A claim often made in popular culture is that the Saami people of Fennoscandia and Northern Russia are the last indigenous Europeans. I saw some guy blurt this out on a random cooking show the other day. But it's been obvious for a while now, thanks to analyses of modern-day DNA, that the Saami, and indeed almost all other Uralic-speaking groups in Europe, have a somewhat more complex population history than the majority of non-Uralic-speaking Europeans.
Now, ancient DNA is helping to cement these findings. The quotes and figure below are from a new preprint at
bioRxiv by Lamnidis et al. [
LINK] focusing on the spread of Siberian ancestry across Northeastern Europe from the late stone age onwards. It's a phenomenon that had the biggest impact on the Uralic-speaking populations of Fennoscandia, and is, in all likelihood, related in a profound, albeit complex, way to the ethnogenesis and expansion of the proto-Uralic people. Emphasis is mine:
The six ancient individuals from Bolshoy show substantially higher proportions of the Siberian component, which comprises about half of their ancestry (49.4-65.3 %), whereas the older Mesolithic individuals from Motala do not share this Siberian ancestry. The Siberian ancestry seen in EHG probably corresponds to a previously reported affinity towards Ancient North Eurasians (ANE) [2,24] , which also comprises part of the ancestry of Nganasans. Interestingly, results from uniparentally-inherited markers (mtDNA and Y chromosome) as well as certain phenotypic SNPs also show Siberian signals in Bolshoy: mtDNA haplogroups Z1, C4 and D4, common in modern Siberia 18,25,26 , in individuals BOO002, BOO004 and BOO006, respectively (confirming previous findings [18] ), as well as Y-chromosomal haplotype N1c1a1a (N-L392) in individuals BOO002 and BOO004. Haplogroup N1c, to which this haplotype belongs, is the major Y chromosomal lineage in modern North-East Europe and European Russia, especially in Uralic speakers, for example comprising as much as 54% of Eastern Finnish male lineages today [27]. Notably, this is the earliest known occurrence of Y-haplogroup N1c in Fennoscandia.
...
We formally tested for admixture in north-eastern Europe by calculating f3(Test;Siberian source, European source) using Uralic-speaking populations - Estonians, Saami, Finnish, Mordovians and Hungarians - and Russians as Test populations. Significantly negative f 3 values correspond to the Test population being admixed between populations related to the two source populations [34]. Additionally, the magnitude of the statistic is directly related to the ancestry composition of the tested source populations and how closely those ancestries are related to the actual source populations. We used multiple European and Siberian sources, to capture differences in ancestral composition among proxy populations. As proxies for the Siberian source we used Bolshoy, Mansi and Nganasan, and for the European source modern Icelandic, Norwegian, Lithuanian and French. Our results show that all of the test populations are indeed admixed, with the most negative values arising when Nganasan are used as the Siberian source (Supplementary Table 3).
...
Consistent with f3-statistics above, all the ancient individuals and modern Finns, Saami, Mordovians and Russians show excess allele sharing with Nganasan when used as Test populations. Of all Uralic speakers in Europe, Hungarians are the only population that shows no evidence of excess allele sharing with Nganasan, consistent with their distinct population history as evidenced by historical sources (see ref 35 and references therein).
...
While the Siberian genetic component described here was previously described in modern-day populations from the region [1,3,9,10], we gain further insights into its temporal depth. Our data suggest that this fourth genetic component found in modern-day north-eastern Europeans arrived in the area around 4,000 years ago at the latest, as illustrated by ALDER dating using the ancient genome-wide data from Bolshoy Oleni Ostrov. The upper bound for the introduction of this component is harder to estimate. The component is absent in the Karelian hunter-gatherers (EHG) [3] dated to 8,300-7,200 yBP as well as Mesolithic and Neolithic populations from the Baltics from 8,300 yBP and 7,100-5,000 yBP respectively [8]. While this suggests an upper bound of 5,000 yBP for the arrival of Siberian ancestry, we cannot exclude the possibility of its presence even earlier, yet restricted to more northern regions, as suggested by its absence in populations in the Baltic during the Bronze Age.
...
The large Siberian component in the Bolshoy individuals from the Kola Peninsula provides the earliest direct genetic evidence for an eastern migration into this region. Such contact is well documented in archaeology, with the introduction of asbestos-mixed Lovozero ceramics during the second millenium BC [47], and the spread of even-based arrowheads in Lapland from 1,900 BCE [48,49]. Additionally, the nearest counterparts of Vardøy ceramics, appearing in the area around 1,600-1,300 BCE, can be found on the Taymyr peninsula, much further to the east [48,49]. Finally, the Imiyakhtakhskaya culture from Yakutia spread to the Kola Peninsula during the same period [18,50]. Contacts between Siberia and Europe are also recognised in linguistics. The fact that the Siberian genetic component is consistently shared among Uralic-speaking populations, with the exceptions of Hungarians and the non-Uralic speaking Russians, would make it tempting to equate this component with the spread of Uralic languages in the area. However, such a model may be overly simplistic. First, the presence of the Siberian component on the Kola Peninsula at ca. 4000 yBP predates most linguistic estimates of the spread of Uralic languages to the area [51]. Second, as shown in our analyses, the admixture patterns found in historic and modern Uralic speakers are complex and in fact inconsistent with a single admixture event. Therefore, even if the Siberian genetic component partly spread alongside Uralic languages, it likely presented only an addition to populations carrying this component from earlier.
This generally looks like a very solid preprint, so I don't expect any major changes between now and formal publication. I have to be honest though, the qpAdm analysis looks like crap. Also, the authors are using the Russian sample set from the Human Origins dataset, which comes from the Kargopol district in Northern Russia. This was actually an Uralic-speaking region until not long ago. No wonder then, that they're inferring that Russians are very similar to Uralic-speaking populations.
But I know from my own analyses that there's quite a bit of genetic substructure within European Russia. For instance, Russians from southwest of Moscow are much less Uralic-like than the Kargopol Russians, and indeed very difficult to distinguish from other East Slavs, and even West Slavs. Hence, it might be useful to sample and run a couple more regional ethnic Russian groups for comparison. This might help to strengthen the argument that Siberian ancestry is somehow intimately intertwined with the expansion of Uralic languages in Europe.
Citation...
Lamnidis et al.,
Ancient Fennoscandian genomes reveal origin and spread of Siberian ancestry in Europe,
bioRxiv, Posted March 22, 2018, doi: https://doi.org/10.1101/285437
See also...
The Uralic cline in the Global25
210 comments:
1 – 200 of 210 Newer› Newest»*Background. Ryu claims Bolshoy genomes with N1c in this paper aren't representative of the people who spread Uralic languages into Fennoscandia. Instead, he claims this populations was a Siberian/Andronovo-like mix similar to Mezhovskaya and arrived after 1500 bc.
@Ryu,
I ran the models you did for Saami and Finnish at Anthrogencia.
https://anthrogenica.com/showthread.php?13821-Siberian-aDNA-and-Turkic-Iranic-and-Uralic-populations
I'm getting different results. I use Global25's scaled spreadsheet. It's much better than the normal Global25 spreadsheet. Here are the results I get for Saami, Finnish, Karelian, Ingrain.....
3.3339"
Saami
Nordic_IA,42.3
Nganassan,20
EHG,15.2
Scythian_AldyBel,11.4
Narva_Lithuania,7.9
Mezhovskaya,2.2
4.0285"
Ingrian
Nordic_IA,66.8
Scythian_AldyBel,10.1
Narva_Lithuania,6.8
EHG,5.7
Mezhovskaya,5.2
Globular_Amphora,2
Karasuk,1.7
Nganassan,1.6
3.0713"
Finnish
Nordic_IA,76.2
Scythian_AldyBel,6.9
EHG,5.2
Narva_Lithuania,5
Nganassan,3.1
Mezhovskaya,2.4
4.1929"
Karelian
Nordic_IA,64
Scythian_AldyBel,12.2
Narva_Lithuania,9.8
EHG,6.3
Nganassan,3.8
Sarmatian_Pokrovka,3.1
Srubnaya_outlier,0.6
Globular_Amphora,0.2
It is a pity that all Leväluhta samples are females. However, their mtDNA is not Finnish nor Saami. Three haplogroups are K1a4a1b, U5a1a1 and U5a1a1a'b. Therefore, the presumption that they spoke a Saami language relies on the known existence of Saami toponyms all over Finland and their genetic affinity with modern Saamis, although, on the PC map, Leväluhta is closest to Chuvash.
As for the roots of the Saami, I find it very interesting that the well-documented Saami burial from Chalmny Varre in Murmansk which dated only to the end of 18th century revealed yDNA I2a1 and mtDNA V7a1 in a Saami male. The other Chalmny Varre mtDNA, U5b1b1a3, was also a typical Saami mtDNA.
Moreover, one of the mtDNAs in Bolshoy was U5a1d which was found in Mesolithic Scandinavians. Therefore, I see that in Saamis there is gene-flow from the Mesolithic inhabitants of the area and the known linguistic substrate in Saami languages could be related to I2a1.
@ Sam
I am using the scaled sheet as well. You will see in the closest populations list that the closest population to Saami is Mezhovskaya.
EHG is not a proper contributor. There are Combed Ware, Narva, and so on available, which represent the mesolithic inhabitants of the region at the time.
[1] "1. CLOSEST SINGLE ITEM DISTANCES"
Mezhovskaya Srubnaya Scythian_AldyBel
0.03980681 0.04278583 0.04393002
Nordic_IA Karasuk Scythian_Samara
0.04669220 0.04986604 0.05072782
Scythian_Pazyryk Sarmatian_Pokrovka
0.05346288 0.05416290
Here is a fit that is noticeably better than yours, and makes much more sense:
[1] "distance%=1.8387 / distance=0.018387"
Saami
Mezhovskaya 25.45
Baltic_BA 20.30
Nganassan 20.00
Nordic_IA 12.40
Comb_Ceramic 11.40
Scythian_AldyBel 7.75
Srubnaya_outlier 2.70
Like I said, if you would like to read up on the ethnogenesis of the Uralics, please do and stop spreading misinformation. The authors have highlighted that the values for admixture between Nganasan and local HGs in the Bolshoi people are far too old, and took place between the wrong populations (EHG and Nganasan); all Uralic populations have admixture between modern European populations and Nganasan and the dating is within the metal ages.
@Ryu,
"The obvious change between Bolshoi people and Iron Age Saami is an introduction of ancestry similar to modern Europeans, with CHG and EEF. If Mansi-like populations extended from Central Siberia to Fennoscandia, then the change from then to now is a Europeanisation of the gene pool throughout European Siberia, including among Uralics."
Iron age Finland genomes pull away from Bolshoi towards northern Europeans. Specifically they pull towards JK2065 who is one of the Iron age Finland genomes but has no Siberian ancestry and clusters with Lithuanians.
The shift from Bolshoi to Iron age Finland/modern Saami was certainly not caused by an Andronovo-Siberian mixed Uralic population. It was caused by Indo European populations who moved up into Fennoscandia from somewhere further south.
@Ryu,
I see what the problem is. When I copied your results I used the "SET 5: Samoyed-like Uralics" references not the "SET 7: Finnics and Saami" references.
Your SET 7 references for Finnics and Saami doesn't include any Siberians. That's absurd. No wonder in your results Saami and Finns score crazy Sycthian numbers. Without Siberian references they'll score high in anyone with significant Siberian ancestry.
@ Sam
Lol, this massive expansion of Indo-European ancestry throughout the Volga and Fennoscandia, such that Saami became 70% Indo-European and Estonian became almost 90% Indo-European, which even reached Mansi in Central Siberia (~10% Indo European) led to no language shifts throughout all of Siberia, rather the stone-age primitive hunter-gatherer Uralics retained their language even after IE pastoralists from advanced metal age Sintastha and Andronovo replaced up to 90% of the local ancestry. Right.
Fact of the matter is, early Fennoscandians had high levels of Steppic ancestry, with Mezhovskaya probably a close but not exact representative of them, and assimilated Nganasan-like Bolshoi Oleni Ostrov substrates as they moved into the Taiga and Tundra. This is the reason why all Uralic populations have EEF and CHG ancestry and form a cline stretching from Selkup to modern Europeans and clearly do not form a cline from Selkup to WHG, SHG or EHG. Go check the PCA of this paper, which could not show this any more clearly than it already does. This is also the reason why all of them had metalworking traditions or used to have such traditions, herded reindeer, were full-blown pastoralists (which was the case for some historical Finno-Ugric populations, some were outright Steppe people, like the proto-Magyars who brought the Hungarian languages) or farmed, except for Nganasan who were the only pure hunter-gatherers who were recently Uralicised Yukagir. This is also the reason why they have huge numbers of proto-IEs and proto-Indo-Iranian loanwords at the proto-language level. This is also the reason why they have star-shaped N Y chromosome phylogenies, which are a feature of agriculturalists or pastoralists. Such phylogenetic patterns never appear among Hunter-gatherers, you seen the Y HG phylogenies of Mesolithic Europeans?
Anyone who thinks that Uralic is a "Mesolithic Stone Age language family" is deeply misguided.
False, you see the Nganasan in there? This is the second set of analyses, which are better than the first round. Stop replying emotionalistically and actually consider the facts.
Ryu, here's what I get using the exact same pops.
2.7657"
Saami
Nganassan,21.9
Baltic_BA,19.8
Nordic_IA,18.5
Comb_Ceramic,17.6
Mezhovskaya,17.3
Scythian_AldyBel,4.7
Srubnaya_outlier,0.2
I must repeat that we still lack all ancient references from forest Volga and West Siberia as well as from Fatyanovo and Abashevo which should all be relevant for Uralics. Moreover, Nganasans are a modern population and a Neolithic sample with similar ancestry would be much better. Therefore, I doubt that these calculations are conclusive.
Furthermore, we should also much better understand the expansion of I1 in Fennoscandia. We still have only a couple of BA I1 samples from Sweden and one from Neolithic Hungary as reference, excluding the late Iron Age and Medieval samples.
@Samuel Andrews and re:"The shift from Bolshoi to Iron age Finland/modern Saami was certainly not caused by an Andronovo-Siberian mixed Uralic population. It was caused by Indo European populations who moved up into Fennoscandia from somewhere further south."
The main immigration movement into Finland after BOO was Uralic speaking, from South/South-East, speaking both Proto Saami and Proto Finnic. Saami language, as a matter fact, expanded into Lappland as late as in the Iron Age.
@Ryu,
Explain to me why there is a Lithuanian-like person in iron age Finland and that modern Saami fit halfway between this person and Bolshoy?
@Ryu,
Using the references you just used to model Saami. Finns, karelians are basic northern Europeans with Siberian admixture. There's no excess Steppe in them.
1.6935"
Karelian
Baltic_BA,47.1
Nordic_IA,31.6
Mezhovskaya,13.9
Nganassan,6.8
Scythian_AldyBel,0.6
Srubnaya_outlier,0
Comb_Ceramic,0
1.6394"
Finnish
Nordic_IA,55.1
Baltic_BA,30.6
Mezhovskaya,9.3
Nganassan,5
Scythian_AldyBel,0
Srubnaya_outlier,0
Comb_Ceramic,0
"the presence of the Siberian component on the Kola Peninsula at ca. 4000 yBP predates most linguistic estimates of the spread of Uralic languages to the area."
Perhaps the Siberian component (i.e. the component maximized in the Nganasans) arrived with the Comb Ceramic culture and the Proto-Uralic language arrived with the Seima-Turbino phenomenon.
@ Sam
Are you blind or what? You see the Mezhovskaya % + Nganasan % consistently appearing in these? After you take out the parts where they are the same as other IE Europeans in the region, who are mostly Baltic_IA+Nordic_IA, what is left behind?
And do they have excess Steppe compared to Bolshoi or not? 70% excess steppe in cultures associated with Iron Age Saami doesn't seem like a lot to you?
I dunno what you did with the Saami fit, but unlike your fit mine actually coheres with the finding that the closest population in PCA space to Saami is Mezhovskaya. Simple geometry.
@ Kristiina
Interestingly, Parpola has attempted to build a total theory associating archaeological cultures with Uralic and its sub-branches, and he places Mezhovska as a descendant of Cherkask'ul, which in turn derives from Fedorovo and in turn from Fatyanovo-Abashevo, so Mezhovska may give us a glimpse of something close. The uniparentals are an issue for sure, but something like it may approach the thing we want.
“Anyone who thinks that Uralic is a "Mesolithic Stone Age language family" is deeply misguided.“
Generally, have to agree with RK
All (almost) the great languages families are metal age, or at least Neolithic
@ Sam
Your SET 7 references for Finnics and Saami doesn't include any Siberians. That's absurd. No wonder in your results Saami and Finns score crazy Sycthian numbers. Without Siberian references they'll score high in anyone with significant Siberian ancestry.
If you look at the PCA of this paper (or of any PCA centered on Siberians and North Eurasians) you will see that the Uralics form a straight cline between Lithuanians and Yukagir/Nganasan (not EHG, WHG, SHG and Nganasan!) except for one set of populations, that is populations from the Volga region, which are perceptibly shifted South (here represented by Chuvash, which are identical to Volga Uralics). The reason why this is is because Scythians were known to have played very important roles in later cultural (and population) change in Uralics in the Volga area, among people like Komi, Mari, Udmurt, etc, in fact under their influence some Uralic groups became capable of forming Iranic and Turkic-style steppic confederations, like Magyars. Such that all Uralics other than Saami and Finns share rare alleles with Mongols and South Siberians, as uncovered by M Myllyla's analysis.
Considering this and the fact that the earliest stratum seems to be Saami like (Iron Age Saami), the following type of fit makes perfect sense:
[1] "distance%=1.5897 / distance=0.015897"
Komi
Saami 65.45
Scythian_AldyBel 23.55
Sarmatian_Pokrovka 7.65
Nordic_IA 3.35
All other populations at 0%, no need for any further Nganassan or Siberian even after it is included, in keeping with the fact that East Siberian populations were not magically airlifted and dropped there after the Iron Age.
You need to be historically aware of what was going on, you can't just throw things at the target and see what sticks.
@Sam "Explain to me why there is a Lithuanian-like person in iron age Finland"
By the way, that Lithuanian like JK2065 individual carries the Neolithic mtDNA K1a4a1b which today seems to be found somewhere in Scandinavia and the British Isles.
However, one could also say that this "Lithuanian"-like hunter individual who was buried in a swamp in the middle of nowhere between 400-800 AD is all what was left of Corded Ware in Finland. :)
I am sure that the bulk of the Finnish DNA is not from the remnants of Corded Ware in Finland or from Bolsoy people in Kola peninsula.
Kristinia,
Finns are mostly Lithuanian like. So.....
Or Lithuanians are mostly Finnish like. So ...
Lithuanians lack Siberian admixture. Why are you against the idea Finns are normal Europeans?
Using D-stats instead. Using the same model with Global 25 scaled Finns and Saami do score 20-25% Andronovo. Using Global 25 Estonians socre the exact same way as they do here.
2.6942"
Saami
Lithuanian,55.8
Nganasan,24
Motala_HG,10.3
Eastern_HG,9.9
Andronovo,0
0.3307"
Finnish
Lithuanian,86.8
Nganasan,7.5
Andronovo,3.6
Eastern_HG,2.1
Motala_HG,0
0.3283"
Estonian
Lithuanian,87.8
Andronovo,10.2
Nganasan,2
Eastern_HG,0
@Kristiina
I am sure that the bulk of the Finnish DNA is not from the remnants of Corded Ware in Finland or from Bolsoy people in Kola peninsula.
I totally agree with this statement. Compare the pattern of closest populations for Lithuanian, Belarusian, and other IE populations vs Saami.
LITHUANIAN
[1] "1. CLOSEST SINGLE ITEM DISTANCES"
Baltic_BA Nordic_IA Srubnaya
0.02271316 0.04718692 0.05480191
Scythian_AldyBel Mezhovskaya Sarmatian_Pokrovka
0.05912284 0.06672548 0.06700302
Scythian_Samara Comb_Ceramic
0.06965996 0.07135279
BELARUSIAN
[1] "1. CLOSEST SINGLE ITEM DISTANCES"
Baltic_BA Nordic_IA Srubnaya
0.03105950 0.04235659 0.04884099
Scythian_AldyBel Sarmatian_Pokrovka Mezhovskaya
0.05131226 0.06026641 0.06112499
Scythian_Samara Scythian_Pazyryk
0.06388016 0.06432686
SAAMI
[1] "1. CLOSEST SINGLE ITEM DISTANCES"
Mezhovskaya Srubnaya Scythian_AldyBel
0.03980681 0.04278583 0.04393002
Nordic_IA Scythian_Samara Scythian_Pazyryk
0.04669220 0.05072782 0.05346288
Sarmatian_Pokrovka Karasuk_outlier
0.05416290 0.05534864
^^(no Baltic_BA)
Fits for IE populations using the same set of populations as Saami and Finnish fits seen above have no Mezhovskaya and no Nganasan, what little East Asian ancestry they have very Southern-shifted:
[1] "distance%=1.2136 / distance=0.012136"
Lithuanian
Baltic_BA 70.15
Nordic_IA 25.45
Scythian_Pazyryk 3.15
Altai_IA 1.25
[1] "distance%=1.5008 / distance=0.015008"
Belarusian
Baltic_BA 58.80
Nordic_IA 26.50
Altai_IA 9.20
Scythian_AldyBel 3.00
Scythian_ZevakinoChilikta 1.95
@Ryu,
Using, Global 25 Saami & Finns can be successfully modeled with excess Andronovo stuff. They also can be modeled successfully as Northern European, EHG, Siberian.
It's really up to opinion which model is right. D_stats though seems to say that there's no excess Andronovo stuff.
I'm open to the idea Uralic(s) had huge amounts of Andronovo ancestry. But I do find it unlikely. You argued star-like Y DNA explosion can't happen in hunter gatherers. 3 of 3 Bolshoy belonged to the same y DNA type the vast majority of modern Finns and Saami belong to. You might be suprised. So, you might be wrong.
We know from historical sources that Finnish speaking Finns were farmers who cleared land for farming and spread towards the north and northeast very late. This farming population came from the south. You cannot show that the Lithuanian Corded Ware people were farmers who cleared land for farming in Southern Finland and later on their fields were taken over by Bolshoi hunters from Kola Peninsula.
@Ryu,
Closest population matches means jack squat. My closest match is French yet I have no French ancestry. All we know is, Saami and Mezhovskaya are a somewhat similar European/Siberian mix.
@ Sam
If you want high resolution for recent drift (i.e. you wanna split differences like Germanic vs Celtic etc) Global25 is what is needed, and there the PCA distances make the pattern very clear.
Three genomes from the sites close by doesn't tell us anything, when we have many then we know, e.g. Iron Gates. Star shaped expansions that are super clear like those among Uralics or even among aDNA from Bell Beakers are only known from pastoralist and agriculturalist contexts, whether modern or ancient, and there are good anthropological reasons why this should be the case. Either way Uralic star-shaped expansions are not all that old either.
Wait a minute. L389 belongs to this quite young starcluster. 4700 year old TMRCA according Yfull. https://www.yfull.com/tree/N-L1026/
This is the most biggest starcluster in haplogroup N. It is like M269 for R1b or M417 for R1a.
If this people were not Uralic speakers then what language they were speaking.? Those folks had a dramatic expansion all over North Eurasia. Reaching even Kamchatka and Chukotka. That is quite puzzling that they are not Uralic speakers. I am really surprised.
@Kristiina,
"You cannot show that the Lithuanian Corded Ware people were farmers who cleared land for farming in Southern Finland and later on their fields were taken over by Bolshoi hunters from Kola Peninsula."
Clearly, Bolshoi marks the introduction of N1c & Siberian ancestry into Fennoscandia. I think we can safely assume that Bolshoi hunters imposed their Y DNA on those farmers. Did they impose their language? It isn't much of a stretch to say yes to that too.
Oops I made a typo. I should write L392 instead of L389.
Closest population matches means jack squat. My closest match is French yet I have no French ancestry. All we know is, Saami and Mezhovskaya are a somewhat similar European/Siberian mix.
Dude I don't give a damn whether they mean anything for you, you are an American mixed-ethnicity individual, of course it means nothing.
This kind of thinking (Uralic component peaks in Nganasan, Nganasan must be first Uralics!) is just naive. One would expect you to know better. Just to let you know, I was of the same opinion and got into a heated discussion with Kristiina and Shaikorth, they showed me evidence (none of which supports this by the way) and then I switched. If you are not capable of doing the same thing, you have no right to be in the business of informing others. None at all.
@Ryu,
"If you want high resolution for recent drift (i.e. you wanna split differences like Germanic vs Celtic etc) Global25 is what is needed, and there the PCA distances make the pattern very clear."
I agree with that. Yet, when it comes to measuring ancient ancestry, like excess Steppe ancestry via Andronovo-like ancestors, I trust formal stats the most.
The circumstances surrounding Bolshoi is what makes me think they mark the introduction of Uralic in Fennoscandia. Really, their Y DNA is the most convincing thing. If the Y DNA experts confirm Bolshoi belonged to a very specific, widespread, young form of N1c then it's game over. Unless, you suppose the new wave of Siberian/Andronovo mixed people brought the exact same Y DNA type.
@Ryu,
"Dude I don't give a damn whether they mean anything for you, you are an American mixed-ethnicity individual, of course it means nothing."
It does mean something. Populations who carry a similar mixture of ancestry will cluster together yet share no direct common ancestry with each for. For example, northern Italy and Spain cluster fairlly close to each other yet have no recent common ancestry.
Saami and Mezhovskaya seem similar because they are a similar mix. That's not a big deal.
@ Sam
Do you understand the difference between recent drift vs ancient components or not? Why is it that Global25 can distinguish Celtic and Germanic, and D-stats datasheets cannot?
@Ryu,
Test=zombie I created which is an identical mixture of Andornovo, EHG, Nganassan as Mezhovskaya.
Not a huge difference. Technically, Saami is closer to this zombie than to Mezhovskaya. So much for shared drift between Saami and Mezhovskaya.
Mezhovskaya
"1. CLOSEST SINGLE ITEM DISTANCE%"
test Mezhovskaya Scythian_AldyBel Finnish_East Scythian_Samara Scythian_ZevakinoChilikta
9.064838 9.594343 9.628804 11.180561 12.372770 20.300823
Scythian_Pazyryk
22.607802
If the distance is 9%, that worse than even this:
"distance%=8.2995 / distance=0.082995"
Natufian:I1072
Barcin_N:I0745 38
Tanzania_Luxmanda_3000BP:I3726 36
Iberomaurusian 26
Clearly wrong. Recent drift matters.
Ryu,
Mezhovskaya is slightly less to Saami than the Mezhovskaya zombie. If Mezhovskaya shared drift with Saami you'd expect Mezhovskaya to be closer.
"If the distance is 9%, that worse than even this:"
Yes and Mezhovskaya is that distant from Saami.
SAAMI
[1] "1. CLOSEST SINGLE ITEM DISTANCES"
Mezhovskaya Srubnaya Scythian_AldyBel
0.03980681 0.04278583 0.04393002
@Ryu,
Explain how Bolshoi and modern Finns+Saami share N-L1026 whose TMRCA is 4700bp yet Uralic languages come from a different population.
From an objective standpoint, the conclusion should be Bolshoi people are the source of N-L1026 in modern Fennoscandia. Therefore, these hunter gatherers caused a Y DNA turnover. If so I don't see why they couldn't have also caused a language turnover as well.
If all that is true. I say we go back to the model of Saami being intermediate between Lithuanian and Bolshoi.
"SAAMI
[1] "1. CLOSEST SINGLE ITEM DISTANCES"
Mezhovskaya Srubnaya Scythian_AldyBel
0.03980681 0.04278583 0.04393002"
Saami
"1. CLOSEST SINGLE ITEM DISTANCE%"
test Mezhovskaya
9.064838 9.594343
I don't know why there's a difference here. I get a distance of 9.5 you get a distance of 3.9.
The N clades of all Uralic populations expanded at much later dates than 4700BC. E.g. Finnic N is the result of a founder effect that is only 2000 years old, which is in turn different from Saami N, which is in turn different from Volgaic N.
Bolshoi is at best Para-Uralic, like EHG might be para-IE. Proto-Uralic was a single community, responsible for a single autosomal signature.
At some point, other situations like this took place, such as with the Tungusics, most of whom are hunter-gatherers even today, but some of which amalgated with steppe peoples (mongolics) to expand, resulting in a late N-cluster specific to Nurhaci's clan.
^^BP, not BC.
In a similar scenario, Ukraine_Mesolithic has R1a-M417, but the IE expansion has a very specific autosomal signature localising to the Yamnaya and earliest Corded Ware, even though R1a-M417 may be flowing around para-IEs in the Pontic region for some time. No difference here.
@Sam "Clearly, Bolshoi marks the introduction of N1c & Siberian ancestry into Fennoscandia."
We currently have N1c from Kola peninsula and Smolensk area (Zhizhitskaya culture), and Smolensk N1c is 1000 years older. Therefore, my understanding is that N1c approximately spanned the triangle between West Siberia, Kola Peninsula/Karelia and northwesternmost part of Russia.
This area has overlap with Comb Ceramic, therefore the old idea of the Comb Ceramic as the culture introducing N1c to Europe may ultimately be true. We will see.
However, languages are a different matter, and I do not argue that a possible Comb Ceramic N1c spoke Proto-Uralic.
One final point: the Nganasan population here admixed with an EHG population approx 5000 years ago, but it imposed the Y-DNA on the wrong population (EHG) in the wrong context (among Hunter-gatherers) and at the wrong time. The Uralic populations here under ALDER have Nganasan admixing with a different population (Lithuanian and Icelandic-like) in the right context (presumably among already agropastoral peoples, which is what the Lithuanian and Icelandic-likes represent) at a later time, probably analogously to how Turkics and Tungusics took up pastoralism from already pastoral peoples in the metal ages and historical period respectively (with the requisite admixture), leading specific haplogroups and autosomal signatures to expand. So Uralic ethnogenesis as indicated by autosomal gene pools was clearly the result of a separate process that could have introduced multiple clades of N1c onto an autosomal and cultural background suited for expansion that was different and distinct from the event that brought Bolshoi. The Nganasan-EHG admixture did not even produce modern-Uralic-looking autosomes.
Saami is special in that it is dominated by 2 specific mtDNA. This is unusal and should tell us something.
One thing to note about the N found in this paper, specifically N-L392/L1026, is that its subclades are widely distributed across a large number of groups, e.g. Chukchi N in Beringia is closer to Baltic M2783 under L392 than Balts are to Estonian N-Z1933. Many Uralics also do not have N in this subclade, e.g. in Maris the majority is Y9022 (separated >7000 bp), in Udmurts all N is Y9022, same for Veps among Finnics.
This kind of pattern resembles R1a-M417 and R1b in IEs, a few different clades, often quite diverged, expanding in different subgroups at different times. This can testify to the fact that autosomals can show a single pulse, where uniparentals show clades that ride the wave.
H/T Shaikorth.
RK,
just to clarify, Udmurts do have N1b too. Ilumäe study just showed their N1c is all Y9022. Vepsians have high N1b but no Y9022.
The region had low population density to begin with, so it shouldn't be all that surprising to see a young star cluster in the most remote corner of Europe. Always surprised to see such a low rate of R1a in Finland, pretty comparable to the R1b level % wise. I guess CWC didn't leave much of a trace. I'd be curious to know the age of the I1-Finnish clusters, and if they predate Swedish expansions. No doubt the R1b arrived later with either Germanic people or later medieval immigrants.
Sam,
Lithuanians, clear to French and Brits appear to have Siberian admixture. Lithuanians even have a lot of N1 plus Siberian mtDNA. You're getting schooled as usual. Sit and learn rather than making comments like Natufians are Iberomaurusian plus Anatolia bullshit. You neither understand history or statistics.
Just to be clear Parpola's linguistic and archeological model fits very well with modern distribution of N1c and TMRCA dates. aDNA seemed to support this too with Sargat being N1c however BOO definatly makes thing more confusing.
It seems to me that either BOO's L1026 is a result of some kind of pioneering Para-Uralic group (Seima-Turbino, Netted Ware, Comb Ceramic) mixing with some Siberian new comers or European N1c does have its roots in some sort of Circum Polar Siberian like group, in which case I highly doubt they spoke Uralic.
Long distance contacts between Northern Fennoscandia and places such as Taimyr actually did take place even after BOO:
"Metal may have been obtained even from beyond the Urals. Christian Carpelan has pointed out the problem of the so-called imitation
textile-impressed ceramics (waffle ceramics)which occur in a zone extending from the Kola Peninsula to Norrland in Sweden. This ceramic
group is connected in any way to textile-impressed ceramics as such. Its points of origin are in the region of the Lena River in Northern Siberia and Carpelan has suggested that the upswing of trade caused by the use of bronze created contacts of at least short duration between these distant areas in the period from ca 900 to 500 BC
(Carpelan 1970, 1975b, 1979). The route of contact may have passed north of the White Sea and via the Kola Peninsula."
http://www.sarks.fi/fa/PDF/FA3_51.pdf
BOO, on the other hand, is supposed to be connected to Ymyakhtakh Bronze culture, which may be related to Chukchi and Koryak.
https://en.wikipedia.org/wiki/Ymyyakhtakh_culture
Chukchi and Koryak, interesting enough, share the same N-sub lineage than fex Balts (and many Buryats and Mongols). Ymyakhtakh is apparently based on immigration wave from lake Baikal area. Complicated as it may sound, I suggest that paternal N in BOO is related to Proto Chukchi and Koryaks of Ymyakhtakh, coming from lake Baikal area, first to Siberian North, then to Northern Fennoscandia because of trading. However, I also suggest that the lineage was very much Uralic before reaching the melting pot in Baikal area Siberia and especially Yakutia after that.
@Huck Finn
You suggest that the Kola N line belongs to Ymyakhtakh culture(c. 2200–1300 BC) and is from Lake Baikal and before reaching Baikal it was Uralic. In your opinion, to what culture this Uralic speaking N community belongs before reaching Baikal and where it comes from?
@Chad,
Of course, I know Lithuanians have Siberian admixture. But it is tiny so I didn't feel a reason to mention that or force it in when modeling Lithuanians.
Samuel Adnrews:"Yes and Mezhovskaya is that distant from Saami"
Mezhovskaya post-dates the branching-out of Uralics and very likely specifically ancestral to the Ugric branch. This means the Finnish branch probably does not have direct Mezhovskaya ancestry. Instead proto-Finns have some common ancestry with Mezhovskaya.
BTW, the already sampled ancient Siberian populations that can be suspected to be ancient Uralics are all _less_ Siberian that modern Mansi, let alone the Nganasan.
@Kristiina: The paternal N of Ymyakhtakh, coming from Baikal area, was apparently based on Bronze Age movement to Siberia from the Ural area. The original Uralic speaking herder group was assimilated into different ethnic entities, including Proto Chukchi-Koryak Ymyakhtakh but also groups leading to modern Buryats and Mongols.
When those N's then reached Kola much later, they were Siberian to the bone.
I don't know much about Uralic prehistory, but there is linguistic evidence for Uralic influence being the cause of satemization.
@Ryu,
'N-L32 doesn't have a Star-like phylogeny, star-like phylogeny, ha!!!'
Looks like N-L392 has what I call a Separta-phylogeny. A Separta phylogeny has a handful of super popular subclades with star-like phylogenies that peak in different regions. For example, R1b P312 has a separta phylogeny. R1b L21 in british Isles, R1b DF27 in Iberia, R1b U152 in Italy.
Separta phylogeny marks the point right before the mHG/yHG bursted. I usually allude to this meaning the ancestral type (in this case N-L392) expanded into many regions then in each region experienced new star-like expansions. Or, all the star-like subclades were born in one tiny region then suddenly bursted into many regions.
N-L392 is the common ancestor between all the star-like. It has a separta phylogeny. IMO, this suggests Bolshoy represents our best reference for the source population of L392. Not just that but that they represent the expansion of L392 into Fennoscandia which then lead to more recent star-like expansions.
@Shaikorth,
"ust to clarify, Udmurts do have N1b too. Ilumäe study just showed their N1c is all Y9022. Vepsians have high N1b but no Y9022."
That doesn't matter. I'm arguing Bolshoy is ancestral to Uralic(s) in Fennoscandia not anywhere else. I expect 'pure' Asian population similar to Nganasans to be the original Uralic(s). Bolshoy is a mixture between them and EHG/SHG.
All the Finns & others, alluding Bolshoy and SIberia had no importance to Uralic languages sound a lot like the Indians who say the Steppe has importance to Indo Aryan languages.
Uralic languages are not European. N1c is not European. Finns, Karelians, Saami are clearlly a two-way between normal northern Euroepans and some early Uralic population from Siberia.
Bolshoy fits the bill perfectly. Saami & Finns form a line between northern Europe and Bolshoy. Bolshoy goes 100% in line with what I and others have been saying for years. It's a population of mostly recent Siberian origin in Fennoscandia, carrying the most common form of N1c in modern Finno-Urgics. You guys are now doing gymnastics around the facts to still fit your ideas.
Let's get this straight. For years, most of us have been convinced these three things are related. Think about that, then look over Bolshoi DNA results again.
-Siberian ancestry in Finno-Urgics.
-N1c in Finno-Urgics.
-Uralic languages in Fennoscandia.
The daughter clades of N-L392 began expanding 1200 or more years after the clade itself formed and each clade underwent it's own expansion, e.g. the Finnic one started only 2000 years ago, a whole 2700 years after the N-L392 found in BOO itself. This is absolutely different from the R1b phylogeny as it relates to Bell Beaker and Western Europe, where the stars are so thickly nested and flattened, often within a couple centuries or less, as to look like an exponential growth curve were you to construct a tree of R1b clades found in Bell Beaker aDNA samples against time. Unlike the N scenario, the R1b is clearly part of a single sociocultural phenomenon, to the extent that we can speculate that some Bell Beaker groups, if they had detailed oral tales of genealogy, may have even remembered their common ancestor (like among N Africans whose genealogies often span 3 centuries or more), which is simply out of the question for Uralic N.
@Ryu, Often mHGs/yHGs currently a separta phylogeny instead of a stra-like one used to have a star-like phylogeny. For example, in early Corded Ware its R1a M417 had a star-like phylogeney but today R1a M417 has a separta phylogeny.
All three Bolshoi males belonged to N-L392 which today has a separta phylogeny. Back then, it may have been a star-like phylogeny. Bolshoi very likely represents the introduction of N-L392 into Fennoscandia like how Corded Ware represents the introduction of R1a M417 into Europe.
"Bolshoi is at best Para-Uralic, like EHG might be para-IE. "
Ok, so my talk on Y DNA has caused you to soften your rhetoric a little. To explain away the Y DNA link between Bolshoi and Finno-Urgics, you settle with Bolshoi not being ancestral to Finno-urgic but somekind of para Uralic.
Pathetic attempt to tip toe around the facts. Bolshoi is not comparable to EHG. It is comparable to Corded Ware.
EHG didn't carry R1a M417 or R1b L23. They carried dead-end R1b & R1a & J lineages. All three Bolshoi males belonged to N-L392. Like how essentially all Corded Ware males belong to R1a M417.
@Ryu,
I never compared N-L392 to R1b P312 and Bell beaker. Only you did. If anything, N-L392 and Bolshoi resemble R1a M417 and Corded Ware.
Corded Ware folk weren't proto-Slavs but they were ancestral to the proto-Slavs. They were Indo Europeans. So, I suspect Bolshoi were early Uralics who later became Finno-Urgics.
To say that "Uralic is not European" is just stupid. Uralic is the closest language family to Indo European, its closest in terms of possible cognates in core vocab and its the only one with an uninterrupted sequence of loanwords starting from proto-IE (c.f. Uralic words for 'six and 'seven') all the way up to Indo-Iranian and even Aryan (c.f. Uralic word orja 'slave'), tracking the sequence of IE languages on the Steppe almost exactly. Its developed in close proximity to Indo-European speakers for most of its history, the only reasons why we think of Eskimos when you mention them now is because they, like other originally pastoralist groups such as the Turkic Yakut and Dolgans who moved into Siberia as well, gave up horses and cattle and started focusing on reindeer as they moved north, as well as fishing and hunting more and metalworking less.
Boloshi were para-Uralic, they carried para-N1c-L392, etc...
Pathetic attempt to tip toe around the facts. Bolshoi is not comparable to EHG. It is comparable to Corded Ware.
Your 'facts' are not worth tiptoeing around, did you even bring any? And it was Aram who first brought it up, not you. He actually brought new facts to the table, not just more amateurish interpretation.
Bolshoi is clearly more analogous to Ukraine_mesolithic. The EHG+Nganasan mix in them is not even representative of the non steppic part of Saami.
I don't give a damn abt 'winning' or whatever, I care abt the facts and the fact that you are spreading misinformation. This point of view only became convincing to me only after Kristiina and Shaikorth introduced me to the voluminous literature on Uralic, which believe it or not is at least as important for genetic hypotheses for Uralic as the body of evidence was for IE studies.
@Ryu,
I'm open to the idea proto-Uralic had lots of Indo iranian Andronovo-like ancestry. I find the Andronovo scores in Global25 for Finno-urgics interesting.
If you and others weren't such jerks I'd be more open to it. Don't tell me to shut up because I'm "spreading miss information." In everything, I say I make it clear it is not fact but my opinion.
But, you state your opinion on Uralic origins as definite fact and then insult me for dis agreeing.
@Ryu,
"Your 'facts' are not worth tiptoeing around, did you even bring any?"
I never claimed to be the first person to bring them up. But they are facts. Each of us have a different interpretation. You shouldn't insult me for having different interpretation than you.
LOL duuuuude, typically what people do when exposed to new facts is becoming more uncertain about their previous position instead of digging in and disagreeing on reflex. You dont respond to something like 'actually Nganasan are Uralicised Yukagir cos we know from the 17th century ethnography' with the response: 'actually Nganasans cannot be Uralicised, they are by definition Uralic, just like Russians are by definition Slavic. They were never Uralicised just like how Russians were never Slavized. There is no Russian without Slavs, no Nganassans without Uralics. '
??????
"spreading miss-information"
If you worry about the spread of miss information on a DNA blog you care too much about this.
This is a place to exchange ideas. Let's do it peacefully. Let's not insult or tell each other to shut up for having a 'stupid' opinion.
Alllllll right all right, no prob. Pls at least acknowledge countrary facts in future.
Comb Ceramic is way too early for Uralic language expansion.
@Huck Finn
That means a lot of trekking!
In any case, I checked from Ilumäe et al that under L1026 there are four major branches, but there are ver many N3a3'6 haplotypes that fall outside of any of these main clusters. These haplotypes cover a very big area ranging from Karelia/Lithuania to Amur. L1026 formed 6400 yBp which is full Comb Ceramic period. Therefore, I suggest that a big amount of L1026 hunters were wandering around in Siberia and the Bolshoy line is one of these lines, of which the majority probably went extinct. Bolshoy had connections with Ymyakhtakh, but there is no need for only one haplogroup or one haplotype to be present in both. In any case, modern Yukaghirs do not carry L1026 on the basis of Ilumäe et al.
The ancient DNA would be very helpful to resolve these issues.
Ancient DNA shows that y lines usually become extinct. I think that the graph below showing the extinctions in R1a and R1b lines is revealing:
https://media.springernature.com/original/springer-static/image/art%3A10.1007%2Fs00439-017-1773-z/MediaObjects/439_2017_1773_Fig7_HTML.gif
We should not expect yDNA N to be any different. The situation can even be worse because this haplogroup used to live in areas where the climate was harsh.
https://link.springer.com/article/10.1007/s00439-017-1773-z
@Kristiina and re:"The ancient DNA would be very helpful to resolve these issues."
Even the subclade of paternal N in the BOO samples would be very useful. How do you BTW know that Ymyakhtakh was only related to Yukaghirs?
@ Huck Finn
Nobody knows that. It is just logical reasoning that can be right or wrong. They may also have been a multiethnic community.
According to Wikipedia:
A distinctive feature of the culture is the round-bottomed ceramics with waffle and ridge prints on the outer surface. Stone and bone arrowheads, spears and harpoons are richly represented. Armour plates were also used in warfare. Finds of bronze ware are frequent in the burial grounds. The culture was formed by the tribes migrating from the shores of Lake Baikal to the south, merging with the local substrate of the Bel'kachi culture. The carriers of culture are identified either with the Yukagirs ethnic group, or perhaps with the Chukchi and Koryaks, the Ymyyakhtakh culture continued at least until the first centuries of our era. It was later replaced by the Ust-Mil culture.
I used this reference to Wikipedia because it is easy to paste here, but you can google with "Ымыяхтахская культура" for more information.
Here is some more info in Russian: https://bigenc.ru/archeology/text/4915666
... that "migrating from the shores of Lake Baikal to the south" must be a typo and it should be "to the north":
ЫМЫЯХТА́ХСКАЯ КУЛЬТУ́РА, археологич. культура в основном позднего неолита (3–2-е тыс. до н. э.) гл. обр. на территории Якутии, до полуостровов Таймыр и Чукотка, севера Забайкалья и Прибайкалья.
@Ryu,
The Andronovo-like ancestry in Uralic Siberians can be explained by admixture with Indo Iranians. There's various explanations for the data.
If early Finno Urgics carried N1c-L392 yet were of mostly Andronovo-like origin that would mean the metallurgist, pastoralists Y DNA was replaced by Siberian hunter-gatherers similar to Bolshoi. Also, it would mean their language was replaced by Siberian hunter gatherers because we can be certain any Steppe-rich Bronze age people spoke an Indo European language.
If you think that's the case, is it so much of a stretch for you to say the early Uralic hunter gatherers imposed their language and Y DNA on Fennoscandia?
More info:
http://национальныйатлас.рф/cd4/28-30/28-30.html
http://wiradhe.narod.ru/yukaghir_weaponry.htm
http://mion.isu.ru/filearchive/mion_publcations/turov/9.html
http://www.libex.ru/detail/book828381.html
@ Sam
This is what happened in West Siberia:
We present the results of a paleogenetic analysis of nine individuals from two Early Iron Age mounds in the Baraba forest -teppe, associated with the Sargat culture. Paternal kinship appears to have been of greater importance. The diversity of mtDNA and Y-chromosome lineages among individuals from one and the same mound suggests that kinship was not the only motive behind burying the deceased people jointly. The presence of very similar, though not identical, variants of the Y chromosome in different burial grounds may indicate the existence of groups such as clans, consisting of paternally related males.
https://www.researchgate.net/publication/321071660_Kinship_Analysis_of_Human_Remains_from_the_Sargat_Mounds_Baraba_Forest-Steppe_Western_Siberia
We can see northern N1c men and southern R1a1 men in a peaceful kinship relationship and with all probability speaking the same language of the more advanced culture.
The same happened also with the Baltic Finnic N1c but we do not know where and when.
Reich seemed to refer to a similar development in his sketch when he draw arrows from local hunters influencing the Andronovo culture.
@Samuel Andrews
You should read some of the papers on the early achealogical and lingustic origin of Proto-Uralic as your view of PU is a little misguided.
Asuming N-L1026 is not of circum-polar BOO origin and was a new comer then it is quite likely they lived just north of the steppe in the Volga-Kama region (Volosovo/Garino-Bor). Garino-Bor prior to obvious IE influence already pacticed metallurgy and PU has word for copper which is not an IE loan. The northern portion of Abashevo then began trading with Garino-Bor creating Seima-Tubino which spread the Uralic languages.
@Chad,
"Lithuanians, clear to French and Brits appear to have Siberian admixture. Lithuanians even have a lot of N1 plus Siberian mtDNA."
Once again Chad I understand Lithuanians have Siberian ancestry. Yet, it is very insignificant so they are appropriate to use as a European ancestry proxy in Finns & Saami.
Anyways, about mtDNA. Lithuanians and Latvians have about 0.3% Asian mtDNA. I have close to a 1,000 samples from them. Not a noticeable difference from other Europeans. Finns also have very little Asian mtDNA, in my 287 samples 1% do.
Don't talk about mtDNA if you haven't studied it. Just kidding, I'm not an condescending jerk like you and half the people here.
More on mtDNA. Finns look like normal northern Europeans. H1a, V1a, V7a, H4a1a1a, T2b, H28a directly link them to northern Europe and specifically its EEF side. All of those lineages or close approximates have been found in Neolithic farmers.
On my blog, using young lineages found in EEF remains, I estimated Finns have 50% EEF mtDNA.
This goes against the narrative they're of largely Andronovo origin. I'm convinced FInns mostly descend from standard northern Europeans like Lithuanians and Scandinavians.
Looking at the genetic data, the Uralic question looked pretty simple. It derives from Siberians carrying N1c. Bolshoi fits this profile perfectly.
That's why I said it was very probable Bolshoi represents the introduction of Uralic languages in Fennoscandia in the initial post which Ryu attacked.
But, I do find the idea they also had Steppe ancestry intriguing mostly because of the interaction between Uralic and PIE then Indo Aryan. Also, the presence of U5b1b1a throughout Russia and SIberia.
Yet, Ryu, I find your claim that Uralic was mostly of Andronovo-like decent and represented well by Mezhovskaya a stretch. Mezhovskaya look like Indo Europeans who picked up some Siberian admixture. That's it.
Uralic(s) don't carry R1a like Andronovo did. They carry N1c which as we know from Bolshoi is of Siberian origin. Early Uralic(s) must have been of largely Siberian origin & closely related to Bolshoi.
Saami have more than twice as much Siberian ancestry than Mezhovskaya did. I find it hard to believe Mezhovskaya is therefore a good proxy for early Uralic(s). That's really a stretch.
If so, you'd have to suggest there's lots of non-Uralic Siberian ancestry in Saami. I guess you could say from people like Bolshoi who carried N1c-L392, the clade almost all European yHG N belongs to. So, what's going on here? Did Uralic(s) make no Y DNA impact? Or did they have lots of N1c-L392 but not much Sibrian ancestry?
@Ryu,
Also, ryu, Nordic IA isn't the best northern European reference for Finno Urgics. You should use lots of northern Europeans. I see you did for some of your tests.
Look at how Sycthian, Andornovo, whatever scores go down when you included northern Europeans other than Nordic IA.....
Finnish
Nordic_IA 65.30
Scythian_AldyBel 24.05
Narva_Lithuania 5.35
Samara_Eneolithic 3.30
Comb_Ceramic 2.00
Finnish
Nordic_IA 35.40
Belarusian 31.70
Slavic_Bohemia 17.05
Mezhovskaya 8.20
Comb_Ceramic 4.95
Srubnaya_outlier 2.70
Vepsian
Nordic_IA 38.95
Scythian_AldyBel 37.15
Comb_Ceramic 13.35
Globular_Amphora 5.50
Narva_Lithuania 5.05
Vepsian
Belarusian 68.35
Mezhovskaya 17.75
Slavic_Bohemia 7.55
Comb_Ceramic 5.75
Narva_Lithuania 0.55
For instance, Russians from southwest of Moscow are much less Uralic-like than the Kargopol Russians.
Yes, the line Moscow-Saint Petersburg marks the border between ethnical Russians with Slavic and Finnish origin.
When modelling using ancient sources the differences between Mezhovskaya and Saami become apparent.
Saami's Steppe/MN farmer ratio is much lower than Mezhovskaya. Meaning they have ancestry from "EEF rich" ancient northern Europeans.
Mezhovskaya.
Yamnaya-54%
Globular_Amphora-20%
Narva-2%
SHG-3%
EHG-11.5%
Siberian-11%
Saami,
Yamnaya-24%
Globular_Amphora-21%
Narva-14%
EHG-17%
Siberian-24%
@ Sam "Uralic(s) don't carry R1a like Andronovo did."
These are the R1a1 percentages of the Volga Uralics which is the probable area of origin the Proto-Uralic:
Vepsas 35.9%
Mordvinians 26.5%
Mari 31.4%
Udmurts 16.7%
Komi-Permyaks 36.7%
Southern Komis 5.8%
Besermyans 32%
Chuvashes 19.7%
(Underhill et al 2014 and Trofimova et al 2015)
Sargat paper shows that R1a1 is not recent in the Uralic groups.
The Andronovo-signal disappears when I use Swedish, Polish, Baltic_BA as northern European references instead of Nordic_IA. It's up to interpretation whether the Andronovo signal is real. D_stats reject it.
2.6779"
Saami
Swedish,35.1
Baltic_BA,20.6
Nganassan,18.1
EHG,16.5
Yukagir_Tundra,6.6
Narva_Estonia,3.1
Narva_Lithuania,0
SHG,0
Polish,0
Andronovo,0
1.2231"
Karelian
Baltic_BA,42
Swedish,36.7
Polish,6.6
Nganassan,6
EHG,5.3
Yukagir_Tundra,2.3
Andronovo,1.1
Narva_Estonia,0
Narva_Lithuania,0
SHG,0
@Kristiina,
"These are the R1a1 percentages of the Volga Uralics which is the probable area of origin the Proto-Uralic:"
Good point. I meant Finno Urgics. R1a in Volga Uralics could be from Slavic and Indo Iranian admixture.
@Kristiina,
"Sargat paper shows that R1a1 is not recent in the Uralic groups."
Yeah, but does Sargat represent early Uralic(s) or an early group of Uralic(s) who mixed heavily with Indo Iranians.
@Kristiina
Finns don't have much Andronovo ancestry, because if you did, you'd show a decent cut of R1a-Z93. But you don't.
Even the little bit of R1a that exists in Finns is R1a-Z280, which is some sort of minor paternal signal from Baltic_BA or related peoples.
So proto-Uralics, even if they did come from the Volga area, didn't carry a lot of R1a. In fact, most of the R1a in Uralics looks either Balto-Slavic or Indo-Iranian. There are no large R1a Uralic-specific clades like R1a-Z280 or R1a-Z93.
Finno-Urgics have lots of WHG the same as Balts do. This suggests to me, they have ancestry from the south Baltic where Narva lived.
The origins of the European ancestry in Finno-Urgics is a mystery. I won't lie about that. They aren't Balts, Scandinavians, British or any of the large northern European clusters. Associating them with Balto-Slavs always seemed wrong to me.
Their European Y DNA, which would point to regional origins, has disappeared other than their I1.
But, I do think it is quite possible they have lots of ancestry from people similar to Baltic BA because of their high amounts of WHG.
Oh, I forgot to add that Finns also have a little bit of R1a-Z284. That's from Scandinavians.
@Davidski
I have not argued that Finns have Andronovo ancestry. I have instead emphasized that we lack ancient data from the relevant areas and cultures. Finnish ethnogenesis is different from Volga Uralics, but the language probably came from there.
I must add that I was indeed very surprised to see that Leväluhta mtDNA is completely different from Finnish and Saami mtDNA. I did not expect that. I am also myself very surprised to see how recent Finns seem to be in Finland. Even if we don't get any older DNA from Finland, I hope that ancient DNA from Estonia, Karelia and Sweden will help to fill the gap.
Moreover, I do not believe any more that PU was introduced by yDNA N. YDNA N spread it Late Bronze Age/Iron Age in Siberia, but these languages were already its daughter languages that contained various Siberian substrates.
How close are Saami and Finnish language?
@Elliv J
Not that close, it seems, except for the loan words etc. due to later language contact. Even the idea of a distinct Finno-Samic subgroup of Uralic languages is in dispute. See e.g. https://www.sgr.fi/sust/sust266/sust266_aikio.pdf
@Kristiina
In your opinion then where the orihinal PU people R1a? Also then how did N1c become prominant in all Uralic speakers with each language group haveing their own specific subclades?
@Sam, bit wary of wading into this discussion and coming down on anyone's "side", and I'm sure I've missed some backstory but I think if we are talking of early Uralic speakers, it seems like we would really be looking for a position on the Global_25 where we *converge* upon an ancestor for the whole family.
The parsimonious solution for a hypothetical proto-Uralic group is probably to find a "point" that is fairly central to all the Finno-Ugric, and then allow for admixture with varying substrates to take away from that.
So it is probably important to consider the Mari, Udmurt, and Komi samples as well as the Saami and Finnish/Ingrian/Karelian samples you look to have modeled above?
You're probably not looking for a single vector of X -> Saami -> Finnish/Veps/Ingrian/Karelian but a point that can explain Mari, Udmurt, Komi, Saami, Finnish/Veps/Ingrian/Karelian by admixing in different directions with plausible substrates in circumpolar, Baltic, Volga, etc. (Perhaps also Hungarians, if we really, really want to go for it).
The Mezhovskaya individual does not seem like a bad point to begin with*, assuming some different admixtures in different directions: early Saami with Bolshoy like (then some level of more European admixture), Komis with people like the Volga Turkic speakers, Finns with Baltic_BA and Nordic_IA and so on. Or a point between the Mezhovskaya individual and the "Andronovo_outlier". Though I think a more optimal individual might share a little less specific drift with Yamnaya in very high dimensions and have a little more East Eurasian ancestry than Mezhovskaya individual.
(If we assume a point near where Bolshoy would be, assuming they're close to 50:50 Northern East Eurasian and EHG, then seems like you would need ridiculously high amounts of admixture to get to Finns, etc.! As I think Ryu notes.)
(Would also say, re Finns, there's a clear cline Swedish like Finns - HG00346, HG00266, GS000016896, HG00277, S_Finnish-3, Finnish_G253, Finnish_G252, S_Finnish_2, HG00350 - vs the Finnish East like Finns so this is perhaps an issue with calculating an average with them).
Interesting discussion anyway; have to admit, before reading this thread I did not worry too much about Uralic / Finno-Ugric and probably would have assumed something like Sam seems (some strange chance reversal of fortunes involving an encounter between pastoralist Indo-European group at the margins of productivity and a Uralic speaking North Siberian forager population). But the points Ryu has raised about the wider context of the language family as a whole do seem to make this pretty doubtful.
*Basing this on a resampling West Eurasian populations Global25 scores through PCA and looking "by eye", which I just did again following Iberomaurasian samples coming out: https://imgur.com/a/wwTvZ
@ Anthony "In your opinion then where the orihinal PU people R1a"
Maybe Proto-Uralic was influenced by Yamnaya R1b. (https://goo.gl/images/ZiF7fy)
Fatyanovo and Abashevo surely had a strong influence and they were with all probability R1a1.
There is the division between Baltic Finnic N-VL29 + Y28527 and Z1936 on one hand and Volgaic Y9022 on the other. N-VL29 and N-Z1936 are clearly much more frequent in Siberia than the Volgaic lineages. We will hopefully see what happened.
There are too few samples to say their mtdna is different from modern Finns. Estonians share recent mtdna with Finns if that means anything.
The link does not work, but this link to the Allentoft paper shows just us well that Yamnaya inluence radiates to the area of origin of PU: https://drive.google.com/file/d/0B9o3EYTdM8lQdzRlWXBQaXlnMGc/view
At the moment my reasoning goes like this: PU and IE languages are the closest languages to each other and share a huge amount of words and elements. YDNA N is from Siberia and yDNA R is from Circum Caspian. Therefore, their languages cannot have the same origin in the time frame of 15-10 kya and Uralic languages must rather belong to the more advanced cultures of the south just as the majority of their genes.
I know that there are many people who say that R1a1 and R1b are from Siberia. R will possibly be from Central Asia but it is too far back in time for PIE.
However, I will happily change my mind in this regard when the contrary evidence appears.
BTW there you can see a Finno-Ugric band displaying much of the Indo-European warrior spirit:
https://youtu.be/-2WqQY_xSSM
Underhill covered Estonians in their 2014 study
They are mostly of the M558 type of R1a , just like in russians and nearby groups.
It’s most likely pre Slavic, maybe even pre Balto-Slavic
@ Kristiina
"N1c and R1a living in peace"
Hahahahaha somehow I doubt that always held true. The *Arya 'free, noble, independent, peer/companion, kinsman' <----> *Orja 'slave' dichotomy probably tells us a bit about the interaction sometimes.
The 'six' 'seven' loanword point was mistaken, sorry, I pull out things from memory as I post and sometimes mix things up. The number is *sata 'hundred' and the proto-language level borrowings (or cognates) are things like *weti- 'water'.
The Sargat culture remains are very interesting. This cultural permeation (from agropastoralist to HG, leading to pastoralisation) is very interesting, we have a few cases of it (Argentinian Pampas, American West, Turkic and Mongolic Steppe peoples, one branch of Tungusic, and of course the Indo-Europeans) but nowhere do we see in the process as it took place all that clearly, maybe except for Plains Indian ethnography. So the archaeology will be very worth researching.
Well, this is great, there are new facts on the table! So we can move forward.
When I said they were influenced by Andronovo and Sintashta pastoralists, I was not exact. I was using steppe MLBA groups as shorthand for all Eastern European cultures with aDNA samples that had both EEF and CHG ancestry (not just WHG, SHG or EHG), and were agropastoralist and not just hunter-fishers. Later archaeological cultures that have been assigned to Uralic speakers, like Cherkask'ul and Mezhovska of the forest-steppe may have been influenced by Andronovo and Sintashta to a degree, but the cultures most often assigned to proto-Uralic, the Fatyanovo-Balanovo and Abashevo complex, arose from interactions between HGs and Corded Ware in European Siberia, Coded Ware that spread from the Baltics to the Volga region. So the Y-DNA data that Davidski and Kristiina have been talking about (the R1a in Uralics being closer to those in Balts than those in Slavs and Indo-Iranians), as well as Sam's finding that the European ancestry in Uralics is somewhat WHG-rich, akin to Balts, are very interesting in this regard. Baltic-BA may well be what we need.
@ Matt
Cline intersection for Uralics is probably a good idea. The issue is to get enough resolution we may need a Siberian or North-Eurasia-specific PCA, like the one in this paper:
PCA
with Uralic and Turkic clines marked, and position of ancient Scythians indicated in purple (by judging from Global25 closest populations to Scythians, we can position them in this PCA as well):
Link
You can see there is enough resolution to perform cline intersection on Turks, which will cause the proto-Turkic population (or at least the last population from which all Turks derive ancestry) to be akin to Altai populations like Tuvinian, Buryat, etc, because the clines from Yakut to Evenk, from Turkmen to Kyrgyz and from Tubalar & Shors etc to Altaian intersect there, but the Uralic cline is still pretty linear and thus we dunno if the cline is cos of the Asian side spreading West or the European side spreading East.
But if we have a PCA consisting of just European Siberian populations then the cline intersection will cause the inferred ancestral population to be very, very European:
Uralic-centered PCA
The Finnic ENA and Volga Uralic ENA are different, and the Uralic populations fan out from Europeans onto a cline of varied ENA ancestry.
On a larger scale: Eurasia PCA
Annotated with the "fan": Link
PCAs H/T M. Myllyla.
This is validated by a published study, except the populations at the ends of the cline (Mansis and Samoyeds on one and Saami on the other) were not sampled in this case:
Uralic-centered PCA
In fact I even suspect the population may even have had a very small amount of ENA originally. I've heard a rumour that there is some aDNA from proto-Finnic populations in the Baltic area that is basically European, with some HG admixture from Combed Ware and so on.
@ Sam
If you use modern or recent pops they are gonna be super attractive to all populations, e.g. add Chechen or Dai or even England_Roman and see what happens.
About Nganasan being recently Uralicised Yukagir and Evenk from their ethnographic history, the autosomes corroborate this as well:
Pugach et al PCA
@ Davidski
Hey David, do you mind performing a Global25-style PCA for just Eurasians (strip all out African-admixed pops) once you get the aDNA for this paper, maybe raise the dimensionality to 30? We have so much aDNA from North Eurasia and Siberia, this will flesh out the higher dimensions specifc to Eurasia, and will be very popular I'm sure.
The link to the Eurasia PCA without annotations is broken, relinked here.
Kristiina: "I was indeed very surprised to see that Leväluhta mtDNA is completely different from Finnish and Saami mtDNA."
In fact, the two Saami-like Leväluhta samples are in the PCA (Supp. Fig. S3) still somewhat pulled towards BOO.
The outlier JK20165 is - other than reported by Sam - , not plotting atop of Lithuanians, but instead clustering with Norwegians, Orcadians, Scotch and Icelandic. Czech, English and a few Hungarians are also not far apart. My best guess for her is a "Varangian" origin (note that the PCA neither includes Swedish nor Danish samples).
This might Germanic (Swedish) settlement of Ostrobothnia push a bit further back in time than commonly assumed, but would otherwise not be such a surprise in view of the strong commonalities already shared by Swedish and Finnish Battle Axe Cultures.
@ Davidski
What is Baltic_HG?
@Ryu,
It would make sense Finns are a combo between Saami and Estonian type pops meaning proto-Finns had little Siberian ancestry. It's important to note Saami aren't Steppe rich. It looks like they have about 25%. Their most important European contributors might be northern hunter-gatherers like Narva and Comb Ceramic. But if they have Baltic ancestry, their Steppe & Anatolian (almost 20%!) & Narva, ancestry would have come together.
@Kristiina
"Maybe Proto-Uralic was influenced by Yamnaya R1b"
But then why does the Uralic language family exist at all? R1a and R1b have been seperated for over ten of thousand years but still 95% of both speak IE languages so what made Uralic unique?
Not to mention N1c can easily be linked to the spread of PU while R1 in Uralics is easily explained by contacts with their IE neighbors (U106 in Finns from Germanics, etc)
"There is the division between Baltic Finnic N-VL29 + Y28527 and Z1936 on one hand and Volgaic Y9022 on the other. N-VL29 and N-Z1936 are clearly much more frequent in Siberia than the Volgaic lineages. We will hopefully see what happened
Sure, and there are Siberian, Volgaic and Balto-Finnic branches of Uralic and modern lineages point to a Volgaic origin for L1026 even more so that a Kama one (all though maybe due to European sampling bias)
What would have happened to the orighinal R1 in Uralics then? This is the same argument as the continuation hypothesis, HG Siberian N1c replaces more advanced R1 metal working cultures to the south who likely lived in greater numbers.
"The Finnic ENA and Volga Uralic ENA are different, and the Uralic populations fan out from Europeans onto a cline of varied ENA ancestry".
This fact alone should prove N1c was originally Uralic even if PU does have some ENA like ancestry the vast majority of eastern ancestry in modern Uralics came from different sources not a single N1c carying BOO source.
@ryukendo kendow
You can look up the individuals that make up the population averages in the non-averaged datasheets, after which you can find the source of each sample using their IDs.
Baltic_HG:Spiginas4
https://www.nature.com/articles/s41467-018-02825-9
@Rob
The R1a in Estonians isn't pre-Slavic or pre-Balto-Slavic. They mostly carry the usual young subclades under R1a-Z280 that are found in Balts and NW Rusians.
@ Dave
IMO, Z92 is pre-Slavic
@ All.
Using only ancients, this is what we could get for some of the groups
https://imgur.com/a/xU9VB
The inflated HG-ancestry in East Baltic and Volgaics is from BA Baltic and extra EHG, not Narva or CCC.
its hard to see how Uralic could be linked to anything other than something east of the Urals.
Second set of fits for Uralics done. This set is very clean.
The Saami % <--> Mezhovskaya % + Nganasan % phenomenon is pretty interesting to see. Anthony Haken, your point that Mezhovskaya seems associated with Ugrics seems connected to the high percentages nMonte assigns them.
A very, very nice point to see confirmed is that Finnics and Saami do not have ancestry from Scythians in this set, while all other Uralics do. This is in conformity with the discovery that all Uralics have rare allele sharing with Mongolians, except for Saami and Finnics. Fairly impressive that rare allele-based data is confirmed by nMonte fits, if you ask me.
@ Rob
You need to add Saami to the sources, all Altai IA and EHG becomes 0. If not Saami, then Mezhovskaya + Nganasan will work as well.
The Scythian ancestry is probably better represented by AldyBel, Altai_IA is very East Asian and has too much ancestry from the Middle East/Iran.
@Rob
IMO, Z92 is pre-Slavic
But Estonians don't carry Z92 per se. They carry subclades derived from Z92, which are the same young, highly derived subclades that are found among Balts and Slavs.
So there's nothing pre-Slavic about their R1a. The issue is with the resolution in most studies published to date, with the authors not testing any subclades beyond Z92.
If you look at R1a phylo trees based on full sequence data, you'll see that Estonian R1a lineages aren't basal to those among Balts and Slavs.
@ Dave
Z92 and derived clades arent pan-slavic, and are therefore pre-Slavic, unless your'e suggesting that the Slavic ethnogenesis occurred in Novgorod
@ Ryu
Yes I know Saami will diminish the amount of EHG and Altai needed, although not completly. Although I find using moderns somewhat unsatisfying :)
I also find adding Scythian does odd things (oddly elevated). I doubt these groups have much Scythian ancestry
@Rob
Estonians carry highly derived R1a-Z92 lineages that they picked up from admixture with Balts and Slavs after the formations of these groups. Hence, definitely not pre-Slavic.
@ Rob
Then Altai_IA will be even more unsuitable, because it is just before 0AD and was dug up on the borders of China, in the territory of the Yuezhi. If other populations are included they displace it immediately.
You should check out some of the fits I linked to; Scythian_AldyBel behaves very well, peaking Volga Udmurts and decreasing smoothly among geographically close populations, just like a real contributor would.
@ RK
I was using it as a very crude / broad standing for eastern ancestry and differentiator from other local / western components; not pretending that it is *the* source or answer for all / any FU groups
Yes I saw - nice fits
What do you suppose Nganasan is standing in for ?
Looking at the current paper's PCA, it lies at the end of the Uralic cline, and so it (and Yukaghir as well) may be a good proxy for "circumpolar ENA".
There are many PCAs with ENA clines stretching from Southeast Asians up to Siberians, extending to Native Americans; some PCAs even extend the cline to ANE and WHG. The migrations that crisscross Eurasia connect populations from the West to a particular position on the ENA cline. It seems judging from BOO that there was some EHG-Nganasan continuum in the Tundra and Taiga, and migration of European-like populations onto this continuum created the Uralic "fan".
Hmmm, I'm gonna make a prediction just for the heck of it. ¯\_(ツ)_/¯ ¯\_(ツ)_/¯ ¯\_(ツ)_/¯ Let's see how much of this holds true in the coming years.
Proto-Uralics: Baltic_BA + EHG + teeny bit of BOO
Proto-Finns: Proto-Uralics + even more EHG, Combed Ware and such
Proto-Saami (not in Finland, but further into Russia): Baltic_BA + EHG + even more BOO
Estonians = Proto-Finns
Finnish and other Finnics = Proto-Finns + proto-Saami
Volga Uralics = proto-Saami + Scythians
Ugrics = proto-Uralics + Kets and Yukagir-like people
Samoyeds = Ugrics + even more Yukagir-like people
Nganasans = pure Yukagir-like people
BOOM
@ryukendo kendow
The consensus is that Pre-Proto-Uralic came from Siberia, and then Proto-Uralic formed in the southern Urals.
I haven't seen anything in the ancient DNA to date to contradict these theories.
So I can't see Baltic_BA, basically Proto-Balto-Slavic Trzciniec culture, as the main contributor to Proto-Uralic.
@Davidski
Baltic_BA may be the closest we have to Fatyanavo-Balanavo people? Also Abadhevo likely had contact with PU.
Has there been any word on the progress of the IA Estonian samples from Tarand graves?
They will likely be the next Uralic related samples we see and will be vital to understanding Proto Finnic as well as PU. Probably even more so that this paper.
It will be interesting to see how much if any BOO/ENA they carry.
The latest wave of Turkic peoples expanding out from South Siberia, whose segments were tracked in Yunusbaev et al., was comprised of a population that was ~60-70% derived from Scythian_Pazyryk or people very like them, with the rest like Mongolians, Daur, or other East Asian groups south of the Gobi.
In fact East Scythian ancestry, or alternatively the ancestry of people speaking Northeast Iranic languages, is best preserved in Turks today. The distribution of East Scythian ancestry and Turkic languages virtually coincide.
So I think this kind of shift can definitely happen. Pre-proto-Uralic itself was probably spoken by the other half of the almagating community in Fatyanovo-Balanovo, the Volosovo hunter-gatherers with some of the clades of N1c and N1b found in Uralics today. I guess they may have been very EHG-shifted versions of BOO, and may have traced their ancestry far to the East at some point. Some of the similarities that Juha Janhunen uses to hypothesize that Uralic has had contact with dialects of the Altaic language area also extend to Indo-European and at some very remote past they may have had common origin or contact, in "deep EHG" or smth like that.
@Anthony Haken
Baltic_BA may be the closest we have to Fatyanavo-Balanavo people? Also Abashevo likely had contact with PU.
Not really. Abashevo and Fatyanavo-Balanavo will be very similar to Poltavka_outlier, Sintashta and Andronovo, and there probably will be a lot of R1a-Z93 in Abashevo.
The Baltic_BA effect in Baltic Finns is mostly produced by a high level of various types of Baltic HG ancestry plus indirect gene flow from various Corded Ware groups, mostly from Scandinavian Battle-Axe. Estonians do have a lot of Baltic_BA ancestry, but this is mediated via Baltic and Slavic admixture, rather than Abashevo/Fatyanavo-Balanavo.
You can see this in non-Finnic Y-halogroups that Finns and Estonians carry. Finns have a lot of Scandinavian I1, while Estonians a lot of Balto-Slavic types of R1a, with practically no Z93.
The CHG-signal is unique to Indo Europeans. UkraineHG didn't have it, Comb Ceramic didn't have it, Bolshoi didn't have it. Uralic languages may have originated near IE languages but that doesn't mean they were genetically similar. If anything their European ancestry would be mostly EHG & ANE not classic Steppe.
@Samuel Andrews
Uralic speakers have been in contact with IE speaker since the PU stage judging by
Proto-Aryan loanwords. Thus PU likely did carry "classic steppe" DNA.
@ Samuel
That is just incorrect.
Even unadmixed Mansis and Khanties in central Siberia have ~25% modern European ancestry, with CHG, and they weren't even in contact contact with Russians until well after the Middle Ages.
@Anthony Haken & ryukendo kendow
There's no reason to assume that Proto-Uralics incurred admixture from early Indo-Europeans to any significant degree, considering the widespread presence of Corded Ware and Yamnaya-related Andronovo and Scythian peoples across Central Asia and Siberia, whose descendants probably mixed into a wide variety of native groups before the Russian expansion east.
For instance, the Baraba steppe in Western Siberia was a major contact zone between Indo-European steppe peoples and Siberian foragers from the Middle Bronze Age onwards, well after the generally accepted Proto-Uralic ethnogenesis.
Like I said, it's likely that the Pre-Proto-Uralics came from Siberia, and the Proto-Uralics from the southern Urals. In all likelihood they were rich in N1c and various stripes of Siberian genome-wide ancestry, and had minimal European admixture until they were deep in Europe and already breaking up into different groups.
@ Davidski
I have to disagree with you here. People like Carpelan, Parpola, Hakkinen and so on, even after breaking with the "meolithic continuity theory" for Uralic and switching to the "short chronology", so to speak, were still talking about Volosovo Hunter-Gatherers, a subgroup developing from the Pit Comb Ware culture, as proto-Uralic, with proto-Finnic and proto-Saami reaching Scandinavia among the hunter-gatherer Textile Ware culture that developed from Volosovo.
The placement of a proximal homeland in Volga region derives from linguistic paleontology considerations, and is fairly secure.
The chronology, however, is dealt a crushing blow by the fact that Saami people are now known to reach Fennoscandia only in the Iron Age, and Finnics even later than that. Something like this was never foreseen by them at all. So the chronology is even shorter than those who proposed a short chronology themselves had dared to imagine.
Furthermore, we also know that the Pit-comb ware culture (also known as Combed ware) were pure EHG-WHGs and had no Siberian ancestry of any kind, giving us a prediction for Volosovo. The Bolshoi Oleni Ostrov people, dated and worked on properly only a decade ago, is connected to the Ymiakhtakh culture associated with Yukagir and related peoples practicing a circumpolar hunter-gatherer lifestyle (the significance of this will become clear later).
Juha Janhunen's school, which was the one believing that Proto-Uralic loaned words to proto-Yukagir, believed that Yukagir was in Northeast Siberia all along, so not only does pre-proto-Uralic have to be in contact with Altaic (due to shared features) proto-Uralic has to be in contact with Yukagir. This makes him place things very far East.
But now we can see that the BOO people show influence from Yukagir-like (which Nganasan represent) people autosomally as well. So contact between the two groups could have taken place much further west than previously assumed.
This means Janhunen's ideas are no longer supported, and at the same time, some of the considerations holding the "short chronology" school are also lifted, such as the fact that every single branch of Uralic now has a spread associated with agropastoralists, and none with hunter-gatherers only. So Uralic is likely to be much younger and much more associated with agropastoral people than prior scholars made it out to be.
Furthermore, the fact that Estonians and likely the Iron Age proto-Finns are only shifted towards EHG (you can see this in the Eurasia plot whose link was broken the first time I posted it), Saami towards both EHG and Nganasan, and other Uralics strongly towards Nganasan is also explained under this scenario by the dispersal of a population that had very little or no Nganasan ancestry originally, but quite a bit of EHG.
Its very interesting because this is one of the cases where aDNA shatters an academic consensus (for Saami and Finnics) in a quite shocking way (unlike the case of IE). Will be very interesting to see how the scholars react.
^^ Just to add, I think that the shared features of the languages that make up the Altaic linguistic are, plus Uralic (which is what Janhunen focused on) but also including Indo-European and some languages associated with this bunch (Nivkh, Eskimo-Aleut etc.) which Greenberg calls "Mitian", from *m- for 'I' and *t- for 'you', and which he thought was most closely related to Amerindian languages at some remote time, now seem to have some kind of association with EHG-ANE. Maybe they reflect an ancient linguistic spread or contact area associated with the Mammoth Steppe.
Diverging From the topic at hand, just saw a video by G.Chaubey regarding India. He says the preliminary results of the ancient DNA from rakhigarhi suggests against any migration into India. He mentions a 'Dravidian' invasion into north west India. I suppose this means that rakhigarhi samples had some ASI perhaps more than most modern natives of that region. More interestingly when asked about evidence for an out of India migration he mentions a paleolithic migration Of haplogroup H from India to Iran and west asia. Avoids mentioning r1a and then specifically says there were no large scale population movements out of India either.
@ryukendo kendow
Nothing's been shattered. Everything still makes good sense if you consider that even in Northeastern Europe closely related Finnic groups have somewhat different origins.
You're making this too complicated. Focus on the shared ancient ancestry between the Finns, Estonians, Saami and the Uralics of Asia, which does include N1c and Nganasan-related input for all of them except Hungarians, and it's all rather straightforward.
I don't find anything that I've read here that deviates from the general consensus about Proto-Uralics very convincing.
@Jijnasu
Honestly, I couldn't care less what Chaubey thinks. I'll take the word of Reich or someone else from Harvard over his any day. In any case, the ancient DNA is already speaking for itself.
The linguistic consensus regarding the spread of Saami and Finnic has quite literally been overturned by this new discovery. We will have to wait to see how this impacts linguistic models.
I also think you'll find Latvians need no more ENA than what the mostly EHG Combed Ware people had, insignificantly more at most, and this is likely the case for the Iron Age Finnic (now probably proto-Finnic) samples as well, who may be purely Baltic_BA + Combed_Ware.
The rest will await aDNA, let's see then.
Agreed, but in any case it seems that they're backing down from previous claims of an r1a expansion from India something OITers should note
@ryukendo kendow
Well, I don't know about Latvians, but Estonians sure have some Nganasan-related admixture, and the reason they don't have much more is because they mixed a lot with nearby Baltic and Slavic groups, like the early Latvians, who had none.
You need to take this sort of stuff into account.
The idea that Iron Age Finnic and Proto-Finnic groups were basically Baltic_BA + Combed_Ware is downright ridiculous.
Jijnasu,
Can you share a link to the video ?
There was no migration of ASI to the north. There was a bounce back in native ancestry in Northern South Asia after Neolithic farmers from Iran mixed more readily with local foragers, just like in Europe during the Middle Neolithic.
Obviously, Chaubey's confused because at least some of the Harappans are more Iran_N-like and have less ASI than modern Indians in the region. Duh.
Some Harappans or all Harappans sampled so far ?
So the cat is coming out of the bag. Has there been any Harappan sample with more ASI than modern samples ?
A DNA paper on the roots of Estonians is coming! Other papers from relevant areas will hopefully also be published. Therefore, I do not want speculate too much at this stage, but of course I broadly agree with Ryukendo and think that this is what the future ancient DNA will confirm.
Populations (and different yDNAs) adopt new languages. We have R1b-L23(xM412) in Yamnaya, R1a1-Z283 in Corded Ware, R1b-L51 in Bell Beakers, I2a2 in British and Central European Bell Beakers, I2a2 in Czech Corded Ware, J in ancient Greeks, I1 in ancient BA Scandinavians and N1c in the forest Volga Ural, if the ancient DNA so reveals.
@Jaydeep
The cat has been out of the bag for a long time on this blog, but you've been ignoring it.
There was an invasion of India by peoples from the steppe during the Bronze Age. That's the real story and you'll have to deal with it sooner or later.
@jaydeep
https://t.co/mIg681v7xH?amp=1
He speaks at around 27 mins
@Kristiina
The main paternal marker of Corded Ware males can only be described as R1a-M417, because Corded Ware samples belong to both R1a-M417 (Z645+) and R1a-M417 (+L664,xZ645,xZ283).
Very few belong to R1a-Z283 thus far, probably because it's too young to be an important Corded Ware marker.
Here is an article regarding the Estonian interdisciplinary research project, covering linguistic, archaeological and genetical aspects:
https://opetajateseminar.files.wordpress.com/2018/03/eestlaste-kujunemine-2018-tekst.pdf
David,
I asked you a simple question. Why don't you answer that ?
Did they find any ASI in the Harappan samples ?
Lol I have a feeling that the first samples will have 10-0% ASI and the later samples much more, akin to the EEF and WHG pattern in Europe in early vs middle Neolithic, and this will be endlessly, endlessly spun.
@ Rob
Interestingly, the culture commonly associated with para-Uralics is Comb Ceramic, which will also explain why the Finnic branch of Uralic is so conservative and changed so little (cos the substrate experienced by this branch as it moved was typologically, lexically and phonologically similar). All other branches, including Saami, have obvious substrate influences, with Samoyed having the most, to the extent of having large parts of its vocab replaced. The substrate of the Saami was probably the language of the BOO people.
@Ryu and re:
Proto-Uralics: Baltic_BA + EHG + teeny bit of BOO
Volga Uralics = proto-Saami + Scythians"
Why wouldn't Mezhovskaya (being itself an outcome of an admixture event between something Siberian and the local Ural area genepool) do the trick?
Proto-Uralics: Mezhovskaya
Volga Uralics: Mezhovskaya + Scythians
^^ Whatever cognatic connections between Uralic and IE were probably mediated by some "EHG unity" at some point too ancient to be traced.
Alternatively we may be getting the same kind of situation with "Amerind languages" in the Americas, where the pronouns and some vestigial vocab show that they are all obviously connected, its just too statistically unlikely otherwise, but its the nature of HG spreads that languages change super fast and the signal gets attentuated very quickly, beyond the ability of the comparative method to discover it.
@Jaydeep
It's not a simple question until the ancient data come out.
Have a look at that panel discussion. There's no mention at all of any of the ancient data from Eastern Europe with loads of R1a.
Chaubey totally avoided that topic and instead was blathering about Y-haplogroup H, seemingly trying to suggest a migration from India to Iran and Armenia. What is that, a new version of OIT?
This is no longer very funny. I see crazy people.
@Anthony Haken
"But then why does the Uralic language family exist at all? R1a and R1b have been seperated for over ten of thousand years but still 95% of both speak IE languages so what made Uralic unique?"
The apparent linguistic connection of R1a and R1b is nowhere near as old as the haplotypes themselves. They happened to be both present in the region where PIE formed, then big population expansions made the variants present there dominant in a wide geographical range.
For example R1b-V88 carriers do not speak IE and they never did.
Also there is a lot of R1a and R1b carriers today that do not speak IE languages (various Turkic groups to begin with), so Uralic speakers are not special on that matter, there your question is not a valid one.
BTW, somebody talked about the Nganasan as something classic Uralic. I just revisited the PCA-s I made with Davidsky earlier data-set. What is very striking with Nganasan is not that they are far away in extreme Siberian direction from the bulk of Uralics (that could be still a cline), but that there are important-looking dimensions where they do not form a cline at all. In Eurogenes World PCA PC4 and PC7 they are pulled completely out of any Uralic cline in a direction where they are father away from all other Uralics than Buryats are. It is especially interesting in PC4, where all Uralics cluster very closely, except the very far off Nganasan and the Nenets, who are halfway.
I think Nganasan (and in lesser extent Nenets) have some kind paleo-Siberian ancestry that other modern Uralic speakers lack.
It's not a simple question until the ancient data come out
Ok. But what is the inside info that you've received so far tell you ?
If you remember, you have been saying that an ancient migration from South Asia to the steppe is impossible because there was no ASI admixture on the steppe. And I have argued that ancient South Asians from North India / Pakistan might not have had ASI in the Bronze Age.
So am I being proved wrong there ?
@ryukendo kendow Do you think any part of the Corded ware culture was Uralic speaking? Before the recent genetic papers in 2015, Corded Ware was assumed to have spread Uralic languages in northern Europe. Even now it enjoys support in some models. (see https://indo-european.info/ie/Indo-European for example)
@Chetan
Even if Corded Ware were N1c-carriers they couldn't have been Uralic, the time depth for the language family just isn't there.
@Chetan
Before the recent genetic papers in 2015, Corded Ware was assumed to have spread Uralic languages in northern Europe.
No it wasn't, except maybe amongst some fringe academics. The academic consensus was always that Corded Ware people were the first Indo-Europeans in Northern Europe.
This hasn't changed.
And it won't change thanks to ancient DNA, because Corded Ware was too old even to be archaic Uralic and Corded Ware was almost exclusively R1a, with no N1c, so not typically Uralic in terms of Y-haplogroups.
You seem more than a little concerned with what kooky Carlos is scribbling at his blog. The guy is making a fool of himself, and you are too by taking him seriously.
Uralic is a super young language family, only as old as Indo-Iranian at the most.
(IIr including Aryan).
@Shaikorth Isn't the Uralic family too old for an Iron Age diversification? Proto-Uralic is usually dater around the same time as Proto-IE further evidenced by loans from late PIE in the earliest stratum of PU. So the languages must have been in contact much before 2000 BC.
Also, there is linguistic evidence pointing to a Uralic adoption of late PIE dialects in the Urals region causing palatization/satemization. I find this argument reasonable at least in the case of Indo-Iranian. That is, Uralic/related language speaking Abashevo adopting the speech of R1b Poltavka chiefs in the Urals region and giving rise to the Indo-Iranian community.
Granted there is no genetic evidence as yet supporting this unequivocally, but still I think that is a possibility.
@Davidski
We only have samples from Western Corded Ware, nothing from Russia, so I would not say that the almost exclusivity of R1a is a settled matter.
Also it is not too old for proto-Uralic at all. The age of Uralic is estimated to be over 4000 years, a Corded Ware subgroup as a predecessor would be actually a reasonably good fit by time.
If I have to decide from current data, I'd also say Uralic is from Siberia with a very high probability, but strictly speaking neither YDNA nor time are dead-blows fro the CW theory for now.
@David Maybe I am taking it too seriously, but I think Carlos' work can't be dismissed so fast. He does make strong arguments in his model. But let's just leave it at that, at least until more genomes keep coming :)
@Chetan
As Ryu said, it's comparable to Indo-Iranian in age so pre-Iron Age but still too young to be Corded Ware. The early PIE loanwords are controversial.
@Jaydeep
My information is that there is ASI admixture in the Harappan samples. But this is irrelevant, since I also know that the Harappan genomes are mostly Iran Neolithic.
So your theory that there was a Proto-Indo-European migration from India to the steppe doesn't work, since it's not only horribly contrived, but there is no Iran Neolithic ancestry in Eastern Europe.
http://eurogenes.blogspot.com/2016/07/early-neolithic-genomes-from-eastern.html
@David And it seems some of Reich's team read his blog (as I'm sure they read this blog too). A Nick Patterson recently made some comments there!
@Chetan
Maybe I am taking it too seriously, but I think Carlos' work can't be dismissed so fast. He does make strong arguments in his model.
No, he doesn't. He's either totally insane, or he's running some sort of propaganda effort. Or both.
Kooky Carlos picks whatever he can find to back up his flimsy models. He even had to resort to citing some fringe Finnish academic, I forget his name, to try and claim that Corded Ware was Uralic.
But no one with any sort of reputation will ever argue that the Corded Ware people carried Uralic languages all the way across North Central Europe up to the North Sea. This is insane. I've never even seen any claims by anyone serious that Abashevo was Uralic.
You're falling for online propaganda by a crazy person.
@Shaikorth&Ryukendo
This is probably an over-estimation, but just an example how much the young age of Uralic is not a matter of fact consensus you seem to imply:
https://onlinelibrary.wiley.com/doi/abs/10.1111/jeb.12107
Nevertheless it is definitely older than Indo-Aryan.
I just found this: Radiocarbon Chronology of Complexes With Seima-Turbino Type Objects (Bronze Age) in Southwestern Siberia
https://doi.org/10.1017/RDC.2017.24
Assuming Seima-Turbino has anything to do with the spread of Uralic languages, this would make the expansion and diversification older than 4000 years, but still quite younger than N1c and Siberian arrival into Fennoskandia.
@ Chetan
Where has RK demonstrated that FU is ‘super young “?
The former idea was Mesolithic now contemporaneous to PIE (~ 4000 BC- Hakkimen)
Lol does Indian have a Uralic substrate ?
But seriously, there’s has been proposeals that Germanic has an FU substratum, as well as Balto-Slavic (Thomas & Kauffman).
Yes, obviously Kooky Carlos is a genius, and Uralics spread with Corded Ware, all the way to France.
True story.
@Rob
Years ago I was straightforward called an idiot by a Hungarian academic linguist for mentioning that proposal (Uralic substratum of Germanic). I could not even find a link or reference for it back then.
@ Slumberry
Eg it was mentioned in 'THE PRECURSORS OF PROTO-INDO-EUROPEAN: THE INDO-HITTITE AND INDO-URALIC HYPOTHESES' Conference @ Leiden University, 9-11 July 2015
I have the Abstract book
@Rob
I have no connection to Kristiina outside of this blog and I have no idea who is she (or anybody else here, as I intentionally never tried to look up anybody here, just to emphasize to myself that it does not matter, arguments and data matter).
I have to go now.
@Slumbery "The apparent linguistic connection of R1a and R1b is nowhere near as old as the haplotypes themselves. They happened to be both present in the region where PIE formed, then big population expansions made the variants present there dominant in a wide geographical range. "
This summary. Couldn't have said it better
Uralic languages are associated with N1c and Siberian admixture. So if Proto-Uralic is older than, say, 4,000 years, then we already have the Proto-Uralics in the ancient DNA record, and no, it's not Corded Ware.
If Proto-Uralic is much younger than 4,000 years, then a bit more digging around in Northeastern Europe, around the Urals and in Western Siberia will put the Proto-Uralics into ancient DNA record. And they won't be like Sintashta, Srubnaya or Andronovo; they'll have a fair whack of Siberian ancestry.
It's just common sense folks.
@ Slumberry
My mistake . I got my wires crossed
The point was - there is an abstract book online of the said conference which discusses the issue
@Sumberry
Whoah man so defensive (?!)
For some reason I confused you with shaikorth - my confusion
The point was - there is an abstract book online of the said conference which discusses the issue
@Rob I didn't say Uralic was "super young". You are talking to the wrong person lol)
No Uralic substrate has been proposed in Indic so far and whatever word exchanges happened usually went in the opposite direction. So yes that is an apparent block for Carlos' model.
I tried suggesting that maybe the R1a in Indic language speakers came from a population related to the Ukraine sample maybe that's why it didn't contain significant substratum effects.
It has been proposed that satemization in general was a result of Corded Ware groups adopting IE (from Poltavka in case of Indo-Iranian and from Polish beakers in case of Balto-Slavic). That is the crux of Carlos' model.
Kooky Carlos' models are highly theoretical and horribly contrived to fit the ancient DNA data.
They basically rely on a lack of ancient DNA sampling, but as soon as some of the relevant samples come in, they look worse and worse.
@The substrate of the Saami was probably the language of the BOO people.
That would be logical considering the high amount of Siberian in Saamis.
However, the only male sample (CHV002) that is Saami at 100% probability is yDNA I2a1 as Mesolithic Scandinavians. He is from Chalmny Varre in Murmansk Russia and he clusters closer to modern Saamis than Leväluhta people.
It seems that there was the BOO language that is from the east and a Scandinavian hunter language represented by I2a1. Nobody knows if the subtrate in Saami comes from both.
If you are interested, you can read Ante Aikio's essay "An Essay on Saami ethnolinguistic prehistory", in which he identifies several possible substrate words in Saami languages.
@ Dave
Yes CWC was R1a with some I2a, so it obviously cannot be FU (and thus not worth discussing) but somewhere near it’s eastern end where there was N1c etc , a post CWC group might have admixed into FU groups.
Well no doubt that some descendants of Corded Ware mixed with Finno-Ugrics in a whole bunch of places, from Scandinavia to the Urals and maybe beyond, depending on if Andronovo actually came from Corded Ware.
But to suggest with a straight face that Corded Ware was Uralic, and then all sorts of Corded Ware groups were Indo-Europeanized in different parts of Europe, by Poltavka here and Bell Beakers there, and just happened to retain linguistic and genetic ties (like Balto-Slavs with Indo-Iranians).
It's beyond stupid. It's the work of an insane mind.
On a side note, I really dislike the labels ANI/ASI. They are both misleading. The component we call ASI was THE major Indian component before the Neolithic. It's to India what WHG is to Western Europe.
There aren't many geographical barriers in the Indian subcontinent that would justify the existence of two vastly separated population for millenia. Admittedly there are some barriers, but none so impermiable as to justify a paleolithic/ mesolithic ASI/ANI separation. Like the mountain ranges of Central Asia that kept East and West Eurasians genetically distinct for millenia.
Rather we should expect a similar situation to the European paleolithic in India with a WHG/SHG/EHG like continuum across the Indian subcontinent but overall showing affinity to the ASI signature. The differences in current population structure date to only from the Neolithic - c. 6000 BC in South Asia with the arrival of Middle Eastern farmers.
This ANI/ASI terminology is outdated and I hope future papers from India, when they start coming, should avoid it.
@Rob
I saw not a hint of offence in your mistaken and deleted comment and I did not intend to be defensive. It was just an associative thought-train that came out in writing.
And thanks for the info. At least now I know who said it if I need to refer. :)
@Davidski
Yes, they have Siberian and they are not Shintashta. But not like Nganasan either. The ancient populations that can be suspected to have anything with them (Mezhovskaya mainly) came out much less Siberian than modern Uralics from Siberia and (based on your PCA data) lack a certain ancestry those Nganasan/Nenec markedly have. The (proto-) Uralic homeland was not in the Northern Tajga or the Tundra, more likely in the South-Siberia- Altai region (the latter seems to be the source area of Seima-Turbino by the way).
@ Slumbery
The problem with that South Siberian Seima Turbino theory is that we already have quite a lot of ancient yDNA from Altai and South Siberia and not one N1c has been detected. Pazyryk has yielded N1b and Iron Age Afontova Gora N(xN1a,N1b, N1c) but most of the ancient samples in Altai are Q, R1a or J. Three Okunevo samples in Khakassia were N1c but they were specifically the Khakass line N-B187 which is non-existent in Uralics. Baikal Neolithic will probably also yield yDNA N, but it was defined K(xN1c) in the paper that was published. Moreover, the Finnish and Saami Siberian is not from Late Bronze Age Altai. It is of different type.
Linguistic arguments are presented here:
http://www.elisanet.fi/alkupera/UralicEvidence.pdf
1. Indo-European - Uralic contacts began during the 3rd millennium BC.
2. Late Proto-Uralic was contemporaneous with Late Proto-Aryan = ca. 2000 BC.
3. Uralic expansion to the west seems to have reached Upper Volga around 1500 BC; then and there Mordvinic, Finnic and Saamic branches dispersed.
4. Saamic to Finland and Finnic to Estonia probably didn't arrive until the 1st millennium BC.
Any genetic model concerning languages should take this arguments into consideration.
Could the later Uralic ethnogenesis be related to Huns mixing with Goths?
@Johan Grip
Could the later Uralic ethnogenesis be related to Huns mixing with Goths?
No.
But N1c is East Asian in origin? Any other route possible than the same that Huns and Mongols came? Just earlier like Seima-Turbino?
@Johan Grip
Well, do you think that the ancient 3,000 year-old sample from this Lamnidis et al. preprint with East Asian-related ancestry and N1c was related to Huns?
Well those are old Hungarian claims that there is a connection.
But it does look like N1c was brought in by some primitive hunter gatherer group following the Arctic coast.
It could be that Indo-Europeans introduced them to agriculture and technology so that in exchange they could provide them with furs or other trade goods.
N1c might have ended in the Baltic as slaves by the Corded Ware people and eventually assimilated.
I'm not sure what the issue is here. Siberian is in all Northern Europeans, and also Hungarians. Latvians can be modeled at at least 10% Saami, but they are Baltic speakers anyway. There are no Uralic speakers without Siberian ancestry.
@Johan Grip
Must have been some pretty popular slaves in order to keep their language and convince their CW masters to switch theirs. They also eventually became the dominant haplogroup of the region, with an N1c carrying ruler.
Post a Comment