly R1b-269.
@Kristiina
If an Indo-Iranian protolanguage was spoken c. 2500 BC in Pakistan how do you explain that a completely different population was speaking similar language c. 2500 BC in Poland. Separated by thousands of kilometers and masses of completely different R1b-269 dominated proto-IE groups independently developed many common features? It just doesn’t make any sense.
EastPole, this is not my theory, so you should put this question to Johannes Krause and not to me. He apparently has access to ancient data that I definitely do not have.
However, if a Balto-Slavic(-Germanic) protolanguage was spoken c. 2500 BC in Poland how do you explain that a completely different population was speaking similar a language c. 2500 BC in Pakistan.
You have a similar problem either way, but it is not true that there is no R1a1 but only R1b in the Near East between Eastern Europe and India (http://dienekes.blogspot.be/2014/03/major-new-article-on-deep-origins-of-y.html). There is a southern trail of R1a1 from Eastern Europe to India: Caucasus Megrels 9%, Georgians 11%, Caucasus Adyghes 11%, Abkhazians in Georgia 33%, Syunik Armenians 9%, United Arab Emirate Arabs 7%, Iranians 12.67%, Iraquis 6,5%, Iraq Kurds 12%, Pathans 45%, Pakistan Sindhi 49%, Pakistan Makrani Negroid 18%, Pakistan Mohanna 71%, Nepalese Chitwan 70% (https://en.wikipedia.org/wiki/List_of_R1a_frequency_by_population) (These are not R1a-Z93 but general R1a frequencies!)
Davidski I just want to make sure if I'm banned or if something is wrong with my Internet because I can not see the latest posts. Sorry for bothering you with this but an answer will be hugely appreciated.
@Kristiina
"However, if a Balto-Slavic(-Germanic) protolanguage was spoken c. 2500 BC in Poland how do you explain that a completely different population was speaking similar a language c. 2500 BC in Pakistan.
You have a similar problem either way"
Not if you take into account autosomal DNA. DNA from modern populations strongly suggest a 'Corded Ware-like' population moved into South and West Asia, and admixed into the local populations.
There is no autosomal DNA evidence at all that suggests the reverse. Y haplogroups simply strengthen the case for a Europe to West Asia movement of people.
And the linguistics also make sense.
To suggest that a movement from Pakistan to Poland is equally likely is total nonsense. It is not equally likely.
Karl, this Max Plank model does not suggest any movement from Pakistan to Poland. Where did you get that? If this model is true – which we do not know of course – R1a1-Z93 was incorporated into IEs in Southern Russia and these cultures could have been for example Catacomb culture (ca. 2800–2200 BC) and Poltavka Culture (2700—2100 BC) and from there southward down to Pakistan and India. According to this model, Proto-IE language was spoken earlier in Southern Caucasus / Northwest Iran in a culture that was probably predominantly R1b or even J or other unidentified haplogroup. The most interesting aspect of this model is that it has replaced the Central Asian route of expansion with a West Asian route of expansion.
@Kristiina
"According to this model, Proto-IE language was spoken earlier in Southern Caucasus / Northwest Iran in a culture that was probably predominantly R1b or even J"
I don't understand why you are so fixated on Y haplogroups. Look at the autosomal components. It is very hard to make sense of it the way you ae doing it.
Sure, for example Punjabis seem to have very little light blue Steppe component, maybe only 5-7% and are heavily West Asian, and this Max Plank figure gives the impression that Indo-Aryans c. 2500 BC were autosomally West Asian. For me it is okay either way. I do not have any preferences. Let's wait for the paper to see what they have really detected. Maybe they suggest that Indo-Aryan languages were introduced by yDNA J which was autosomally West Asian. :-)
One of the big problem with this new model is not about DNA (ancient or modern) is about PIE language itself. PIE as recomposed today must have emerged in a very peculiar society, a Steppic society, with no sea, but with horse, and so on.
If we follow this "hybrid" theory (in fact only a revised Anatolian theory), it would mean that the archeological culture supporting PIE would be in the Fertil Cressant, where we already know that animals, plants and so on are not the same as described in PIE.
So, not only genetically, it's bulshit, but archeologically and linguistically also.
'Maybe they're just trying really hard to accommodate Heggarty and Gray?'
I would say that's the most likely explanation. Russel Gray works at the same institute, and Johannes apparently absorbs his theories to some extent...
I don't think fixating on SOV/SVO makes a great deal of sense given the amount of diversity in word order in other IE languages. There are languages out there with pretty flexible word orders. I wouldn't assume Basque is necessarily representative of all neolithic languages either though.
Re: Kristiina's points - the genetics of the model in this paper are essentially the same as those of the standard Kurgan model, more or less. The only way to distinguish the two would be to check the autosomal DNA of early Anatolian IE speakers. Under the Anatonial model, early Hittites and Luwians should not have steppe component, or at least not greater than their non-IE peers. Under the Kurgan model, these people should have an inflated EHG component. Until we can compare pre-IE Anatolians with early IE Anatolians there's really no way to prove or disprove one or the other. Modern Y-DNA isn't all that helpful simply because modern Y-DNA diversity is so damned high in Anatolia to begin with, even for R1a.
That being said, I definitely think the Kurgan model is more likely. It looks like an early wave of R1b IE speakers radiated out from the Pontic Steppe, with a second wave of R1a IE speakers spreading Satem IE languages. I think the linguistic evidence against the Anatolian hypothesis is actually stronger than the genetic.
On the R1b-V88 discussion, I don't linked it to proto-Afro-Asiatic makes all that much sense. For one, it seems to be rare or absent among Omotic, Semitic and Cushitic speakers - http://www.nature.com/ejhg/journal/v18/n7/fig_tab/ejhg2009231t1.html#figure-title
More likely seems and entry into the North African Afroasiatic gene pool some point during the neolithic (and thus well after the break-up of proto-AA), probably from hunter-gatherers resident around Tunisia and Libya at the dawn of the Holocene, who spread south with the improving climate. It couldn't have been common in the Levant or else it would have spread more widely with agriculture, which is why I think an Italian or Iberian route to North Africa is more likely.
E-M215 seems like a far more likely proto-AA marker.
For some people here, it is very difficult to digest the possibility that a movement from South from either Iran or SC Asia, could have contributed the CHG-like ancestry into Eneolithic steppe and the Yamnaya. It is almost like a heresy to them and they get very much offended by the mere suggestion of such a possibility.
They need to keep in mind the following facts :-
1. Right from the time when Haak et al 2015 paper came out, it became clear that a Near Eastern/South Asian component had contributed to the ancestry of Yamnaya which was absent from earlier HG populations. This was a very important finding since, it is the Yamnaya which is credited with bringing IE languages into Europe. It clearly brought into picture the possibility that this NE/South Asian component itself might have brought the IE languages into the steppe.
2. Ancient DNA has shown that this Near Eastern/South Asian ancestry component has been present in the Caucasus and in Iran since the start of the Neolithic and even earlier. Ancient DNA from Central Asia and South Asia may as well show us the same thing. This ancestry component which is associated with the 2nd group of Early farmers, appears to have been spread out over a very vast region, from the Caucasus through Iran & Central Asia and into South Asia. Most of this vast area also had early and advanced prehistoric civilizations.
Is it therefore, so outlandish to think, that some people from this vastly spread out group could have migrated into the less advanced and thinly populated HG regions of the steppe and influenced and transformed their religion, culture and language ?
3. Lazaridis et al 2016 suggestion that a group closely resembling Chl_Iran contributed to the ancestry of Eneolithic steppe and the Yamnaya, looks like a very real possibility. There is also an archaeological evidence for such a movement. Read this - https://en.wikipedia.org/wiki/Cave_of_Dzhebel. That this group could have radically transformed the culture, religion and language of the steppe people appears quite certain. It is therefore very much likely that this is how IE languages entered the steppe.
4. Further evidence of movement into the steppe from the South comes from the fact that the native steppe cattle, has admixture from South Asian Zebu cattle. This is not a recent admixture but an old one. The question is how old ? If it came with the Yamnaya, things will become very interesting.
5. There are two main lines of sheep mtDNA, A & B. mtDNA B predominates across Europe and is probably of Near Eastern/Anatolian origin. However mtDNA A has its highest diversity in South Asia and is likely to have originated there. This mtDNA A is present in significant proportion among the sheep of Eastern Europe and the Caucasus. A point to ponder again.
6. Recently, aDNA of a dog from a corded ware site in Germany, showed admixture into the dog dna from Indian/Iranian dogs as well as Indian/Iranian wolves. This again needs explanation.
-----------------
When the evidence such as above, exists, I think that it is very hard to keep denying the very strong probability of Central Asian/Iranian people contributing to the formation of the Yamnaya.
@Davidski:
What do you think of this work:
http://www.cs.rice.edu/~nakhleh/Papers/81.2nakhleh.pdf
You can say that they used only a limited *type* of linguistic information, but it's refreshing to see linguistic *networks* being discussed at all. On page 391, authors begin discussion of relevance of "wave model".
Notice that the software produces phylogenetic language almost-trees that are trees plus fewest number of "admixture" edges -- exactly alike to some of the trees+admixture edges you've produced in recent posts via genetic data...
@ Folker
The region suggested by this model isn't the Fertile Crescent, amd in fact had steppic environment and a culture based on metallurgy and pastoralism.
In other words, what we might expect From linguistic palaeontology for PIE
(Google "fertile crescent" if your unsure of its basic geography)
I'd like to make a point about linguistics, genetic data, and "race as a social construct". I've pointed out that the Natufians do in fact have about 25% of what is clearly an African component. Based on the Y-DNA and the split between East African E-V1515 and E-PF1962 within E-Z830, right at the height of the LGM at 19,200 ybp, and the fact that the Natufians were E-M123 / E-M34 before the Younger Dryas, it would seem that Proto-Afroasiatic arrived in the Near East right at that time.
YFull E-Z830 tree
This corresponds very nicely with the fact that the Beja language of Northeast Sudan may not be Cushitic, but rather a separate basal branch of Afro-Asiatic. The locus of greatest diversity of Afro-Asiatic is in fact somewhere between the western Ethiopian Highlands (Omotic) and the Beja region of Northeast Sudan.
Eurogenes Lazaridis (2016) PCA closeup showing Natufian affinity with Africans
Eurogenes Lazaridis (2016) 3-D PCA overview showing Natufian origins in Africa
This East African component later spread to Neolithic Anatolia, where it became incorporated into the Northwest Anatolian population, and later Neolithic Europeans. Craniometry shows that some Starcevo and Funnelbeaker people were practically identical to the Natufians in their skull shapes, and even rather different than Neolithic Anatolians. LBK Stuttgart girl also has a clear autosomal African component. This is the "LBK-like" ancestry found in the Llorente (2015) Mota study.
However, in spite of the clear evidence of an African origin of the Afro-Asiatic languages, which is now supported by aDNA from the Natufians, we get articles like this:
Saharan remains may be evidence of first race war, 13,000 years ago
Which two "races" were in conflict in Sudan 13,000 years ago, contemporary with the Natufians?
(There's no evidence of "LBK-like" Eurasian ancestry in Africa this early. This ancestry clearly arrived with the First Farmers, at least 5,000 years later.)
Why does this pass for "science" in the popular press?
So our present-day "Racial" constructs influence not only how we interpret the data, but also what we choose to see.
@Jaydeepsinh Rathod
When the evidence such as above, exists, I think that it is very hard to keep denying the very strong probability of Central Asian/Iranian people contributing to the formation of the Yamnaya.
I realize that you have difficulty interpreting genetic analyses and lack the ability to analyze data yourself.
But this is beyond a joke now, because you've been saying the same thing for months, despite not being able to put together a coherent argument in favor of your claims.
There's absolutely no chance that what you're claiming is true. You need to accept reality and move on.
@MomOfZoha
Yeah, it's not bad.
*If* this map is based on unreleased aDNA, which is far from clear, then it raises a number of questions. Where did Greek, Armenian, and Albanian come from? Why is the arrow from SC Asia to NC Asia dated at 500 BC? Do they have DNA from there showing Iran_N type ancestry didn't arrive until the Iron Age? That would really be unexpected. If there really are Harappan samples ready, they are supposed to be from Rakhigarhi, which is located near the eastern edge of the Harappan territory; it could have a higher ratio of ASI than the norm.
@Atridhr
Could you quit trying to be cute and just say what the hell you mean? Your vague posts littering the comments are annoying.
@ OG
"I'd like to make a point about linguistics, genetic data, and "race as a social construct". I've pointed out that the Natufians do in fact have about 25% of what is clearly an African component. Based on the Y-DNA and the split between East African E-V1515 and E-PF1962 within E-Z830, right at the height of the LGM at 19,200 ybp, and the fact that the Natufians were E-M123 / E-M34 before the Younger Dryas, it would seem that Proto-Afroasiatic arrived in the Near East right at that time."
Thank you, very interesting , and certainly strengthens the arguement for AA
"Which two "races" were in conflict in Sudan 13,000 years ago, contemporary with the Natufians"
Agreed. A local skirmish over a watering hole, more likely. Although one can imagine the Journalists were merely trying to "spice up" anthropology, they should be more careful with their wording, perhaps by getting counsel from anthropologists
@ Capra
Lol
The Balkan IE lects won't find a simple solution, because they are too structured & diverged vis-a-viz each other ,
Despite attempts by earlier scholars like Darden to do so. We need layering, multiple loci and a more sophisticated version of the Tree model. And I don't mean to be cute, but this requires a formal paper
@Davidski:
I thought the algorithmic novelty would have been more than what I'm seeing (Appendix A), but then again it's a linguistics journal. The characters and their coding, on the other hand, will take a long time for me to understand...
@ Ric Hern
"Gioiello is there a link between the VSO Sentence Structure of Celtic and Afro Asiatic ? Basque people are the closest genetically to the Neolithic Farmers but they do not have a VSO Sentence Structure in their language. This makes me think that Mesolithic people of Europe could have had a VSO like language that spread via R1b(V88) into Proto-Afro-Asiatic...."
Actually people thought the other way around, that the VSO construction of Celtic (but only the Isles ones) demonstrated a presence of a Berber substratum in the Isles.
Unfortunately we don't know which language the R-V88 who migrated to Sahara spoke (that in my theory, till a few days ago I was the unique to think that, even though the same Cruciani many years ago wrote to me that, even though he believed (and wrote) in the Middle Eastern origin, he shouldn't discarded my hypothesis).
Of course there have always been migrations from and to Europe and Northern Africa, but we don't know where the Afro Asiatic languages came from. Ted Kandell is sure of a Northern African origin. I don't know so far, but perhaps from genetics we'll be able to understand something, and from aDNA.
Certainly some words made me think, for instance "berra", if it is also in Maltese surname Habeyer from Arab of Sicily "hebejjer(?)" (einjähriges Schaf).
David,
You're only proving my point. What indeed is a joke is the steppe theory of IE origins. That the most dominant and widespread language family came out from a bunch of nomads from the thinly populated steppe and that it managed to overturn the entire linguistic and cultural landscape of Central Asia, Iran and South Asia. Such a massive turnover yet not much archaeological evidence to support it beyond evidence of contacts between steppe cultures and BMAC, which by no means prove any movement of people from the steppe into Central Asia, much less any cultural influence of the steppe. Yet people believe it to be as true as the day and the night. If this is not a joke than what is ?
On the other hand, you have by no means disproved the possibility of a movement from Chl_Iran into the steppe. You're only making the weak argument that CHG ancestry, and not Chl_Iran, contributed the Southern ancestry into the steppe. And that too by the CHG women. You've ofcourse proven none of this. CHG can itself be modelled as a heavily admixed with Iran_N.
A movement from across the Caucasus into the steppe is probable. And as Mariya Ivanova has very clearly shown, the Maikop culture itself seems to have been its antecedents further SouthEast from Iran and Central Asia, going all the way even to Indus Valley. The presence of South Asian y-dna L1a in Chl_Armenia as well as the discovery of an Indian mtDNA M52 in Maikop also corroborates Ivanova's findings. Keep ignoring all of this, and somehow all of it is going to go away.
So far, it is only the presence of R1a in ancient samples from the steppe, from which you're acting jubiliant. But ancient DNA from India has not come. The geneticists like Haak, Lazaridis & Reich, have only argued for a movement of Neolithic Anatolians and later the steppe people, into Europe after they had analysed the pre-Neolithic genomes from Europe. Not before that.
Yet people do not want to exercise the same restraint for South Asia. Hold back your horse. aDNA from Rakhigarhi could be as old as 5500 BC, i.e. comfortably older than Yamnaya & Eneolithic steppe. Let us see if those old samples are like ASI or similar to modern population of that area. Do not be so presumptuous to think that South Asian aDNA will turn out to be ASI-heavy.
----
And please do explain how South Asian Zebu and South Asian sheep managed to introgress into the steppe. Also explain the dog aDNA from corded ware with South Asian dog and wolf admixture. Does such an evidence exists of steppe sheep or steppe cattle admixture into South Asia or Iran ?
@Jaydeepsinh Rathod
I'm going to re-iterate a few points now, and you have to tell me whether you understand them or not. If not, then tell me why you don't understand them and I will try and explain them to you in another way.
However, make no mistake, these are problematic truths that you have to deal with sooner or later.
- Yamnaya and Afanasievo are a mixture of EHG, CHG and something more western than CHG
- the non-EHG half of Yamnaya and Afanasievo only superficially resembles Iran_ChL, because Iran_ChL is also a mixture of CHG and something more western than CHG
- Yamnaya and Afanasievo cannot be a mixture of EHG and Iran_N or Iran_ChL because Yamnaya's and Afanasievo's non-EHG mtDNA doesn't match the mtDNA of Iran_N or Iran_ChL
- there is no Iran_N ancestry in extant Northeastern Europeans, who generally have the highest ratios of Yamnaya-related ancestry among Europeans, so it's impossible for Yamnaya to have Iran_N ancestry
Please confirm that you understand these facts, and the consequences that they have for what you've been arguing here for months.
If you don't believe what I'm saying, then e-mail Nick Patterson, whose e-mail I know you have, and ask him whether these points make sense or not.
Btw, cows, sheep and dogs don't speak languages, so if there was no human migration from Iran or Central Asia to the steppe, then clearly Indo-European languages were not introduced there by cows, sheep and dogs. Domestic animals can be traded and this doesn't require human migration or language change.
@ Jaydeepsinh Rathod
Davidski explained you the question by an autosomal point of view, I try to explain you the same by an uniparental and a linguistic point of view:
1) of course what you say, that we have to wait for aDNA from India is important, and we all are waiting for that (from India as from any other place: more and more aDNA is welcome)
2) I wrote in the past about a weird R1a1a from India and its strange haplotype, and said that, very likely, not only the age of the haplogroup is older than it is thought so far (even though not much more, perhaps for an 1.17 factor), and that India has very likely some old R1a haplotypes too, but that the mass of R1a are recent, whereas the oldest haplotypes are in Western Europe, thus India or Iran aren't the source of the hg. (Iran isn't also for the aDNA we know, about India we'll wait for the results)
3) Of course from a linguistic point of view the separation of Sankrit from the European languages, also the satem of Eastern Europe, is probably older, because Sanskrit had many changes that required more time
4) but very likely the people who spoke "Sanskrit", or anyway the dialects from which Sanskrit was constructed, very likely belonged to the IE migrated from Samara to Central Asia (I spoke above of possible links between an extint language perhaps linked to Sankrit in Proto-Uralic and some Turkish languages)
5) thus I am expecting that this people, hg. R1a downstream R-Z94 but also other hgs., came from Central Asia. Also Indo-Iranian languages may have had a different pathway, and also their separation is very likely older than it is thought.
@Gioiello
Actually people thought the other way around, that the VSO construction of Celtic (but only the Isles ones) demonstrated a presence of a Berber substratum in the Isles.
VSO word order was a common feature in Gaulish (continental Celtic) too and increasingly became more common as the Gaulish language got simplified and regularized over time. See: http://www.moderngaulish.com/welcome#!__word-order
David,
Thanks for putting forward your arguments. Let me answer them one by one.
-- CHG itself was modelled by Lazaridis et al as 71 % Iran_N, 7 % WHG, 22 % EHG. The Western component is the EEF or Anatolian_N. This was already present in Iran-Chl, the major reason why it makes a good fit for the southern component in Yamnaya.
-- Superficial or real, Iran_Chl makes a good fit for the Southern component in Yamnaya. Iran_Chl does not necessarily derive a major part of its ancestry from CHG, because CHG itself can be modelled as having major ancestry from Iran_N like population.
-- We have the presence of J2a in Iran_ Hotu which is also present in one Karelia EHG sample. Plus Iran_Chl has J2a and G1a, both of which are present on the steppe. More sampling from Iran, Central Asia and South Asia will likely reveal more Y-dna and mtDNA correspondences.
-- I am not sure how NE Europeans do not have Iran_N ancestry. Do they also not have CHG or Iran_Chl admixture ?
-- Lastly, cows, sheep and dogs migrate in large nos. to leave a demographic impact only if there is an associated human migration. Mere trade cannot leave a big enough demographic impact that lasts for millenia. Therefore the South Asian genetic signature into these animals therefore signals a possible human migration from Central Asia or Northern Iran because South Asian Zebu and likely the sheep as well, were already spread into Central Asia and Iran in the Chalcolithic.
Gioiello,
Two points,
- South Asia is massively undersampled. To give you an idea, the Underhill paper that argued for the Iranian/Anatolian origin of R1a, had about 1,700 samples from Iran, a country of 75 million people, but it only had about a 1,000 samples from the combined population of India & Pakistan which is about 1.5 billion i.e. 20 times that of Iran. So if we are on the lookout for the rare older haplotypes of R1a, such an undersampling will understandably not yield much results from South Asia.
- Sanskrit is not the mother of all Indo-Aryan languages in India. Most of the modern Indian languages are heavily influenced from Sanskrit but they are not directly descended from it. They're descended from Prakrit languages, which shared features with Iranian and even European languages which are absent in Sanskrit. Clearly, the IE linguistic diversity in India is very much underestimated. Most comparisons only focus on Sanskrit and will have you believe that it is the mother of all Indo-Aryan languages when it clearly is not.
@Jaydeepsinh Rathod
It's not relevant that CHG was modeled as mostly Iran_N, because they're distinct populations and Yamnaya has ancestry from CHG not from Iran_N.
Thus, no Iran_N in Yamnaya = no South Central Asian ancestry in Yamnaya.
And you keep ignoring the fact neither Yamnaya nor any other Bronze Age steppe groups show the type of mtDNA seen south of the Caspian, including in Iran_ChL. So the logical conclusion is that the statistical fit featuring Iran_ChL is a coincidence.
You can't ignore this fact, because no Iranian uniparental markers in Yamnaya = no Iranian admixture of any kind in Yamnaya.
Also, nope, there's absolutely no evidence of any ancestry from the South Caspian in Northeastern Europeans. CHG yes, Iranian no. You surely know what this means.
Btw, I don't know much about Asian domestic animals, and since I can see fatal flaws in your arguments about human population structure, I'm very skeptical about your claims here.
@ Jaydeepsinh Rathod
I said that aDNA from India, as from everywhere, is welcome, and we all hope having much more, but India had many samples in the 1KGP, thus I think that it isn't a few tested. Anyway we are here for next proofs.
About languages I know very well what you said, and of course the IE languages in India are a complex phenomenon, but I think it isn'0t wrong to take Sanskrit for our comparisons (even though a synthesis, an artificial language; but every literary language is a construction, also to-day), because what is interesting are some data that permitted me to compare words like "si/e/asare" in Uralic of the word for "daughtar" in some Turkish branch.
Anyway we are open to any proof that will come. We may do hypotheses as to the data at our disposal. I announced Villabruna ten years before that he were tested. We'll see in the future.
@ Onur
"VSO word order was a common feature in Gaulish (continental Celtic) too and increasingly became more common as the Gaulish language got simplified and regularized over time. See: http://www.moderngaulish.com/welcome#!__word-order"
I am not able to see your link (when someone asks me to upload something, I of course don't do it), but perhaps you should try to translate this:
MARTIALIS DANNOTALI IEVRU VCVETESOSIN CELICNON ETIC GOBEDRI DVGIION TIIO VCVETIN IN ALISIIA
But are you Dinc,er? If you knew how many morons write on linguistics without being linguists!
@Gioiello
That article I linked above is well written with lots of citations and excerpts. Here is the most relevant part:
These extracts from the most important continuous texts from the ancient Gaulish corpus show clearly that the Gaulish language frequently used constructions where the verb was followed by the subject, often in turn followed by the object (VSO word order). While there is variation on the theme, and while various translations continue to be fought over bitterly without consensus, there is no doubt that a word order placing the subject after the verb was in common usage. Furthermore, the Chateaubleau text, which is the most evolved Late Gaulish text available to date, is virtually exclusively VSO (see Mees 2010 for the parsing and translation).
The above observation is supported by Isaac (2007), who asserted that Gaulish had developed a tendency towards VSO word order by the 1st cent. BCE (Isaac 2007). The examples clearly show that verbal forms are followed either by a cliticised subject pronoun, apparently in accordance with Wackernagel’s Law and Vendryes’s Restriction, or, when pronouns are not used, by what is clearly recognisable as a subject (e.g. Rèze, Larzac).
Yes, I am Onur Dinçer you know from Facebook and from Dienekes' blog before that.
Onur, but what they say
The above observation is supported by Isaac (2007), who asserted that Gaulish had developed a tendency towards VSO word order by the 1st cent. BCE (Isaac 2007). The examples clearly show that verbal forms are followed either by a cliticised subject pronoun, apparently in accordance with Wackernagel’s Law and Vendryes’s Restriction, or, when pronouns are not used, by what is clearly recognisable as a subject (e.g. Rèze, Larzac).
seems to agree with an internal and late development, as it is in Wackernagel's law.
Now I am going to eat two eggs with a little bread and a glass of wine...
@Jaydeepsinh Rathod
True most Modern Indian languages don't directly derive from classical Sanskrit, but they can all be traced to Old Indic dialects closely related to Rg Vedic. Some OIT theorists like Talageri suggest that the IA languages are far more diverse having amongst them features of various IE subfamilies. However except for traces of a lost kentum language in the Bangani dialect, there is no evidence of any of this.
@Gioiello
seems to agree with an internal and late development, as it is in Wackernagel's law.
VSO order was still pretty common in Gaulish even before its near-standardization. Proto-Celtic probably also harbored VSO along with other word orders.
Now I am going to eat two eggs with a little bread and a glass of wine...
Bon appetit!
Valid Points Jaydeep!.
Gioiello,
Why don't you check his academia.edu profile. Don't expect to be spoon fed every time ;) . You are old and I love to help the old scholars but, you are not that old that you even can't get that article!.
@ Nirjhar007
I thank you, but, tell me, if I read all the paper, may I reach other conclusions rather than those I reached from this quote?:
The above observation is supported by Isaac (2007), who asserted that Gaulish had developed a tendency towards VSO word order by the 1st cent. BCE (Isaac 2007). The examples clearly show that verbal forms are followed either by a cliticised subject pronoun, apparently in accordance with Wackernagel’s Law and Vendryes’s Restriction, or, when pronouns are not used, by what is clearly recognisable as a subject (e.g. Rèze, Larzac).
In other blogs and in other threads I have to do with Ted Kandell/Open Genomes, who reads BAM files etc. I think that I won't be able to get the programs that YFull has, thus I wait the a BAM file is there.
Many have PC terapowered than mine, but give me two haplotypes with only a few markers, and I'll understand what their programs won't ever do.
Oh comon just read the research by Bomhard! :) .
@Jaydeepsinh Rathod
explain how South Asian Zebu and South Asian sheep managed to introgress into the steppe. Also explain the dog aDNA from corded ware with South Asian dog and wolf admixture.
The same way the WHG cultures such as Ertebolla that went along during the Neolithic got pigs with Anatolian mtDNA: Trade, or perhaps even raids.
@Open Genomes
I've pointed out that the Natufians do in fact have about 25% of what is clearly an African component.
Yet they don't snug to Africans in D-stats.
@Jijnasu said...
"True most Modern Indian languages don't directly derive from classical Sanskrit, but they can all be traced to Old Indic dialects closely related to Rg Vedic"
This is a myth ..NIA is often left uncharacterized. Talageri does not profess to be a linguist he has merely referd to others. As I read word lists say Masica's X list I see that its so hopelessly inadequate to explain extra vedic components in NIA. Vedic is good for reconstructing PIE but beyond that linguists have not cared to delve too much.
@jijnasu
As a simple exercise:
how do we derive some thing as basic as NIA (bhAI, bhAU, bhaiyyA) from sanskrit bhrAtarah.
What's the etymology of bhIR/bhID meaning crowd?
Perhaps its been done and I am ignorant.
@jaydeep
how do we know if zebu admix in italy is not recent or due to gradual diffusion from say anatolia. what are zebu levels in todays anatolian cows?
The Water buffalo population in romania and Italy would be a good isolated reference population.
@postneo
The poor representation of the Prakrits and Apabhramshas in addition to the overalap between various regions has made it difficult to trace the evolution of NIA accurately. However as far as I am aware there is very little to suggest that NIA dialects owe their structure to other branches of IE (There is however evidence of influence from neighbouring non IE languages and some unknown substrates). It seems hardly likely that such features would have been completely missed by all those studying IA.
I am not an expert but as to your question
BhAI, BhAU can be traced to the OInd BhrAtR
the vowel R is replaced by i or u in the prAkRts
likewise consonant clusters were often assimilated giving rise to forms like Bhatu or Bhati (attested in the early M Indic pALI)
elimnation of consonants between vowels was common in later prAkRts and apabhramsha and elimnation of the T probably gave rise to the present form
I am not sure of the origin of bhID though
@jijnasu
you really don't have to try very hard.
Just look at the english word thirst and sanskrit trishNA, bengali tEshTA with retroflex nasal flap or simple stop. its a regular derivation from classical sanskrit to bengali, Its there in germanic as well but missing in non germanic and hence PIE.
ver curious, wont say more ..
@postneo
I don't get it, how is tṛṣṇā specifically related to thirst? They both have the root in zero grade *tṛs- but with different suffixes.
@capra
the root is the same as you said.
the meaning of thirst and tṛṣṇā are the same.
Can you tell me the difference between the suffixes in the two cases
btw I thought *tṛs was not part of PIE.
@postneo
Well, I'm no expert, but as I understand it they are just two different suffixes to form a noun from a root. The Sanskrit one from *-na and the Germanic one from *-tis, both literally something like "dryness".
The root *t(e/o)rs- "dry" is well-attested, e.g. Ancient Greek térsomai "to dry up", Irish tart "thirst", Albanian ter "dry", Icelandic þurr "to become dry", Latin torreō "scorch, roast" (hence English toast), Old Aremnian tʿaṙam "withered". Also Latin terra and Irish tir "land" are both thought to come from the same root, in the sense of "dry land". (Most of this just from Wiktionary.)
I'm not sure if the form with syllabic r occurs in any branch other than Germanic or Indic, though, if that's what you are getting at.
@postneo
Is it possible that the bengali word you mentioned is derived from an alternate form like tRSita or taRSita? The word though seems to have a complex history and looks like at has been reborrowed at some point into MIA or the early NIA predecessor of Bengali since it retains the consonant cluster ST rather than it becoming TTh as expected. It is possible to go on about this but let me end it on this note. A more detailed study of NIA is required but given the lack of sufficient evidence so far there seems little reason to suggest that India was a region with great diversity of IE. Genetics at this stage seems to very strongly support the standard model and makes the OIT very unlikely. The very improbable event of discovering r1a in Rakhigarhi and an autosomal makeup very similar to modern Indians would be the only reason to completely revise our understanding of Indian linguistic history
Rob
"The copper in the Khvalynsk Chief's grave is of Balkan provenance, not Urals"
Ty, I'd missed it had been tested.
Which lead me to this.
https://books.google.co.uk/books?id=nKNOAAAAIAAJ&pg=PA44&lpg=PA44&dq=khvalynsk+grave+copper+beads&source=bl&ots=BVnbRy9GMq&sig=xonm8njrE_8QHtNT7-pxf95zJQU&hl=en&sa=X&ved=0ahUKEwj_9pLKi9rPAhVCAsAKHWlVCdM4ChDoAQgsMAY#v=onepage&q=khvalynsk%20grave%20copper%20beads&f=false
"At the same time, however, it is unlikely that eastern imports exclusively took the form of finished objects. It is not possible, for example, to find direct parallels for the shell-shaped pendants in the Tripolye metalwork assemblage; such artifacts are probably derived from the earlier, local tradition of steppe metal working."
"Preliminary data from metallographic analysis of these ornament (Ryndina and Ravich) indicate that the hammering and casting methods are distinctive even in comparison with those of the Tripolye culture."
"In an examination of the metallurgy of subsequent historical periods , we never again come across a similar situation, in which the peripheral focuses of a particular province were markedly different in terms of their metal production and metal-working technology."
So apparently although the copper used was from the Balkans the artifacts themselves had a unique steppe style of their own hinting at an earlier metal working tradition.
So maybe a local tradition developed from but limited by relying on found natural copper from rivers in the south Urals rather than mining and smelting and then later using more abundant imported copper - paid for with horses? - while maintaining the traditional steppe metal working style?
David,
It is very probable that CHG and Iran_N sprang from the same ancestral source. The sharing of ydna J2a between CHG & Iran_N also points towards this likelihood.
You have ignored the sharing of y-dna markers between Iran_Chl/Iran_N and Eneolithic_Karelia/modern Russia such as G1a & J2a.
As far as the sharing of mtDNA markers is concerned, you know it well that samples from Iran have been limited and even in that limited dataset, Lazaridis did not disclose the mtDNA from their study. So we only the Broushaki paper. Here we had an mtDNA T2c, which was present in one of the Yamnaya samples published by Haak et al. Another mtDNA was J1d6 which is present in modern day Russia.
For the domestic animals, you already know about the dog sample, so let me just give you some links which highlight the origin and also the present spread of Asian cattle and sheep.
http://www.nature.com/hdy/journal/vaop/ncurrent/full/hdy201679a.html
"By analysing the genome-wide SNP markers, McTavish et al.
(2013) and Decker et al. (2014) also reported an indicine influence on
Italian cattle breeds. Using whole-genome sequences of ancient human
DNA, Jones et al. (2015) and Haak et al. (2015) suggested massive
migration of Yamnaya steppe herders as a source of dispersion of
Indo-European languages to both northern-central Europe and India.
These herders might also have mediated gene flow between Indian
zebu and Ukrainian steppe cattle."
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0077858#
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4576706/pdf/msv139.pdf
(figure 2B for the spread of mtDNA A in Eastern Europe and Caucasus)
postneo,
"how do we know if zebu admix in italy is not recent or due to gradual diffusion from say anatolia. what are zebu levels in todays anatolian cows?"
We know this because the Italian cattle in which we find the Zebu admixture is the Podolian cattle which is considered to have its origins in the steppe.
https://en.wikipedia.org/wiki/Podolian_cattle
@Jaydeepsinh
The Laziridis mtDNA which wasn't in the preprint was published in the completed paper. There were only a couple of the Iranians missing in the preprint, most of it is Natufian and PPNB.
The only Iranian mtDNA sharing is J1c - a J1c11 in Iran_N and a J1c1b1a in Sintashta - but J1c1 was present already in the European Early Neolithic. (And that was in the preprint.)
There is now Natufian J2a2, and J2a2a in 2 of the Afanasievo, but the connection is not clear.
T2c1 was also present in the European Neolithic.
Of course this is all from the central Zagros and does not represent the rest of Iran.
@Jaydeepsinh Rathod
It is very probable that CHG and Iran_N sprang from the same ancestral source.
But obviously this happened during the UPPER PALEOLITHIC and is irrelevant to the question of Yamnaya origins or the PIE debate.
Why do you bring up such irrelevant points?
You have ignored the sharing of y-dna markers between Iran_Chl/Iran_N and Eneolithic_Karelia/modern Russia such as G1a & J2a.
We don't have any samples from Eneolithic Karelia, we have samples from Mesolithic Karelia, and they look like unadmixed Eastern European Hunter-Gatherers (EHG).
So again, you seem to be trying to divert attention away from the issue at hand by bringing up irrelevant points and indeed just plain nonsense.
As far as the sharing of mtDNA markers is concerned, you know it well that samples from Iran have been limited and even in that limited dataset, Lazaridis did not disclose the mtDNA from their study.
Face the facts.
Yamnaya and other Bronze Age steppe groups do not show any direct explicit links to Neolithic Iran or even Chalcolithic Iran. The uniparental markers don't match, because there's a striking paucity of South Caspian-specific markers in Bronze Age and modern Eastern Europeans.
Easy enough to see.
https://1.bp.blogspot.com/-YquIP_NuqEA/V8z9vO5RkOI/AAAAAAAAE3E/sAmtYvdSCIEYQqGW3I0l8IK8AURGk-MLQCLcB/s400/Steppe_EMBA_vs_Iran_ChL%252BN.png
Please stop embarrassing yourself here. You need to read the relevant papers more carefully and note the caveats that the authors have expressed, and if you still don't believe me that there's no clear evidence of any population movements from Iran to the steppes, then e-mail the relevant authors and ask them directly whether there is such unambiguous evidence.
@Jaydeepsinh Rathod
Still keeping abay of this Echo chamber.
However just noticing that for such a large, massive, Uber event of migrating people from the steppe to Europe... their cattle didnt really left a big mark in the overall broad region, did they?
Follow metadata or even maps for percentages of cattle Haplotypes (t, t1, t2,t3).
In fact follow percentages of exclusivly north african haplotype T1 and you can even make a trail pretty close to bell beakers dispersal. Check maps of larger BB sites and you will have larger percentage of North African T1. 15% portugal, 12% spain, 10% Italy, 7% france , 8% switzerland, lower on balkans etc... Just a coincidence I supose.
Yes, because obviously migrating cows spread Indo-European languages across Eurasia during the Bronze Age.
Nevermind the lack of any evidence of admixture from Iran on the Bronze Age steppes, but plenty of evidence of a massive expansion of steppe males both into Europe and Asia at the time. They were obviously mutes.
Mooooo!!!
@Davidsy,
sure. We all know that Balanovsky guy, or this Johannes Krause idiotic does not know shit about Ancient Dna. Its only you that have master the subject. That is a paradigm here in Echoland.
Balanovsky stating that Y steppe haplotypes didnt go anywhere and this Krause retard making a map that does not show the steppe as the center of the universe must be abject and a total monkey. Go figure one of this days someone will even dare to say that part of the EHG in CWC didnt come ALL from steppe (was part local or balkan) or that after all Steppe was just kurgan, hence didnt went further than pretty much the western part of Black sea (as archeology shows) .
yeah ... EEECCCCCHHHHHHOOOOOO
The data is speaking for itself.
You can't push shit up a steep hill, neither can Balanovsky or Krause, because it will come rolling back down.
@jijnasu
"Is it possible that the bengali word you mentioned is derived from an alternate form like tRSita or taRSita?"
No the meaning seems different. thirst, tEshTA and trshNA mean exactly the same thing and are nouns and have a retroflex suffix, The form you are suggesting shares the root but seems to be an adjective like thirsty not thirst and has a dental suffix.
for bhrAtar to bhAI etc a possible derivation suggest itself. Panjabi has prA from bhrA an abbreviated form.
the suffix is yA or similar forms is often found as a term of endearment. e.g. bETI, biTiyA (small girl). this trait is even more common in marathi.
similarly mAtA, mA from mAtarah also has the forms mayyA, mAI
Making blanket statements like there is no diversity in NIA ... "experts must have looked" etc serves no purpose Its better to just check little by little. Every small example does not prove or disprove steppe. They form their own reality. Taken as a whole they might start to show a pattern, but no need to jump the gun.
@OM
"That is a paradigm here in Echoland."
David allows for a large amount of disfunctional discussion to nevertheless take place. Counter-opinions pop up as per default on any thread. You yourself are an example of that.
To call this "Echoland" is a grave insult.
Post a Comment