search this blog

Wednesday, November 22, 2017

Ancient genomes from NE Europe suggest in tandem spread of Siberian admixture and Uralic languages into the region >3,500 ya

Max Planck's Thiseas Christos Lamnidis recently tweeted this image of a part of a poster that he's presenting on the population history of Northeastern (NE) Europe at the Human Evolution 2017 conference in Cambridge, UK (for the tweet see here):

If you can't make out the text in the image, this is what the introduction says:

European history has been shaped by migrations, and subsequent admixture. Evidence points to migrations linked to the advent of agriculture, and the spread of Indo-European languages [a b]. Little is known about the ancient population history of NE Europeans, specifically Uralic speakers. Here we analyse eleven ancient genomes from Finland and NW Russia and a high-coverage modern Saami genome, and show that northern Europe was shaped by gene flow from Siberia that began at least 3,500 ya. Today, this ancestry is found in modern populations of the region, especially Uralic speakers. Additionally, we show that ancestors of the Saami inhabited a larger territory in Finland during the Iron Age than today.

It's intriguing to me that Max Planck is looking so closely at these issues now, because back in 2015 I ripped into Max Planck's Paul Heggarty for some comments that he made about the potential link between Yamnaya-related admixture and Uralic languages (see here). This is what I said back then:

These are exceedingly naive and stupid comments from someone representing the Max Planck Institute. Perhaps as an ardent supporter of the Anatolian hypothesis he's feeling more than a little desperate at this point and clutching at straws? That's because anyone with even a basic grasp of European linguistics and genetics should know that:

- present-day Hungarians and Estonians speak Uralic languages, but they are of course overwhelmingly of Indo-European origin, which is easily seen in their genome-wide and uniparental DNA

- other Uralic speakers, further to the north and east, in the forest zone away from Indo-European influence, are clearly distinct from the vast majority of Indo-European speaking Europeans, because they show significant levels of recent Siberian ancestry, which was missing among the Yamnaya and Corded Ware people, and appears to be an Uralic-specific genetic signature

- therefore, it's highly unlikely that Uralic-speakers were also part of the Yamnaya > Corded Ware movement; rather, early Uralics in all likelihood began to move west across the forest zone well after the Yamnaya and related expansions from the steppe.

All of this is probably just a remarkable coincidence, but in any case, it's nice to see that the good people at Max Planck are now beginning to understand the processes that have shaped the genetics and linguistics of NE Europe.

See also...

Uralic genes

Genetic and linguistic structure across space and time in Northern Europe

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


«Oldest   ‹Older   201 – 263 of 263
Davidski said...


Are you able to use the Gedmatch Lazarus kits at all for your PCAs or anything like that? Anything I should beware of with them?

Don't know what that is. I've lost track with what's happening at GEDmatch.

Ryan said...

@David - Gedmatch lets you construct artificial kits using someone's relatives. Did that myself for a close relative but when I plug it into your Eurogenes calculators I get some screwy results (that I can't for sure rule out). All the fits are about the same quality too (like 3.7 to 3.9). Is 8% Ashkenazi on your JTest higher than noise?

Davidski said...


Sorry, can't help you with that, because I'm not familiar with the methods.

Obviously, 8% Ashkenazi can be a real result. But if there's a problem with the methods, then it might be false.

Chad Rohlfsen said...

If that person isn't getting a chunk of Ashkenazi at any other place, like 23andme or Ancestry, and not showing a load of Jewish matches, then it's not real.

Chad Rohlfsen said...

You can check with David, but I don't think it even was true Ashkenazi cluster, but one that peaked at 23% in Askenazi. Everyone got it to some level and Greek were close to Ashkenazi, IIRC.

Ryan said...

@Chad - I can't check them anywhere else, as it's a kit that was generated artificially by Gedmatch. Good point on the cluster peaking in the 20s though. And yah, the question is how reliable are the methods of creating the kit.

There are a few Jewish matches. Not that many, but if the signal was real and ~4 generations back I guess you'd only expect 1/16 matches to be Jewish.

The trouble is I know very little about this relative. I know he is R1b-M222 and probably at least half Irish but that's about it.

Davidski said...

Anyone with an Ashkenazi ancestor four generations ago will practically get IBD hits with every Ashkenazi in any database.

Ryan said...

@David - Gotcha.

I should probably just come out and say it - this is my gamete donor I'm talking about. I've created a fake kit for him using my DNA, my half sister and a few of our shared matches to get it up to the minimum 1,500 cM. Not sure why it wouldn't run for just my half sister and I - we share more than 1,500 cM of DNA. I'm worried my choice of shared matches may skew it.

From my own DNA I can infer the donor was probably part Irish and part Scandinavian of some sort, but there rest is hard to tell. I come back in the calculators (ie Family Tree, Ancestry and EurogenesK13) as about ~6% West Asian, Middle Eastern or South Asian. I can explain half of that as being from my grandfather who seems to be 1/8 Chechen or Tatar or something, but not the rest. Such a small slice it may be noise but boy does it make it hard to figure out the other ~94% of the donor since Oracle in Gedmatch latches right on to it.

Al Bundy said...

That's what I meant.Since a lot of them were in what is now Germany at that time, modern Germans must have some ancestry from those Celts unless they left lock, stock,and barrel.I guess I actually had too much Guinness.I tried, atleast give me that.

Chad Rohlfsen said...


You're not just trying to limit it to sections where you match? Does your mother have a sibling or are her parents alive?

Chad Rohlfsen said...

Also, I second what David said. Real Ashkenazi ancestry will overwhelm you with Jewish "cousins". They're never as closely related as predicted, of course.

Ryan said...

@Chad - this Lazarus kit feature on Gedmatch does just that - limits it to the sections of DNA where my sister and I overlap. That's where I'm getting this less than specific Irish + something anywhere from Norway to Romania result though. I have my mom's DNA too yah.

Chad Rohlfsen said...

Okay. If you want some help using other methods, send me a message at

Davidski said...

I don't think qpAdm or formal stats will be helpful for Ryan. He needs those files in a decent IBD/haplotype run. Modeling with nMonte with modern pops with something like Global 10 data might also be useful, because he's focusing on recent ancestry and drift.

Al Bundy said...

Ok the hangover wore off I was thinking insular Celts because you mentioned Irish,my bad.

Ebizur said...

I recall having noticed that discrepancy regarding the haplogroup assignment of the ancient specimens from Miaozigou when that study was published a little over four years ago. In any case, archaeologists seem to be in agreement that the relics excavated at the Miaozigou site indicate connections both with the widespread Yangshao culture (Neolithic of the interior of North China along the Yellow River) and with the Hongshan culture (Neolithic of southwestern Manchuria and southeastern Inner Mongolia), so even if Miaozigou were demonstrated to be the origin of haplogroup N-M128, the ethnolinguistic identity of that progenitor would remain somewhat unclear.

Quoting one of my previous comments,
"N-P43 and N-F1154 share a MRCA in N-L666 9,000 [95% CI 7,700 <-> 10,200] ybp according to YFull YTree v5.08. N-F1154 subsumes N-M128, one of the first subclades of Y-DNA haplogroup N to be defined. Underhill et al. (2000) discovered N-M128 in a sample of Japanese (1/23) and in a sample from Central Asia & Siberia (1/184). Hammer et al. (2006) found it in a sample from Korea (1/75), a sample of Northern Han (1/44), and a sample of Manchu (1/52). Xue et al. (2006) found it in Manchu (2/35), Xibe (1/41), and Buyi (2/35). The Y-DNA of one individual from Beijing, China, one individual from Guangdong, China, one individual from Japan, and two individuals from Vietnam has been tabulated under N-M128 on the YFull YTree v5.08. As you probably already know, N-P43 is another subclade of Y-DNA haplogroup N (besides N-L708) that has a significant presence among present-day speakers of Uralic languages."

According to Supplementary Table 1 of Rootsi et al. (2007), N-M128 also has been found in 15/185 = 8.1% of Kazakhs (pooled from 3 locations), 1/54 = 1.9% of the Uzbek sample of Karafet et al. (2002), 1/94 = 1.1% of a sample of Komis, and 181 = 0.6% of a sample of Khakas. According to Kang Hu et al. (2015), N-M128 has been found in 1/219 North Han, 4/499 East Han, 2/173 South Han, 3/1126 Gansu Han, 12/874 Shandong Han, 2/34 Han from Wuchang, Heilongjiang, 1/286 Hui, 1/506 Amdo Tibetans, and 1/264 Gannan (South Gansu) Tibetans. Kang Hu et al. also provide the only data regarding N-F1154*(xM128) of which I am aware besides the single example on YFull. They have found N-F2759(xM128) in 1/84 Northeast Han, 2/219 North Han, 11/1126 Gansu Han, 5/874 Shandong Han, 1/23 Bargud (Barga Mongols) from Hulun Buir, 1/264 Gannan Tibetans, and 1/149 Qiang. They have found N-F1154(xM128) without determining the status for F2759 in 3/499 East Han. Furthermore, they claim to have found one example of N-F1154(xF2759) in a sample of 62 Ööled Mongols from Hulun Buir. N-M128 and N-F1154*(xM128) have been estimated by YFull to share a MRCA approximately 4,800 [95% CI 3,800 <-> 5,900] ybp, which is not significantly earlier than the TMRCA of extant N-M128 in China, Japan, and Vietnam.

Ebizur said...

N-P43, another extant subclade of N-L666, has been found in a few people in Northeast Siberia, such as 4/31 = 12.9% of a sample of Eskimos examined by Kharkov in his PhD thesis (2005) and in an Aleut originally from Medny Island (now uninhabited; Medny Island had recently been colonized by people from Attu Island, which also is now uninhabited). The Medny Island Aleut individual (whose DNA sample also has been labeled as "S_Tlingit-1" in the spreadsheet at; I have assumed that this labeling is erroneous) is id:ERR1347715 on the YFull tree, and his N-B226 Y-DNA has been estimated to share a MRCA with an individual from Kirov Oblast 1,650 [95% CI 1,050 <-> 2,400] ybp. N-B226 is estimated to share a MRCA with the N-Y24382*(xB226) Y-DNA of a Khant approximately 3,200 [95% CI 2,300 <-> 4,100] ybp and with the slightly more genetically distant N-Y23785*(xY24382) Y-DNA of an individual from Sverdlovsk Oblast not many generations prior to that. It looks like this Aleut member of N-P43 is descended in the paternal line from someone who inhabited the vicinity of the Ural Mountains prior to being assimilated by the expanding Russian Empire and eventually producing a child with an Aleut woman.

Rob said...

@ Kristiina
Thanks !

Tesmos said...


''There's no mystery here - a small number of IE speakers assimilated Europeans in Eastern and Western Europe. Elite dominance is a thing.''

I wouldn't call 40-54% Yamnaya-related ancestry in Eastern and Northern Europeans ''elite dominance''. There was a large replacement in Northern & Eastern Europe. We do not have even our hands on samples from Western Yamnaya yet.

Chad Rohlfsen said...

I think IBD and formal stats can compliment each other with Ryan. I can do the formal side, if you want to do the IBD side.

Chad Rohlfsen said...

qpGraph.. f3, f4ratio, Dstats, the lot.

Kristiina said...

@Tesmos "I wouldn't call 40-54% Yamnaya-related ancestry in Eastern and Northern Europeans ''elite dominance''. There was a large replacement in Northern & Eastern Europe. We do not have even our hands on samples from Western Yamnaya yet."

You should read that comment in tandem with the previous comment as it was irony from my part. However, I presume that the Finnish high Yamnaya procentage is Uralic-related.

Sofia Aurora said...


Please allow me to disagree.

Romanova and Dyakonov in 2016 wrote a paper named "the Stepe Saga of Horsemen in the Arctic: From prehistoric times to now" and found the influence of the kurgan groups to the Arctic including the river Ob to the river Lena.
They found a superstratum of pontic steppe cultural and archaeological presence stretching through the entire arctic of Northern Asia.
Sakha had played an important role in the spreading of that phenomenon from Yakutia to Karelia.
They do not relate these peoples with the Yukaghir or Ket or Kerek or Kamassian exclusively.
Probably these people were bilingual and spoke both their native tongues and an Indo-European language serving as serfs and allies to various Indo-European clans that pushes so north in Arctic Asia

Tesmos said...



Sofia Aurora said...

Come on lads!

We are supposed to have a serious discussion and you bring up..."Dienekes"!!
The synonym of propaganda, arrogance and ignorance?
Let him rot as far as i am concerned!
He has been deliberately lying and misguiding people through his blog for 11 bloody years!!
He was opposing everything that went against his imaginary pioneer Mediterranean midgets and the Lux Ex Orientes that they brought to the World (including his country Greece)!
Yeah! i agree with Shahanshah that Pontikos has AND NOT HAD issues with everything that prove his dogma wrong.
It's nice that he has shut his mouth finally now and for the last 4 years so we won't have to read his nonsense anymore on any topic what so ever!

Huck Finn said...

@ Sofia Aurora and re "Sakha had played an important role in the spreading of that phenomenon from Yakutia to Karelia."

With all due respect to Sakha, have you found anything to prove this laughable claim of your?

(Степная сага коневодов Арктики: от древних времен до недавних событий is BTW available at, for those interested in Yakutia and Yakuts)

Kristiina said...

I think that this paper will show that Siberian in Finns and Saamis is native to Finland. It probably did not arrive before the Comb Ceramic period, but, however, it preceded the modern languages spoken in the area (Finnish, Saami).

The typical Finnish N1c-VL29 could even be an autochthonous line in the area which originally spoke a paleo-language and only later on adopted the modern Uralic language.

Samuel Andrews said...


"However, I presume that the Finnish high Yamnaya procentage is Uralic-related."

I guess that's plausable if Uralics mixed with Yamnaya like groups. But how do you explain the large amount of Middle Neolithic farmer ancestry in Finns? Finns have both EEF and WHG. Finns can be modelled at least 30% Middle Neolithic European farmer. One way or another this camne from outside of the Steppe and from outside of Finland.

It confirms Finns have some sort of "recent" mainland European ancestry. It also suggests their Steppe ancestry came from Steppe+MNF admixed people from mainland Europe similar to Indo European speaking northern Europeans.

Kristiina said...


EEF could be from the Kiukainen Culture (2000–1500/1300 BC).

In that case, it could be related to Scandinavian TRB farmers.

Kristiina said...

The English text is very short. The Finnish version says about the Kiukainen Culture that :
- tools resembling the tools of the Comb Ceramic period were used as well as Corded Ware type axes
- ceramics was a mixture of Comb Cearmic and Corded Ware elements
- fishing was important but also farming as pollen from wheat, barley and oates has been found on the sites as well as millstones and sickles of Swedish type

Kristiina said...

It is also important to note that Kiukainen Culture was not a metal culture. They did not know how to make metal tools, and this is also a good reason to presume that they were not IE speaking and did not derive from the Scandinavian Bronze Age.

Huck Finn said...

@Kristiina and re: "The typical Finnish N1c-VL29 could even be an autochthonous line in the area which originally spoke a paleo-language and only later on adopted the modern Uralic language."

Question of definition, maybe, but if that's the case, how do you explain those Buryats, Mongols, Koryaks etc. in the ancestral form N-Y6058 and Komi in the preceding N-CTS10760?

The expansion area somewhere next to Ural mountains explain very nicely both western and eastern forms of paternal N, especially now that we know that Uralic expanded both to west and east from that place.

supernord said...

All hard followers of the Anatolian hypothesis are trying by all means to convince them that in Eastern Europe before the 2nd Millennium BC lived the Uralics.

Kristiina said...

The first N line in Finland need not be VL29. It can be an extinct line. There could have been several groups coming from the east with different N lines. The ancient yDNA has in any case shown us that most ancient y lines went extinct.

You are right that VL29 is not so very old. If the Komi and the Finnish line arose somewhere in Russian Karelia, it must have happened c. 3000 BC at the earliest. However, it could somehow be related to early use of bronze in the area between Karelia and Volga Kama.

In any case, VL29 + CTS10760 (upstream VL29) is not very frequent in Volga Ural and we also know that Finns are autosomally quite different from Volga Uralics as if we had our separate history.

I do not really argue that this is what happened, but my aim is rather to try to widen the perspective.

Rob said...

What about Botai culture being pre-proto-FU ?
(I know it sounds odd but fits geographical spread of FU tree)

Huck Finn said...

@Kristiina and re: "You are right that VL29 is not so very old. If the Komi and the Finnish line arose somewhere in Russian Karelia, it must have happened c. 3000 BC at the earliest."

Even 3000 BC is too old, I'm afraid. N-CTS10760 TMRCA is only 4100 yrs, the ancestor of the lineage may very well have been moving towards west somewhere in Upper Volga still in Bronze Age, with some Seyma Turbino type and related bronze gadgets.

Mittnik et al did not find any N in their BA samples. That is hardly a coincidence.

Kristiina said...

"N-CTS10760 TMRCA is only 4100 yrs,"

Yes, but it should mean that 2100 BC it was somewhere between Finland and the Komi Republic.

If it was moving in Karelia-Komi, it need not reach Latvia early. I do not think that the absence of yDNA N in Latvia before c. 300 BC means that it must be recent in Finland.

In any case, it is not necessary that all the Uralic N lines stayed put in Volga Kama and then some decided to go to Finland and reached their destination at the same time during the Iron Age.

Of course, we do not know when Y16323 reached Altai, but it has not yet been found there or nearby, so it may have been in a more westerly location still in the Iron Age.

Z35352 may have arrived to Chukotka as late as 1000 AD as it is at that time that reindeer hunters with knowledge of bronze arrive.

I do not insist and I hope to get soon some hard reference points.

Kristiina said...


But the latest news was that Botai yDNA was O2!

If N-M2126 (TMRCA 6400) will be found there, then you are right. Let's wait and see.

I am a bit sceptical, but never say never. :)

Kristiina said...

I think that due to its climate and sparse population, yDNA extinctions have been the norm in Finland. During the great famine in 1695–1697, one third of the population died. Y full is based on modern populations, i.e. in Finns in Finland, and Finns are a population that expanded with farming only recently. I am confident that the YDNA N tree will look different if we get samples from prehistorical populations in Karelia and Lapland.

supernord said...

"But the latest news was that Botai yDNA was O2!"

It is the duck, do not trust.

Palacista said...

Getting back to ergativity in proto IE. It is possible that some features of proto IE could be explained by an earlier ergative state even if it does seem to be a bit forced, but as I see it the main argument against is Hittite. Hitite does have an ergative case, however this is clearly a derived feature using an otherwise unknown suffix. This can be comfortably explained as an areal feature as Hittite was the most eastern of the Anatolian languages and almost surrounded by ergative languages. As Hittite split off from a very early proto IE why would it need to assimilate contact ergative features if this was already part of the language?

Ebizur said...

supernord wrote,

"It is the duck, do not trust."

I cannot attest to the reliability of the website on which the claim of finding O2 (what would now be called O1b-M268) has been published, but there are some findings from modern populations that would be consistent with an ancient presence of some sort of O1b-M268(xO1b1a1a-M95, O1b2-M176) in Central Asia:

Wells et al. (2001)
O-M175(xM119, M95, M122)
14/45 = 31.1% Koreans
1/16 = 6.3% Tajiks/Dushanbe, Tajikistan
2/41 = 4.9% Uyghurs
1/22 = 4.5% Crimean Tatars
1/40 = 2.5% Tajiks/Samarkand, Uzbekistan
1/54 = 1.9% Kazakhs
1/68 = 1.5% Uzbeks/Surkhandarya region, Uzbekistan
1/70 = 1.4% Uzbeks/Khorezm region, Uzbekistan
(The Koreans in this case probably belong to O-M176.)

Di Cristofaro et al. (2013)
O-M175(xM119, M95, M176, M122)
1/18 = 5.6% Iranians/Teheran, Iran
2/37 = 5.4% Tajiks/Badakhshan province, Afghanistan
1/97 = 1.0% Mongols/northwest Mongolia

1/20 = 5.0% Mongols/northeast Mongolia

Haber et al. (2012)
4/18 = 22.2% Tajik/Takhar province, Afghanistan (the extreme east of ancient Bactria around Ai Khanoum, and part of the northwestern slopes of the Hindu Kush that form its southern border; Badakhshan and the Pamirs are located to the east and northeast)
1/9 = 11.1% Tajik/Badakhshan province, Afghanistan
1/60 = 1.7% Hazara (father of the O-M175 individual is from Kabul, Afghanistan)
1/13 = 7.7% Balush (also known as Baloch; father of the O-M175 individual is from Kandahar, Afghanistan)
(These cases of O-M175 from Afghanistan do not necessarily belong to the same type of O-M175(xM119, M95, M122), but some of them might. I would say that the likelihood that some of them might belong to that unusual type of O-M175 is increased by their having been found mainly among Tajiks in the northeastern corner of the country rather than among the Pashtuns and Turkic peoples in Afghanistan. That would also be congruous with Di Cristofaro's finding of O-M175(xM119, M95, M176, M122) Y-DNA in 2/37 = 5.4% of another sample of Tajiks from Badakhshan province, Afghanistan.)

The Tajiks speak a dialect of Persian, and Di Cristofaro et al. also found an example of O-M175(xM119, M95, M176, M122) Y-DNA in their small sample of Iranians from Teheran, Iran. It seems to be a minor haplogroup found at present mainly among speakers of Persian.

Elsewhere, O-M175(xM119, M95, M176, M122) Y-DNA has been found mainly among Han Chinese in the form of several subclades of O1b1-K18(xO1b1a1a-M95), which in total comprise about 5% of the extant population of Han Chinese males.

The history of this haplogroup among Persian speakers (and among Central Asians in general) is still murky, but I suppose it might make that Russian report of O1b (former O2) in remains from the Botai culture somewhat more plausible.

Samuel Andrews said...


Thanks for sharing here. It's nice to see someone who is an expert in their area. I look at Finno-Urgic genetic origins from purely from a genetic perspective. I don't know anything about lingustics or archeaology. From a genetic perspective, I can't wrap my head around how Finns can be so similar to other northern Europeans yet not be of largely Indo European origin.

The only major Y DNA haplogroup other than N1c in Finns is I1. Plus, Finns have their own unique form of I1. This suggests pre-historic, pre-Norse and even pre-Germanic, ancestry from Scandinavia. I1, like R1a-M417 and R1b-L151, expanded sometime between 2600 and 2000 BC, and likely in a Steppe-MN mixed population similar to modern northern Europeans. So, could this mean Finns are a I1-European, N1c-Siberian mix?

In my opinon, Saami pretty obviously have more "ancient" Fennoscandian ancestry. Their huge amount of Siberian, their SHG admixture, suggests they trace a lot to people like the Iron age genome from Finland and the HGs from the Kola Peninsula. Finns trace more back to somewhere else in Europe, either in Scandinavia or Baltic states, that had heavy MN farmer and Steppe.

Rob said...

@ Sam

" Plus, Finns have their own unique form of I1""

Was gonna say I don;t see, but they do too.
Under I -L287

Ryan said...

@Al Bundy - No worries lol.

@David, Chad - Thans guys.

Rob said...

I'd be surprised if Botai has hg O, but not surprised if it has R1a, Q, some J even, N with enough samples.

Derek said...

Somebody on Anthrogenica posted an abstract of an upcoming paper on Lombard DNA in northern Italy:

The PCA may not be perfect, but here the Lombards look very North European, more like the Scandinavians than the Germans. The 6th century Hungarian skeletons, surprisingly, look to be divided into NW European and S European subgroups.The two lone Avars look Polish.

Matt said...

@Derek, yeah, I thought that was pretty cool, using the Novembre Popres PCA for that function, and really all the analysis of the site in their poster coming together like that to add an adna layer to conventional analysis; identified families buried together, high family at each Lombard cemetary site looks Scandinavian rather than Pannonian or N.Italian, warriors similar but based on PCA I would guess perhaps a slightly more diverse background, other samples tend to look more Roman / local, stronger grave goods status and ancestry covariance at the Pannonian site than Collegno.

That said, I'm not totally sure about their use of ADMIXTURE though, and may be better for them to have used qpAdm or a PCA based method to work out the ancestry proportions of the samples. Will have to read paper to try and understand that reasoning better.

I wonder if those two "Polish" Pannonian Avars would look as Polish in a world context, or pick up some fraction of Siberian ancestry though. Be pretty interesting if they did not...

Relative absence of genetically Hungarian type samples at Szolad may be due to Roman settlement or slavery from across Roman world or something?

Kristiina said...


If the Finnish I1 + EEF is from the Neolithic farmers from Sweden and N1c + Yamnaya/EHG/Siberian from the metallurgical centres of the Urals + Fatyanovo admixture + extra Siberian from the Finnish local population, we get more or less the genetic ancestry of Finns. We do not need Trzciniec and Scandinavian Bronze Age for the autosomal ancestry of Finns.

I think it is important to take into account that Finns were farmers from the south. The inland population before the Finnish expansion was probably specialized in fur trapping and, IMO, they surely did not speak Finnish.

Angantyr said...


Are there any archaeological signs or even hints of a migration of EEF/TRB farmers from Sweden to Finland? I have never heard of any. In CWC/Battle Axe/Boat Axe times there were likely contacts across the Baltic Sea, but if there was any major migration is uncertain, and going by the most-likely-CWC guy from Ölsund, Hälsingland, the only hint we have is that there might have been one from east to west.

Also, the earliest Swedish I1 we have is from the late neolithic, post-CWC/Battle Axe. From the mesolithic we have only I2 and a possible pre-I1. From TRB we have unfortunately almost no Y DNA, only a possible I2. From PWC we have only I2. And from Battle Axe we have 100% R1a.

Kristiina said...


I already proposed Kiukainen Culture (2000–1500/1300 BC).
- tools resembling the tools of the Comb Ceramic period were used as well as Corded Ware type axes
- ceramics with textile decorations of Swedish type
- fishing was important but also farming as pollen from wheat, barley and oates has been found on the sites as well as millstones and stone sickles of Swedish type

We have very little ancient I1. In the Scandinavian Bronze Age samples I1 was only found in Sweden and not in Denmark.

Angantyr said...


Yes, but why would we expect the people of a culture that appears to be a fusion of CWC and Comb Ceramic to be genetically like TRB/EEF people, when TRB had disappeared hundreds of years earlier, and the nearby CWC/Battle Axe (or PWC for that matter) genomes that we do have don't look much like the TRB people?

But indeed, we have far too little ancient Fennoscandian Y DNA to be able to determine when and where I1 spread to become as frequent as it is today.

Angantyr said...


But just to check, is your hypothesis that an unusually EEF-rich Battle Axe group dominated by Y HG I1 (and not R1a) migrated east from Sweden to Finland (where the earliest CWC people that had arrived from the south had failed to expand much) and mixed with the Comb Ceramics?

supernord said...

Kristiina said...
"If the Finnish I1 + EEF is from the Neolithic farmers from Sweden"

Impossible,TMRCA I1 only 4700 ybp, it is Bronze Age, but is not Neolithic! In Finns there are latest subclades.

"We do not need Trzciniec and Scandinavian Bronze Age for the autosomal ancestry of Finns."

But this is necessary. Ugro-Finnic is N1c + Siberian components. You will not can to associate Pra-Finns with Proto-Finnic language with anyone else.

Rob said...

I1, or rather pre-I1, could relate to an early dispersal of Magdalenians or Hamburgians, before becoming attenuated during one of several extinction/ replacement cycles in Late Paleo Northern Europe (eg. by later I2a1 Azailian- Fedemessergruppen folk). As such it could have harboured in the Scandinavian post Fosna / Komsa groups, or in a Carpathian nook.

Present I1 is a Bronze Age expansion from either post-LBK groups in east Europe or the local Swedish Pitted Ware which survived and expanded at the expense of TRB in east Sweden, then integrated into BAx. One group then expanded to Finland

Kristiina said...

@ Agantyr

"But just to check, is your hypothesis that an unusually EEF-rich Battle Axe group dominated by Y HG I1 (and not R1a) migrated east from Sweden to Finland (where the earliest CWC people that had arrived from the south had failed to expand much) and mixed with the Comb Ceramics?"

Yes, I think that there was a migration of EEF+WHG rich I1 bearing group to Finland, but I suggested that it was related to Kiukainen Culture which seems to have come from Sweden and was not a metal culture. Baltic Finnics are not direct descendants of the Finnish Comb Ceramic because they have a lot of Yamnaya admixture and I think that Baltic Finnics acquired it somewhere between Northwestern Russia and Volga.

I see that Mr Myllylä's new analysis is the following:

Khanty_Mansi 0.00669230541442569
Saami 0.0318001424720861
Scandinavian 0.0406288973530398
Eastern_Baltic_Finnic 0.372195068297064
South_Baltic_Finnic 0.547727866737746

West_European 0,6080567694
Baltic 0,203714483
North_Baltic_Finnic_G3 0,092400554
North_Baltic_Finnic_G1 0,0431817267
East_Slavic 0,0387924755
Saami 0,0096464481
North_Siberian 0,0042075434


Kristiina said...

This map that I once made might be interesting:

In this map you can see the area of Western Uralic hydronyms and Baltic hydronyms and the ancient N1c that was found within the Zhizhitskaya culture.

Simon_W said...

@ Matt

"Relative absence of genetically Hungarian type samples at Szolad may be due to Roman settlement or slavery from across Roman world or something?"

I guess it means the Hungarian type genepool formed later, with the addition of Slavs, a Magyar elite and later German admixture.

The fact that even in Szolad there is a person that goes towards Cypriots is interesting. Must be either from Roman colonists of south Italian provenance or (more likely) be evidence for the circum-Mediterranean character of the late imperial Roman empire that induced migrantion from the east Mediterranean area, analogous to the Bedouin-like outlier from Roman Age Yorkshire.

I would say the rather north European character of the ancient Bavarians (Baiuvars probably) is also fascinating. It seems to imply that modern Bavarians are a mix of very differentiated elements: rather north European Germanics and very dissimilar, more southern provincial Gallo-Romans.

The place of the only "Roman" in the PCA is also interesting, closer to Iberians than to the Italian average. Of course in his case it would be essential to know what time and place he (or she) is from. If that Roman is a local from Collegno, his position in the PCA wouldn't be earth-shattering.

Simon_W said...

The Roman sample is probably from late antique/early Medieval Piedmont as well:

Quotation: "For this project we have selected two cemeteries that appear characteristically Longobard, Szólád in Hungary and Collegno in Italy. These have the advantage of having been recently and carefully excavated, we have detailed information on each tomb, and we have been able to obtain stable isotopic data from both sites that provide additional information on the shortterm population demography and lifestyle. Currently we are in the process of extracting and sequencing ancient DNA from twenty samples from each site in the laboratory of the Department of Evolutionary Biology at the University of Florence and in the Institute of Archaeological Science in Tübingen. We will also sequence samples from near-by closely contemporary sites that have different cultural characteristics and are thus considered non-Longobard."

So, probably not a 100% Latin Roman from the early republic. ;-)

Matt said...

@ Simon_W: "I guess it means the Hungarian type genepool formed later, with the addition of Slavs, a Magyar elite and later German admixture."

Yeah, it could totally be that as well, I am just tentative if this cemetery can provide good evidence for that if is atypical and is in a region strongly colonized by Langobards / incomers from the Roman world to Roman Pannonia.

The other phenomena you discuss are interesting as well; good points. I agree I'd infer seem like slightly more change in North:South in Bavaria over time than the comparable Iron Age Brits to moderns show, perhaps from the absorption of those groups you mention. As well, they do seem to suggest that the Roman world had a pretty wide genetic span (among those that actually participated in its urban civilization...?).

It would be useful to repeat all of these with simple qpAdm / West Eurasia PCA results as well - projecting onto Popres PCA is pretty cool for capturing subtle drifts between recent Europeans as we enter the stage of history, but still useful to have a sanity check on deep ancestry as well from methods that are more sensitive to that.

Simon_W said...

There's hardly a place in Europe that saw more invasions and migrating peoples than Hungary, thanks to its central position, its natural lack of woods and the notoriously flat terrain. So the expectation to see a strong genetic continuity down to the deep past is not very realistic. Hungarians in an ethnic sense are first seen with the Magyars who arrived there in 896 AD. Since they were Finno-Ugrians from far eastern Europe while modern Hungarians are no genetical outliers to east central Europe and quite distinct from other Finno-Ugrians, it can be suspected that there was a lot of genetic continuity to the immediately preceding populations, which were Slavic and Avaric. I've also read several times in the comments here that the Hungarians later acquired substantial Austrian and Czech (?) admixture, but I'm not familiar with this part of history. So if somewhere in the deep prehistory a population in Hungary looked Hungarian-like, this is in all likelihood coincidental.

Using nMonte2, I've checked the distances of several ancient samples in the Global 10 PCA to a wide array of modern European samples: To Basque_French, Belarusian, Czech, Danish, Dutch, English_Cornwall, English_Kent, Estonian, Finnish, French_East, French_South, German, Hungarian, Icelandic, Irish, Latvian, Lithuanian, Norwegian, Orcadian, Polish, Scottish, Slovakian, Slovenian, and Swedish.

Now we've got a sample from late Bronze Age Hungary, BR2, which belonged to a local version of the pan-central European Urnfield culture. In the Global 10 PCA he's closest to the French_East, followed to English_Kent. Then we've got the early Iron Age sample IR1 which belongs to a culture colloquially called Thraco-Cimmerian, which thus has more eastern affinities. He's closest to Slovenians, followed by Hungarians. We haven't seen any more samples from LBA or Iron Age Hungary, but it can be said that in the later Iron Age Transdanubia saw a heavy influence from the Celtic La Tène culture. And this was also the part of Hungary that was conquered by the Romans. While to the east of the Danube there were the Iazyges, a tribe of the Sarmatians. So tentatively we can expect some sort of continuation of more western BR2-like affinity in Transdanubia and eastern IR1-like affinity in the east. And then Huns came and went, followed by the Ostrogoths, who in turn were followed by the Longobards. And then the Avars and Slavs.

And as for the North:South changes in southern Germany, I've also checked the distances of the sample Germany_Bronze_Age:RISE471, which is from MBA southern Germany, Tumulus culture apparently. Among the specified modern references he's closest to French_South, followed by Basque_French! While Halberstadt_LBA from LBA northern Germany is closest to the Danes, followed by the Swedes. That's what I call a stark contrast!

Davidski said...

Since they were Finno-Ugrians from far eastern Europe while modern Hungarians are no genetical outliers to east central Europe and quite distinct from other Finno-Ugrians, it can be suspected that there was a lot of genetic continuity to the immediately preceding populations, which were Slavic and Avaric.

Tatar raids largely depopulated Hungary, so no.

Simon_W said...

@ Davidski

OK, I'll buy that. ;-)

@ Matt

The apparently strong diffusion of southeast European-related admixture into Roman Age Hungary looks indeed remarkable. You can easily imagine how this must have been at least similar in Italy. The Roman Empire was in some ways like the EU, with free movement of people and goods, but with even more uniformity in language, as Latin was more widely used than English is in the EU.

«Oldest ‹Older   201 – 263 of 263   Newer› Newest»