search this blog

Saturday, September 22, 2018

Corded Ware people =/= Proto-Uralics (Tambets et al. 2018)


A new paper on the genetic structure of Uralic-speaking populations has appeared at Genome Biology (see here). It looks to me like the prelude to a forthcoming paleogenetics paper on the same topic that was discussed in the Estonian media recently (see here). Although not exactly ground breaking (because it basically argues what I've been saying at this blog for years, like here and here), it's a very nice effort all round and must be read by anyone with an interest in this topic. From the paper, emphasis is mine:

Background The genetic origins of Uralic speakers from across a vast territory in the temperate zone of North Eurasia have remained elusive. Previous studies have shown contrasting proportions of Eastern and Western Eurasian ancestry in their mitochondrial and Y chromosomal gene pools. While the maternal lineages reflect by and large the geographic background of a given Uralic-speaking population, the frequency of Y chromosomes of Eastern Eurasian origin is distinctively high among European Uralic speakers. The autosomal variation of Uralic speakers, however, has not yet been studied comprehensively.

Results: Here, we present a genome-wide analysis of 15 Uralic-speaking populations which cover all main groups of the linguistic family. We show that contemporary Uralic speakers are genetically very similar to their local geographical neighbours. However, when studying relationships among geographically distant populations, we find that most of the Uralic speakers and some of their neighbours share a genetic component of possibly Siberian origin. Additionally, we show that most Uralic speakers share significantly more genomic segments identity-by-descent with each other than with geographically equidistant speakers of other languages. We find that correlated genome-wide genetic and lexical distances among Uralic speakers suggest co-dispersion of genes and languages. Yet, we do not find long-range genetic ties between Estonians and Hungarians with their linguistic sisters that would distinguish them from their non-Uralic-speaking neighbours.

Conclusions: We show that most Uralic speakers share a distinct ancestry component of likely Siberian origin, which suggests that the spread of Uralic languages involved at least some demic component.

...

Recent aDNA studies have shown that extant European populations draw ancestry form three main migration waves during the Upper Palaeolithic, the Neolithic and Early Bronze Age [2, 3, 45]. The more detailed reconstructions concerning NE Europe up to the Corded Ware culture agree broadly with this scenario and reveal regional differences [65–67]. However, to explain the demographic history of extant NE European populations, we need to invoke a novel genetic component in Europe—the Siberian. The geographic distribution of the main part of this component is likely associated with the spread of Uralic speakers but gene flow from Siberian sources in historic and modern Uralic speakers has been more complex, as revealed also by a recent study of ancient DNA from Fennoscandia and Northwest Russia [68]. Thus, the Siberian component we introduce here is not the perfect but still the current best candidate for the genetic counterpart in the spread of Uralic languages.


Citation...

Tambets et al., Genes reveal traces of common recent demographic history for most of the Uralic-speaking populations, Genome Biology, (2018) 19:139 https://doi.org/10.1186/s13059-018-1522-1

See also...

Indo-European crackpottery

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

Genetic and linguistic structure across space and time in Northern Europe

71 comments:

Kristiina said...

According to Fig. 5 (Circos plots of Globetrotter), there are the following events:
c. 580-1000 AD Maris mix with Turkic speaking Chuvash
c. 730-1140 AD Udmurts mix with Komis and Tatars
c. 650-1300 AD Finns mix with Saamis
c. 1000-1300 AD Saamis mix with Finns
c. 1240-1500 AD Komis mix with a European population and to a smaller extent with Udmurts
c. 1200-1440 AD Khanty-Mansis mix with Mansis and Samoyeds
c. 1440-1580 AD Samoyeds mix with Khanty-Mansis and Central Asians/South Siberians and East Siberians
c. 1500-1700 AD Mansis mix with Komis and a European population and Khanty-Mansis
c. 1500-1700 AD Nganasans mix with Samoyeds and E Asians/ South Siberians and Far Easterners

Admixtures take place earlier in the west than in the east and confirm the movement from west to east of Uralic languages. The Siberian admixture in Finns, Saamis, Udmurts, Maris and Komis is not coming from a common source. The authors seem to propose that Proto-Uralics were Khanty-like, but Globetrotter does not show that Finnics, Saamis or Maris have mixed with Khanty-Mansis and the Khanty-Mansi admixture in Komis and Udmurts is almost non-existant (Globetrotter A). The Globetrotter also shows that Samoyedic languages reached South Siberia before heading north and there was gene flow from South Siberians to Samoyeds.

Davidski said...

@Kristiina

The Globetrotter analysis is based on modern DNA and has been proven in the past by ancient DNA to produce very unreliable results, so I wouldn't read too much into it.

The best we can say from it here is that Uralic speakers have Siberian ancestry that non-Uralic speakers generally lack, unless they come from former Uralic speaking regions.

There's a new software coming from the authors of Globetrotter that apparently does a better job of things, but I also wouldn't take that too seriously if the output is based on modern DNA.

The statement that Uralic languages spread from west to east is bullsh*t and you know it, unless it's heavily qualified.

Tesmos said...

Did anyone would even expect that CWC people spoke some kind of Proto-Uralic language?

Ebizur said...

This paper's supplementary materials include data regarding the Y-DNA of new samples of Selkups and Kets.

First, the (Samoyedic-speaking) Selkups:

Selkups (Tambets et al. 2018)
2/43 = 4.7% C2-M217
3/43 = 7.0% I-M170
3/43 = 7.0% N-M231(xTat/M178)
1/43 = 2.3% N-Tat/M178
25/43 = 58.1% Q-L54
3/43 = 7.0% R1b-M269
6/43 = 14.0% R1a-SRY1532/M198 ("All Selkup R1a individuals belong to European-specific subhg R1a-M458")

There is no really significant difference between this sample and the Karafet team's previously published sample of Selkup Y-DNA, although I do note that this new sample contains a few members of I-M170 and a single member of N-Tat/M178. It is interesting that their R1a is not of the Asian Z93 type, but rather of the European M458 type.

Next, the (Yeniseian-speaking) Kets:

Kets (Tambets et al. 2018)
2/22 = 9.1% N-Tat/M178
18/22 = 81.8% Q-L54
1/22 = 4.5% P-M74(xR1b-M269, R1a-SRY1532/M198, Q-L54)
1/22 = 4.5% R1a-SRY1532/M198

Likewise, not much difference from previously published samples of Ket Y-DNA. Most extant Kets belong to the L330+ subclade of Q-L54, while a few belong to other haplogroups, including C2-M217, N-Tat, N-P43, and R1a-SRY1532/M198. One of the Kets in this sample appears to belong to some unusual (for Western Siberia) sort of P-M74, perhaps R1b(xM269) or Q(xL54); note that the P-M74 total in Table S5 is 86.4% (i.e. 19/22), but the numbers for the Ket Q-L54 haplotypes in Table S16 only add up to n=18, so, barring some sort of error, the implication is that one of this study's Kets belongs to P-M74(xR1b-M269, R1a-SRY1532/M198, Q-L54).

Davidski said...

Yeah, the R1a in Uralics is rarely Z93+, so it's not linked to the Sintashta expansion.

It's mostly from Balto-Slavic admixture and subsequent recent founder effects.

But try telling that to Carlos Quiles and his crackpot fans.

Slumbery said...

@Kristiina

These dates are all very late, so they are not relevant to the original spread of Uralic languages.

Nevertheless, even with the new data I stick to my earlier opinion that the pre-dispersion Uralics were probably not or not much more Siberian that the more Siberian subsets of modern Volga-Ural Uralic speaking populations (like Udmurts). This still means however that _in the West_ Siberian genetic ancestry can be linked to the spread of Uralic.
Earlier here the Nganasan were proposed by somebody as a good proxy for ancient Uralics. With the new data that does seem even more unlikely (look at the admixture graph, how the component they think as the most Uralic behaves around Nganasan).

Interesting that Karelians have some 25% direct EHG ancestry above of what comes from Steppe/CWC.

Shaikorth said...

Khantys which this study uses as their K=9 component have some R1a but based on Global25 doesn't look like they have Balto-Slavic admixture, or even CWC admixture from available samples.
Global25 nMonte (scaled, pen=0)
Khanty

West_Siberia_N,44.6
Shamanka_N,44.2
Barcin_N,11.2

Russian_Orel,0
CWC_Baltic,0
Ukrainian,0

Compare to Yukaghir_Forest which does have Balto-Slavic admixture:
Yukagir_Forest

Shamanka_N,48.8
Russian_Orel,44.2
CWC_Baltic,7

Ukrainian,0
Barcin_N,0
West_Siberia_N,0

Some individual Kets & Selkups in the set have modern Russian ancestry too, some don't.

@Slumbery, no ancient DNA in this study but a little test with Global25, what do "Finland-shifted" Sigtuna_84005, which doesn't seem to have AncSaami admixture, and Khanties have in common that CWC-related stuff in Scandinavia and later Anglo-Saxons do not have?



England_Anglo-Saxon

Nordic_LN,98.6
Barcin_N,1.4

Sintashta_MLBA_o3,8
Shamanka_N,0


Sweden_Viking_Age:Sigtuna_84005

Nordic_LN,92
Sintashta_MLBA_o3,8

Barcin_N,0
Shamanka_N,0


Khanty

Shamanka_N,48.2
Sintashta_MLBA_o3,46
Barcin_N,5.8

Nordic_LN,0

Of course without more ancient DNA it's impossible to say whether this is right...

Shaikorth said...

Oops, a correction: the Anglo-Saxon of course has 0% Sintashta_o3 in that fit.

Anthony Haken said...

Unfortunately Carlos is still pursuing the R1a/CWC=Uralic narrative.

Unrelated to that, this study is saying what we have known for a long time. Uralic speakers are closest genetically with their neighbors while while almost all have Siberian, EHG components and N1c.

IMO the best candidate's we have currently are BOO or WSHG as being closest to Proto-Uralics and spread via Seima-Turbino seems most likely.

Slumbery said...

@Shaikorth

I am not sure what you are trying to say. (I am not saying this as some argument, I am really not.)
Your short Global25 test (as far as it goes) shows no special relatedness between Khanty and Sigtuna_84005 that is not common with later Anglo_Saxons too. Either ancient Saami did not have ancestry from Baikal Neolithic or Sigtuna_84005 had only a too low level AncSaami ancestry to detect that.

BTW, just for clarification: when I say "original" Uralics were not as Siberian as modern Siberian Uralic speakers, I mean what "Siberian" it means in the current population, not that they could not have more (more than modern Volga-Ural populations) ancestry from the Geographical Siberia from some time depth.

Ebizur said...

I am a bit surprised by how much of the mtDNA of even the Kets appears to be of Western origin or at least broadly Western Eurasian affinity. Two thirds (44/66) of the present study's sample of Kets belong to Western Eurasian mtDNA haplogroups:

11/66 = 16.7% U5a1
11/66 = 16.7% U4a1
7/66 = 10.6% U4
9/66 = 13.6% H
4/66 = 6.1% N2a
1/66 = 1.5% J
1/66 = 1.5% I5a

An approximately equal percentage of this study's sample of Selkups belongs to Western Eurasian mtDNA haplogroups: 74/120 = 61.7%.

For the other Siberian Uralic-speaking populations, the percentages of Western Eurasian mtDNA are as follows:

Khanty 30/42 = 71.4%
Mansi 26/38 = 68.4%
Nenets 32/79 = 40.5%

Just to the west of the Ural Mountains in the northeastern corner of Europe, the Komi exhibit an even higher level of Western Eurasian mtDNA (as one might expect): 110/130 = 84.6%.

So, at least in regards to their mtDNA, Uralic-speaking populations (with the exception of Nenets and almost certainly also Nganasan) have predominantly Western Eurasian affinities.

The Kets are perplexing, though. If the Uralic language family is assumed to have been spread predominantly by males migrating from east to west, which population movement might have brought Western Eurasian females into the basin of the Yenisei River in central Siberia? Do any of the Kets' Western Eurasian mtDNA haplogroups look like they might plausibly be attributed to ANE people?

Grizzlor said...

I find it interesting that the study suggests that the most northeasterly population of Europe, the Kola Saami, have the most WHG like ancestry of all the modern populations

Shaikorth said...

@Slumbery
There is something, I made a copypasting error with the Anglo-Saxon and assigned it o3 it didn't have. Ancient Saamis are very similar to modern ones for the record.

The same test expanded with WHG and EHG, to make sure Global25 wasn't just using Sintashta_o3 as a proxy for any extra HG ancestry relative to Nordic_LN:

England_Anglo-Saxon

Nordic_LN,98.6
Barcin_N,1.4

Sintashta_MLBA_o3,0
Shamanka_N,0
EHG:Sidelkino,0
WHG,0

Sweden_Viking_Age:Sigtuna_84005

Nordic_LN,92
Sintashta_MLBA_o3,8

Shamanka_N,0
EHG:Sidelkino,0
Barcin_N,0
WHG,0

Khanty

Shamanka_N,48.2
Sintashta_MLBA_o3,46
Barcin_N,5.8

Nordic_LN,0
EHG:Sidelkino,0
WHG,0

Nirjhar007 said...

@ Dave

No offence but it does does seem clear that the dominant ancestry in Uralics, apart from outliers , is CWC. And in contrast to the uniform R1a-Z283 seen in Uralics, the N1 Clades found are divergent, local founder effects. Also Haakinen has demonstrated that Samoyed is not the most deeply divergent Uralic language, but sits symmetrically within a Uralic tree. All in all, a western origin of Uralic, linked to eastern late CWC seems rather well founded.
QED.

Davidski said...

@Nirjhar

The R1a in Uralics can't be from eastern Corded Ware, because it's mostly the result of young founder effects of Balto-Slavic clades and so it has nothing to do with Sintashta, Srubnaya, etc.

On the other hand, the N1 in Uralics is mostly N-L1026, which has the right age and structure to be the paternal signal of the Uralic expansion.

Quit making a fool of yourself. You've been doing it for years. Time to stop.

Kristiina said...

I agree with Nirj. Mesolithic N west of the Urals was probably autosomally WSHG/EHG + SIBERIAN. However, there is no need that Proto-Uralic was spoken by yDNA N. It could have been spoken by another yDNA OR a pool of N and other western yDNAs. IMO, it is not probable that N haplos in Mesolithic Siberia spoke a Uralic language. They probably spoke languages that bore variable resemblance to modern Uralic varieties.

Do you notice that Selkups and Kets with 62% and 66% of Western mtDNA, score 70% and 74% Siberian while Nenets with only 41% of western mtDNA score only slightly more Siberian than Kets, i.e. 79%. This difference between Kets and Nenets must be due to yDNA: Kets are only 9% N-Tat and 91% Q+R (including 82% Q-L54) while Nenets are 57% N(xN-Tat), 41% N-Tat, 1.4% P and 0.7% O.

@Grizzlor The only 100% sure ancient Saami sample is from Kola Peninsula:
Chalmny Varre Murmansk 18-19th cent CHV002 yDNA I2a1, mtDNA V7a1
I2a1 is of Mesolithic Scandinavian origin and we know that Scandinavian I2a1 hunters were a mixture of WHG and EHG.

Davidski said...

@Kristiina

I agree with Nirj.

So obviously you just like agreeing with him and don't care that his statements are contradicted by the current and forthcoming data, which show that...

- the R1a in Uralics is not from eastern Corded Ware or Sintashta, it's just mostly Balto-Slavic

- remains from likely early Uralic Tarand graves belong to N1c not to the "Sintashta" R1a-Z93

- these remains from the Tarand graves also don't resemble Corded Ware people in terms of genome-wide structure, because they have too much hunter-gatherer ancestry and apparently also Siberian admixture


It seems a bit much to make a fool of yourself in public just because you like agreeing with Nirjhar, no?

Slumbery said...

@Shaikorth

Ah, I see. I actually find it to be likely that some Shintashta outliers were Uralic. Shintashta is a possible source of some old loanwords in Hungarian. Of course this test is a bit too indirect to actually prove that.

Huck Finn said...

@Kristiina: If WSHG was involved then I'm not sure to what extent there's a need for extra Siberian, as WSHG is sort of eastern as such?

Kristiina said...

@Huck Yes, we need ancient DNA to clarify things. However, the amount of Siberian seems to increase towards the north, i.e. in the same longitude there is more Siberian in the higher latitudes. This means that the east to west flows of yDNA Q and N during the Mesolithic were probably ANE/Siberian while the Mesolithic flow of R from south to North was EHG.

The northern latitudes are not the easiest land to make a living. I pay homage to Uralics and Saamis in particular for their strength and resilience and respect of ancient traditions and their inventiveness in making their own mix of new and old:

https://www.youtube.com/watch?v=-qA7GDkKZJk

https://www.youtube.com/watch?v=CyA64m_p3PY

Nirjhar007 said...

@ David

the R1a in Uralics is not from eastern Corded Ware or Sintashta, it's just mostly Balto-Slavic

So it is all from after c. 800 AD (the Russification) ? .We shall see with aDNA from the forest zone .
Until then, maybe you, Mikkel and then gang should just chill & be open minded :) .

Davidski said...

@Nirjhar

So it is all from after c. 800 AD (the Russification)?

You really shouldn't share your opinion about things that you have no clue about.

There were Balts in the forest steppe and as far east as the Volga long before the Russification of the region.

There are only minor instances of non-Balto-Slavic R1a in Uralics, and most of these aren't associated with eastern Corded Ware/Sintashta, but with the hunter-gatherers of the region, like R1a-­YP1272.

Corded Ware and R1a-Z645 came from the steppe, not from the forest zone. Uralics and their N-L1026 came from the forest zone, and they weren't like Corded Ware people.

This is what you'll see in the aDNA from the Tarand graves.

Nirjhar007 said...

@ David
Hmm..

“The Z283 in the forest zone can’t be from late, east CWC because they’re “Balto-Slavic” .
Just study that for a second ..

Nirjhar007 said...

This is what you'll see in the aDNA from the Tarand graves.

so what about them ?.

They’re a local, late phenomenon.

Kristiina said...

This map shows the extent of Baltic hydronyms and the probable extent of Baltic languages:
https://historum.com/threads/history-of-the-balts.105388/

This map shows the area of Baltic hydronyms, the area of probably Balto-Slavic-speaking Trzciniec culture and the area of Western Uralic languages:
https://imgur.com/a/ByH0W8w

Davidski said...

@Nirjhar

Not sure what you're mumbling about now, but Balto-Slavs don't come from eastern Corded Ware. This is what Sintashta is supposed to derive from and its R1a is 100% Z93.

So tell us how much Sintashta-derived R1a-Z283 is in Uralics?

EastPole said...

@Nirjhar007
I don’t follow you.
Are you suggesting that Indo-Iranians R1a-Z93 are just slavicized Uralics?
What is your theory? Could you spell it out in more detail.

Davidski said...

@Kristiina

I don't have time to play picture games with you. Different maps by different people show different things, but it's a fact that Baltic peoples lived much further east than they do now.

And they didn't have to live near the Urals for some of their Y-chromosome lines to spread there and beyond before the Russian era.

Fact is, most Uralic R1a is a young subset of typically European, Balto-Slavic R1a lineages, and not Z93, which it should be if Uralics were in large part derived from really eastern Corded Ware and/or Sintashta.

Stop embarrassing yourself.

Davidski said...

@Nirjhar

so what about them ?.

They’re a local, late phenomenon.


LOL

Tarand graves are seen as intrusive in the Baltic region from what is now Russia, and associated with early Uralic speakers.

The remains from the Tarand graves are the earliest to date to show N1c in the East Baltic.

Kristiina said...

No. Look at the map I posted. The oldest N1c in the area is from Serteya site in Smolensk belonging to the Zhizhitskaya culture (c. 3000-2000 BC). That N1c sample is marked on the map.

Nirjhar007 said...

@David

I agree they’ll likely show N1c.
But Tarand graves date to the Roman era - too late for FU !.Moreover they spread from coast inland, not from Russia.
They are an elaboration of Baltic-Finnic traditions...

Davidski said...

@Kristiina

You need to stop grasping at straws.

First of all, you're referring to a result from an unreliable PCR test, which is very likely due to contamination.

Secondly, Smolensk isn't in the East Baltic.

Thirdly, I was referring to the upcoming time transect ancient DNA study of Estonia, which shows that N1c first appears in the East Baltic in remains from Tarand graves, which are generally accepted to be intrusive to the region from Russia and associated with a wave of early Finnic speakers.

You just seem to have made up your mind to argue against all of these facts. It's a waste of time discussing this with you. Go and bother the authors of the new paper when it comes out, see how far you get.

There's always kooky Carlos' blog if you get bored after that.

Nirjhar007 said...

@ David
I agree they’ll likely show N1c .

But Tarand graves date to the Roman era - too late for FU !.
Moreover, they spread from coast -> inland, not from Russia!. They are an elaboration of Baltic-Finnic traditions...

Davidski said...

@Nirjhar

You're not making any sense and ignoring the obvious facts, like you usually do.

The Tarand Grave Culture came to the East Baltic from the east with people who were rich in N1c and are seen as early Baltic Finns.

Connect the dots. Duh.

I honestly think that you have some serious mental problems.

Kristiina said...

@ Nirj

I agree that there is no common N clade that would be shared by the Uralic populations. It is just as you said divergent, local founder effects dated to the late Bronze Age. Therefore, it is possible that R1a1-M558 in Uralics will be proven of Fatyanovo origin and Proto-Uralic developed as a mixture of Fatyanovo and the earlier Volosovo tradition.

EastPole said...

@Davidski
“@Nirjhar
I honestly think that you have some serious mental problems.”

I wouldn’t call it mental problem, it is something else. And the same applies to Carlos and Kristina.
Nirjhar, Carlos and Kristina are very intelligent, knowledgeable and mentally healthy people. I have no doubt about it.
What they are experiencing is mental stress. This kind of mental stress occurs temporarily when people realize that what they have been working on for years and the ideas they have devoted substantial portions of their lives to are not entirely correct.
Carlos for example studied in Germany. And now he sees that some foundations of western historical sciences like cephalic index, linguistic reconstructions, satem-centum division, glottochronology etc. are just BS. Modern science does not seem to support those old theories. He is now trying to connect the dots, but works under stress and has his mind temporarily clouded. I am sure he and others will get over it and will see the sun.

Davidski said...

If anyone missed it, here's the abstract from the Estonian aDNA paper talk that was presented at the ISBA 2018 conference recently...

Demographic processes in the territory of Estonia from the earliest inhabitants to modern times

Tambets et al.

This interdisciplinary project deals with the studies of temporal population dynamics of the eastern coast of the Baltic Sea, in the territory of present-day Estonia. We use the skeletal material from Estonian archaeological collections to characterize the genetic structure of the population in time series starting from the earliest layers of lithic cultures to the contemporary population. The sample consisted of 72 individuals – 24 from the Bronze Age stone-cist graves, 13 from the Iron Age tarand-graves and 35 from the Medieval rural and town cemeteries. We produced low-coverage Illumina whole-genome sequencing data. The resulting data was analyzed in a context of modern Estonian and European genetic variation. Hgs N3 and R1a are the two most common chrY hgs among modern Estonians. While we have previously found that hg R1a appears in Estonia together with farmers of Neolithic Corded Ware culture (CWC) people, the arrival of hg N, which has been proposed to be connected with the arrival of Uralic languages to Europe, is yet to be studied. We found that the Iron Age individuals do in fact carry chrY hg N3 while all 18 Bronze Age males belong to R1a. Furthermore, based on their autosomal data, all of the studied individuals appear closer to hunter-gatherers and modern Estonians than Estonian CWC individuals do. The Medieval period started in the eastern Baltic region much later than in Central Europe and in Scandinavia. The crusades and conquest in 13th century AD brought along vast social, economical and cultural changes, which presumably changed the structure of the local population. While the Medieval individuals buried in rural cemeteries are considered as the representatives of the local Estonian population, those of big towns can often be associated with the new wave of people who arrived, mostly from Western Europe, together with Christianity via the economical, cultural and political networks. We find that there is a clear difference between the genome-wide data of individuals belonging to Medieval urban and rural communities. The urban elite clusters genetically with modern Germans but the rural local class with modern Estonians. We did find a few individuals of mixed genetic ancestry, but the overall admixture between the two classes was limited. Our results reveal several population shifts during the prehistory of the region and show a clear continuity of the population starting at least from the Iron Age.

zardos said...

I wonder about the amount of y-haplogroup N even in modern Balts, especially Lithuanians, with their respective Siberian component being so low. Assuming N was introduced and spread with the Siberian related genetic component, it was really thinned out when reaching North Eastern Europe I guess. But even then, the proportion of N in Balts is even much higher relative to the Siberian component than in the neighbouring Finno-Ugrian ethnicities.

Davidski said...

@EastPole

Carlos is more than stressed, I'd say. But in any case, what I think happened is that he thought he figured everything out when he saw that Karelia_HG belonged to R1a and eastern Yamnaya to R1b.

Of course, he didn't understand the data, because the Karelia_HG R1a is irrelevant to Corded Ware R1a and the vast majority of modern R1a.

But instead of admitting that he was wrong, he kept digging a hole for himself, and now he can't climb out of it without admitting that he doesn't know what he's doing.

Davidski said...

@zardos

There's nothing remarkable about the frequency of N in Balts. It's that high due to a series of recent founder effects.

And these founder effects basically decoupled this lineage in Balts from autosomal Siberian ancestry.

Anthony Haken said...

This study states N1c’s expansion was likely sex biased and demic much like R1b and R1a with IE. I agree with Kristina about Fatyanovo being R1a however I don’t see why it should be connected with PU despite giving them IE loan words and Steppe autosomal DNA.

Carlo’s theory about Siberian’s carrying N1c somehow spreading their genes across all Uralic people’s through recent founder effects makes very little sense to me. Uralic speakers carry N-L1026 which has a TMRCA congruent with PU. Founder effects and bottlenecks have certainly occurred but the fact that N1c has been favoured each time should be enough to disprove PU were half R1a or something.

@Nirjhar007
Tarand graves are not PU but are instead probably proto-Finnic, ancestors of Saami were already in Finland at this time. It’s rise to prominence around the Baltic probably has something to do with metal working.

Shaikorth said...

@Zardos
Northern Baltic region, namely Finns and Karelians have more Siberian (as does the northern part of Estonia via post-bottleneck Finnish admixture) because they've assimilated Saamis post Tarand Grave period. See this abstract and graphs:

https://pbs.twimg.com/media/DndcErtW4AAzfVr.jpg:large
https://pbs.twimg.com/media/DY50sMUW4AApfKb.jpg
https://i.imgur.com/6qhameW.jpg
https://i.imgur.com/OhdE6JB.png

Basically Southern Finland was populated by Saamis, with Finns not yet existing, for hundreds of years after N1c1, Tarand Graves and Baltic Finnic language came to Estonia. Balt-like autosomal composition with N1c1 is expected in the Southern Baltic, Baltic_IA is a decent example.

Anthony Haken said...

@Shaikorth
In this interview Kristiina Tambets says Siberian admixture does appear during the iron age, presumably with Tarand graves although it is hard to understand using a translator.
https://www.linnaleht.ee/885967/kes-me-eestlased-oleme-ja-kust-me-parineme

This Siberian also does not seem to show in leaked PCAs with the Tarand samples so that leaves me a little confused.

Wastrel said...

Tentatively and respectfully, at risk of directing some of the abuse my way too, I just have to question a few things here.

I don't understand how anyone can claim to know that the Balts once stretched to the Volga. Perhaps if I lay out my thinking on this, someone can point out where I'm wrong?

On the one hand:
- Baltic languages are in most respects highly conservative - certainly the most conservative IE languages. If we're talking about 'Baltic' languages spoken a thousand, or two thousand, years before the modern day, we'll be talking about an IE dialect that is even MORE conservative than modern Baltic languages. This means that that dialect will not look all that different from late IE itself.

- there are no attestations of the language of people east of the Balts until very recently (other than a couple of Iranians a long way south). Indeed, even the Baltic languages aren't attested until the 14th century. So there can be no direct evidence of WHAT they spoke.

- indirect evidence in the form of surviving placenames is an extremely blunt tool to identify historic languages of an area, because those placenames get filtered through centuries of re-analysis by speakers of the new dominant language. It's often possible to identify approximate elements, but because all the fine detail has been lost over time, it's extremely hard to be clear on a particular ancestor. For instance, if you look at placenames like Basildon, Langdon, London, Yverdon, Verdun, and even Olten and Thun, it's plain to see that these are all areas that share the common Germanic placename element, -don (sometimes -ton, etc). Except: they aren't. Basildon and Langdon have Germanic -don. But Verdun and Yverdon, and Olten and Thun, have CELTIC -don. Which ultimately comes from the same PIE word with more or less the same meaning - but despite English and Gaulish coming from entirely different IE subfamilies, the words didn't diverge sufficiently to still be distinguishable out of context after the sundry distortions that placenames undergo. [And 'London' certainly looks like it must have either the Germanic or the Celtic -don naming element... except that we happen to know the Romans called it Londi:nium, which makes it very unlikely to come from either element; so we don't know where it's from; conversely, Lyon and Nevers both DO have the Celtic -don element, but so obscured by soundchanges that we couldn't possibly know that if we didn't know what these towns were called two thousand years ago.]

- So in distinguishing "early Balts" from "a non-Baltic IE people" - particularly a non-Baltic Balto-Slavic group, or a non-Balto-Slavic group closely related to the Balto-Slavic branch - we're having to spot really small distinctions. And in doing it through placenames as conveyed through a thousand years of Slavic speakers, wer'e having to spot that really small distinction in a dataset out of which all fine distinctions have been removed with a hammer.

- In particular, a lot of the changes that would most clearly identify something as Baltic - i.e. the development of the Baltic accent systems - is information that would immediately be lost by transference into a non-Baltic population.

- So how can you possibly make that call with confidence?

- Indeed, it's not clear that the call is even MEANINGFUL. It's highly controversial whether "Baltic" is even a meaningful category in a historical sense, other than just meaning "Balto-Slavic people who survived this long but aren't Slavs". That is, it's not clear that the Baltic languages are more closely related to each other than they are to Slavic, which would make 'Baltic' paraphyletic.

Wastrel said...



I'm also confused by your certainty that Balto-Slavs can't be from eastern CWC. Where are they from then? Western CWC? They're not Sintashta, sure. But since they occupy the eastern part of CWC territory, and there's no particular reason to think they came from west of there, surely, "eastern CWC" must be the default assumption? Even if they're not East-MOST CWC (i.e. Sintashta). But then again, just because Sintashta ended up to the east doesn't mean they started out there. They could well have been 'southern' rather than 'eastern' CWC, who just migrated east through the steppe.


I'm not saying you're necessarily wrong about anything. But with respect, I think you're leaping to some conclusions without properly laying the groundwork for them.

Davidski said...

@Wastrel

I meant the Don River. In any case, Balts stretched very far to the east.

https://en.wikipedia.org/wiki/Balts#/media/File:East_europe_3-4cc.png

And in regards to whether Balto-Slavs came from eastern Corded Ware, obviously eastern Corded Ware in this context is far eastern Corded Ware.

You're free to believe that Balto-Slavs may have come from somewhere near the Samara.

Shaikorth said...

@Anthony Haken
Have to wait for the paper to be sure, the way it's worded in the interview could also mean N1c, which is mentioned in the abstract and here: https://opetajateseminar.files.wordpress.com/2018/03/horisont_2018-1_dna.pdf. When it comes to modern Estonians autosomally, things strongly look like what I said.

Wastrel said...

@david, well, I do see your map, yes. But I don't understand what evidence you might have that would give you any confidence in it.

I must say, the degree of confidence with which I would regard the speculative work of 1950s Soviet linguists is exceptionally slight. And 'discovered several hundred placenames of possibly Baltic origin' has somehow transformed into 'definite Baltic presence' remarkably easily, given that you could discover several hundred 'possibly' Baltic placenames by throwing darts randomly at a map. We might also point out that the wikipedia article that map is from is filled from top to bottom with [citation needed]s, and that the article on Baltic languages notes that the range of the INFLUENCE of East Baltic "once possibly" reached to the Urals, "but this hypothesis has been questioned"; honestly, I'd rather go with modern rigour and skepticism rather than Soviet ideologies and Gimbutas.

Your article does mention the idea of Baltic archeological cultures, which is at least something solid to go on; but given that these cultures were all extinct hundreds of years before their languages could be recorded, I really don't see how a settlement's language is to be discerned from its archaeology. Any similarities between adjacent cultures in northern European Russia could be explained by a combination of common environment, areal influence, and shared ancestry, without requiring any particular language family to be present.

Which isn't to say that the pre-Slavic inhabitants of the region were NOT Balts (i.e. members of a linguistic clade closer to both East and West Baltic than to Slavic). Given that East Baltic languages are nearby and there are no other surviving families to associate them with, it's as reasonable a guess as any other. But I can't see it as reasonably to treat it as any more than a guess. There just isn't the evidence to support any confidence in this regard.

At least, I'm not aware of it, or sure what it could be. But you may be!

Davidski said...

@Wastrel

As I said to Kristiina, I don't care to swap maps and notes about the precise geographic extent of Balto-Slavic peoples.

The point I was making was that they were already deep in Russia prior to the Russian expansions east, which can be an explanation for the presence of Balto-Slavic specific R1a clades in many Uralic speaking groups.

The reason I actually claimed that Balts ranged all the way to the Volga is because I'm aware of R1a clades all the way up there in the indigenous groups that look Baltic and Slavic. Strictly speaking I should have said they ranged all the way up to what is now Moscow, but their DNA is now found further east than that. Oops.

So you might now like to find someone else to discuss whatever it is that you want to discuss, because I'm not that person.

Ric Hern said...

How did Komis, Mordvins and Udmurts end up with roughly the same percentage (11-13%) of R1b in their populations ?

Slumbery said...

@Ric Herm

According to the file attached to article it is 13.4% for Mordovians, 4.9% for Udmurts and 5.9% for Komis. I can see no mysterious exact matching that should be explained.

The source is impossible to tell from the data presented, because there is no deeper resolution than "R1b". Mezhovskaya could be a good guess if it is Z2103.

Ebizur said...

Slumbery wrote,

"According to the file attached to article it is 13.4% for Mordovians, 4.9% for Udmurts and 5.9% for Komis. I can see no mysterious exact matching that should be explained."

More importantly, each sample of each population contains different percentages of this haplogroup:

Udmurt (Semino et al. 2000)
2/43 = 4.7% Eu4 E-M35
3/43 = 7.0% Eu7 I-M170(xM26)
2/43 = 4.7% Eu10 F-M89(xI-M170, J2-M172, G-M201, H-M69, K-M9)
1/43 = 2.3% Eu13 N-Tat(xM178)
12/43 = 27.9% Eu14 N-M178
2/43 = 4.7% Eu16 K-M9(xN-Tat, T-M70, L-M11, P-M45)
5/43 = 11.6% Eu18 R1-M173(xM17)
16/43 = 37.2% Eu19 R1a-M17

Udmurt (Tambets et al. 2004)
1/87 = 1.1% I
2/87 = 2.3% R1b
9/87 = 10.3% R1a
49/87 = 56.3% N3
25/87 = 28.7% N2
(1/87 = 1.1% Other) [This individual may be inferred to belong to P-M74(xQ, R1b-M269, R1a-SRY1532/M198).]

Udmurt (Semino et al. 2000 + Tambets et al. 2004 + Tambets et al. 2018, pooled)
2/184 = 1.1% E-M35
2/184 = 1.1% F-M89(xI-M170, J2-M172, G-M201, H-M69, K-M9)
4/184 = 2.2% I-M170
2/184 = 1.1% K-M9(xN-Tat, T-M70, L-M11, P-M45)
31/184 = 16.8% N-M231(xTat/M178)
98/184 = 53.3% N-Tat/M178
1/184 = 0.5% P-M74(xR1b-M269, R1a-SRY1532/M198)
9/184 = 4.9% R1b-M269
35/184 = 19.0% R1a-SRY1532/M198

Udmurt (Tambets et al. 2018, inferred)
6/54 = 11.1% N-M231(xTat/M178)
36/54 = 66.7% N-Tat/M178
10/54 = 18.5% R1a-SRY1532/M198
2/54 = 3.7% R1b-M269

Each of these samples of Udmurts is quite different from the others. Note the high frequency of N2 (i.e. N-P43; 25/87 = 28.7% N2) in the sample analyzed by Tambets et al. 2004 versus the low frequency of potential N-P43 (2/43 = 4.7% Eu16 K-M9(xN-Tat, T-M70, L-M11, P-M45)) in the sample analyzed by Semino et al. 2000. One loses these distinctions when one pools the samples together as Tambets et al. 2018 have done.

Davidski said...

@Nirjhar

It is also recognised by linguists that Balto-Slavic as well as Germanic have Uralic substrate, as per Shcriver and Thomasson & Kaufman respectively.

No it's not. There's no consensus like that at all.

If there is Uralic influence in Germanic and Slavic, then it has to be explained by a different way than by Corded Ware being Proto-Uralic, because the modern and ancient DNA evidence is unequivocal that Corded Ware people could not have been Proto-Uralics.

A new population moved into Northeastern Europe during the Bronze Age that had cultural traditions distinct from the earlier Corded Ware people and were rich in N1c. These were the early Uralic speakers in the region, and ancient DNA is starting to map out their migrations in detail.

Enjoy...

Demographic processes in the territory of Estonia from the earliest inhabitants to modern times

Tambets et al.

This interdisciplinary project deals with the studies of temporal population dynamics of the eastern coast of the Baltic Sea, in the territory of present-day Estonia. We use the skeletal material from Estonian archaeological collections to characterize the genetic structure of the population in time series starting from the earliest layers of lithic cultures to the contemporary population. The sample consisted of 72 individuals – 24 from the Bronze Age stone-cist graves, 13 from the Iron Age tarand-graves and 35 from the Medieval rural and town cemeteries. We produced low-coverage Illumina whole-genome sequencing data. The resulting data was analyzed in a context of modern Estonian and European genetic variation. Hgs N3 and R1a are the two most common chrY hgs among modern Estonians. While we have previously found that hg R1a appears in Estonia together with farmers of Neolithic Corded Ware culture (CWC) people, the arrival of hg N, which has been proposed to be connected with the arrival of Uralic languages to Europe, is yet to be studied. We found that the Iron Age individuals do in fact carry chrY hg N3 while all 18 Bronze Age males belong to R1a. Furthermore, based on their autosomal data, all of the studied individuals appear closer to hunter-gatherers and modern Estonians than Estonian CWC individuals do. The Medieval period started in the eastern Baltic region much later than in Central Europe and in Scandinavia. The crusades and conquest in 13th century AD brought along vast social, economical and cultural changes, which presumably changed the structure of the local population. While the Medieval individuals buried in rural cemeteries are considered as the representatives of the local Estonian population, those of big towns can often be associated with the new wave of people who arrived, mostly from Western Europe, together with Christianity via the economical, cultural and political networks. We find that there is a clear difference between the genome-wide data of individuals belonging to Medieval urban and rural communities. The urban elite clusters genetically with modern Germans but the rural local class with modern Estonians. We did find a few individuals of mixed genetic ancestry, but the overall admixture between the two classes was limited. Our results reveal several population shifts during the prehistory of the region and show a clear continuity of the population starting at least from the Iron Age.

Mouthful said...

@Wastrel

"There just isn't the evidence to support any confidence in this regard."

Mordvinic branch has clear Baltic loanwords although not as many as Finnic branch does and Erzya and Moksha certainly didn't live anywhere near near South Baltic where actual Balts live now. According to Riho Gr√ľnthal earliest Baltic contacts appeared in Pre-Mordvinic stage and then there was Late Baltic influence during Proto-Mordivinic stage.

Nirjhar007 said...

@ David

When it comes to substratum, there cannot be consensus, because we are working with indirect evidence. The quoted authors are leaders in their field, so we will not dismiss their views on the basis of yours, no offence. But I don't think anybody here proposed any simplistic CWC = Uralic , did they ?.
Anyhow, thanks for the reference, I look forward to that study.

Ebizur said...

Underhill et al. 2010
Komis-permyans 22/60 = 36.7% R1a1a(xM458)
Tatars from Tatarstan 23/66 = 34.8% R1a1a(xM458)
Bashkirs 36/126 = 28.6% R1a1a(xM458)
Maris 13/49 = 26.5% R1a1a(xM458)
Chuvashis 30/117 = 25.6% R1a1a(xM458)
Udmurts 23/141 = 16.3% R1a1a(xM458)
Tatars from Bashkortostan 5/39 = 12.8% R1a1a(xM458)
Komis 1/52 = 1.9% R1a1a(xM458)

Underhill et al. 2010
Tatars from Tatarstan 3/66 = 4.5% R1a1a7-M458
Komis 2/52 = 3.8% R1a1a7-M458
Tatars from Bashkortostan 1/39 = 2.6% R1a1a7-M458
Udmurts 3/141 = 2.1% R1a1a7-M458
Maris 0/49 R1a1a7-M458
Komis-permyans 0/60 R1a1a7-M458
Chuvashis 0/117 R1a1a7-M458
Bashkirs 0/126 R1a1a7-M458

Note the huge difference between the above study's two samples of Komi.

It appears that Uralic and Turkic peoples of the circum-Uralic region do not possess a great deal of the typically Balto-Slavic R1a-M458 despite the previously noted presence of R1a-M458 among the Selkups. Does anyone know which subclades of R1a are common among the circum-Uralic populations? Do significant numbers of them belong to R1a-Z93?

Myres et al. 2011
Chuvashes
2/117 = 1.7% R1b-L23(xM412)
1/117 = 0.85% R1b-U106(xU198)

Komis (Perm Oblast, Russia)
7/61 = 11.5% R1b-L23(xM412)
1/61 = 1.6% R1b-S116(xM529, U152)

Udmurts
2/54 = 3.7% R1b-L23(xM412)

Tatars (Kazan, Russia)
1/80 = 1.25% R1b-M343(xM73, M269) [perhaps also xV88]
1/80 = 1.25% R1b-M73
2/80 = 2.5% R1b-L23(xM412)
2/80 = 2.5% R1b-U106(xU198)
1/80 = 1.25% R1b-S116(xM529, U152)

Tatars (Bashkortostan, Russia)
1/39 = 2.6% R2-M479(xR2a-M124)
1/39 = 2.6% R1b-M73
3/39 = 7.7% R1b-L23(xM412)
1/39 = 2.6% R1b-U106(xU198)

Bashkirs South-east (Bashkortostan, Russia)
77/329 = 23.4% R1b-M73
8/329 = 2.4% R1b-M269(xL23)
106/329 = 32.2% R1b-L23(xM412)
1/329 = 0.3% R1b-U106(xU198)
1/329 = 0.3% R1b-S116(xM529, U152)
2/329 = 0.6% R1b-U152

Bashkirs Western (Bashkortostan, Russia)
0/54 R-M207(xR1a-SRY10831,M17)

Bashkirs South-west (Bashkortostan, Russia)
1/54 = 1.9% R1b-M73
9/54 = 16.7% R1b-L23(xM412)

Bashkirs South (Bashkortostan, Russia)
2/79 = 2.5% R1b-M269(xL23)
9/79 = 11.4% R1b-L23(xM412)
1/79 = 1.3% R1b-U152

Bashkirs North (Bashkortostan, Russia)
1/70 = 1.4% R1b-M73
2/70 = 2.9% R1b-L23(xM412)
50/70 = 71.4% R1b-U152

The subclade of R1b that appears in non-negligible proportions of the greatest number of populations in the circum-Uralic region is R1b-L23(xM412), which should be approximately equivalent to R1b-Z2103 according to YFull YTree v6.05.10. However, other subclades are also present, notably R1b-M73 in Tatars and Bashkirs and R1b-U152 in Northern Bashkirs.

Davidski said...

@Ebizur

Practically all R1a-Z280 subclades are typically Balto-Slavic, and this is what most Uralic groups have.

Ric Hern said...

@ Ebizur

That R1b-U152 in Bashkirs North is interesting. Isn't it mostly associated with the Alpine Region in Europe ? Wonder how and when Bashkirs in that region attained a 71.4% ?

Ryan said...

Am I reading this right, or is this showing EHG admixture into LBK_EN? That doesn't seem right.

Bob Floy said...

@nirjhar
"When it comes to substratum, there cannot be consensus, because we are working with indirect evidence."

Uh huh. This is all crap and you know it. Everyone here, with the possible exception of Kristina, can see that you just want to argue against whatever is being said at this blog because you're upset about the death of your pet theory. Your motives are transparent. Have some dignity, you look ridiculous.

Philippe said...

Sorry, very off topic, but is it possible the haplogroup T people in chalcolithic Israel came from Europe? I'm thinking in particular about the Varna culture. Gold metallurgy appears in Israel around the same time these blue-eyed newcomers arrive from the north.

https://www.nature.com/articles/s41467-018-05649-9

Romulus said...

The R1a as Uralic argument never made any sense. However if Uralic arrived in Europe after the Bronze Age with N1c then I don't really see Comb Ware as Uralic or N1c heavy.

Huck Finn said...

N1c has apparently been found in Estonian Iron Age burials, in so called Tarand graves, but it may well be that it still first arrived into Baltics in the late Bronze Age. Those Estonian tarands, on the other hand, are possibly just one version of a common burial type of the period i.e. House of Death. Possibly sort of similar burials, but without a stone basement, were constructed by fex eastern Finnics in Karelia still in the early 20th century.

http://www.ortodoksi.net/images/f/fc/Krobu01.jpg

Slumbery said...

@Romulus

I am playing with the Global25 nMonte Runner that was referred by Davidski earlier. I just did a few tests with Comb Ware + various CWC + Siberian side sources reference populations. In the presence of CWC Baltic, all Uralic populations reject Comb Ware. But OK, CWC Baltic probably have some Comb Ware ancestry. So I replaced CWC Baltic with CWC Czech, they surely did not have Comb Ware ancestry. With that most of the European Uralic populations + Chuvash show some Comb Ware ancestry (Karelians, Finnish_East, Saami the most), with the notable exception of Udmurts. Not the Siberian ones however, neither Bashkirs.

It was just a short test and there are multiple possible interpretations. One is however, that the current area of surviving Uralic languages in Europe have a big cross section with former Comb Ware territory (also in the extreme North direct EHG and SHG ancestry is probably added to Comb Ware ancestry in this setup), but this ancestry does not reach out of that area much.

Slumbery said...

I stay with this topic a little bit yet. I tried to model the Sintashta outliers and Mezhovskaya with the Global25 nMonte runner. None of the Sintashta Outliers takes any Shamanka_N at all. However Mezhovskaya does, despite being more western overall.

Apparently this type of ancestry from the Baikal region (or similar to it) was not yet present in the Sintashta area at the time of the Sintashta outliers, but it was already here at the time of Mezhovskaya.

Another interesting thing that the Sintashta outliers are not just Sintashta + West Siberian. O2 and O1 takes Botai ancestry (O3 does not) and they all take Yamnaya/Poltavka/Afanasievo ancestry pretty well, up to the extreme O1, where Sintashta ancestry completely disappears in the presence of any of these three. This does not work with Poltavka Outlier however, it is completely rejected.


Huck Finn said...

@Slumbery and re "Another interesting thing that the Sintashta outliers are not just Sintashta + West Siberian. O2 and O1 takes Botai ancestry (O3 does not) and they all take Yamnaya/Poltavka/Afanasievo ancestry pretty well, up to the extreme O1, where Sintashta ancestry completely disappears in the presence of any of these three."

Very interesting, thank you. I'd personally guess that Pre Proto Uralic was spoken by people which were WSHG-biased, if not just WSHG, then Proto Uralic by a group diluted by the other groups you mentioned.

gL said...

@Slumbery
Nganassan are Uralics - they are not Finnish, but Samoyedic, which is subgroup of Uralics, but still - Uralics.

@Ebizur
It seems, that your comment does not distinguish clearly Ket from Uralics. Ket belongs to Yeniseian languages and they in turn seems to be closely related to Na-Dene languages, that are located in America(US/Canada). A wild guess would be that Yeniseian were spoken in most of Siberia before others arrived there, as 200 years ago Siberian linguistical landscape was a lot different than now.

@Kristiina
Point of origin for Pit comb ware(they left pottery that dates earlier, than any pottery found in Near East, so that makes them very ancient in this sense) culture can be pointed to Liao peninsula civilization, where earliest ceramics have been found along with N y-dna samples. It is highly unlikely, that earliest N y-dna in Siberia spoke language, that was not similar to Uralic(even if modern Uralic is diluted), as Yukaghir languages are candidates for Uralic-Yukhagir language group. It is hard to tell what kind of language they spoke in Liao, but when Yukhagir(most of those descendants now speak Yakut) split from others near Baikal, others spoke proto-Uralics. Khanty and Mansi have N1a2b, where Finn people have predominantly N1a1(it is new name for N1c, where equivalent in R y-dna would be switching names for R1a and R1b), which points to Khanty and Mansy as later split.

Speaking of languages, I find Carlos Quiles page very graphical and very informative in sense of historical maps - there are not many pages, that provide better timeline for languages/nations. However, along with Ket/Na-Dene, Yukhagir/Uralic ties, there also exist idea, that Basque might be Indo-European language or at very least isolate of IE - not as the whole extinct language group and I don't see this idea shown there. So Basque and Yukhagir existence alone should not make possible existence of PIE-Uralic from which respective IE and Uralic sprouted - actually, that is not shown in Carlos page. Anyway, all the logic would require to search for PIE and Na-Dene/Yeniseian connections first and Uralic connection to other NE/SE Asian languages. Clearly having been for thousands of years in each others neighborhood will influence both language groups, but that won't make them closer on language tree.


@Davidski
It would be interesting to compare R1a of Uralics. If it has been done before, maybe it is worth to do again, as repetition is the "mother of learning, the father of action, which makes it the architect of accomplishment". I've read some time ago, that Mari had R1a that was linked to R1a of modern Baltic people. It would be really nice to have some reference to any R1a data in Uralics. It seems, that it takes about the same effort imagining why R1a appeared in Uralics during migration and assimilation, as claiming and repeating that PIE were originally Uralic speaking, so it would really better to have some point of reference to data.

Davidski said...

@gL

There's nothing remarkable about Uralic R1a. By and large it's a young subset of Slavic and Baltic R1a-Z645.

So it's pointless to argue that the R1a-Z645-rich Corded Ware population was Uralic-speaking, and anyone who does so is a crackpot.

Slumbery said...

@gL

Why are you telling me this? I never disputed that Nganasan are an Uralic speaking group.

Kristiina said...

@gL

You cannot use Ket yDNA as a prove of a long presence of Yeniseian languages in Siberia because the Ket Q haplotype is very recent. In the recent paper "Dispersals of the Siberian Y-chromosome haplogroup Q in Eurasia", 2017 (https://link.springer.com/article/10.1007/s00438-017-1363-8)
the OMRS expansion time of Q1a2a1-L54 in the Yenisei basin is 3.3 kya while the OMRS expansion time of Q1a1a1-M120 in East Asia is 6.4.

From the paper: "However, Yeniseian and Samoyedic samples in this study belonged to L54, which was different from the results of previous studies (xL54). In view of the time estimates (Table 1), we postulated that Q1a2a1-L54 had migrated from the southern Altai region and was assimilated into Yeniseian- and Samoyedic-speaking populations during a recent historical period."

M. Myllylä said...

"Perhaps even more surprisingly, we found that Estonians, who show close affinities in IBD analysis to neighbouring Finnic speakers and Saami, do not share an excess of IBD segments with the VUR or Siberian Uralic speakers. This is even more striking considering that the immediate neighbours—Finns, Vepsians and Karelians—do."

This happens because Finestructure, and other recent tools estimating common haplotypes, is not capable to make difference between shared alleles and shared haplotypes. In other word, haplotype analyses are not yet powerful enough to make accurate datings. The Finns, Karelians and Vepsas, they all share haplotypes with Saamis, while Saamis share common alleles with Siberians and with Eastern Uralic speakers. The Estonians don't share haplotypes with Saamis, but do it with Finns. So this raises the dilemma of where the Estonians share alleles with Saamis and Finns, but haplotypes only with Finns. So, the answer is not the reality but the poor analysis.