Genetic continuity among Uralic-speaking cultures in north-eastern Europe


The recent study of Estonian Late Bronze Age/Iron Age samples has shown, as expected, large genetic continuity of Corded Ware populations in the East Baltic area, where West Uralic is known to have been spoken since at least the Early Bronze Age.

The most interesting news was that, unexpectedly for many, the impact of “Siberian ancestry” (whatever that actually means) was small, slow, and gradual, with slight increases found up to the Middle Ages, compatible with multiple contact events in north-eastern Europe. Haplogroup N became prevalent among Finnic populations only through late bottlenecks, as research of modern populations have long suggested, and as ancient DNA research hinted since at least 2015.

I risked to correlate the arrival of chiefs from the south-west with the infiltration of N1c-VL29 subclades during the transition to the Iron Age, coupled with that minimal “Siberian” ancestry (see e.g. here and here). Now we know that the penetration of this non-CW ancestry started, as predicted, in the Iron Age; that it was highly variable in the few samples where it appeared, with ca. 1-4%, while most Iron Age individuals show 0%; and that it was not especially linked to individuals of N1c-Vl29 lineages.

It is also basically confirmed, based on the (ancient and Modern Swedish) N1c-L550 subclades found among Iron Age Estonians, that N1c-VL29 lineages and the so-called “Siberian” ancestry will be found simultaneously around the Baltic coastal areas, and that different lineages must have suffered later founder effects among Finns, which suggests that these alliances through exogamy brought exactly as much language change in Sweden, Lithuania, or Poland, as they did in the East Baltic region…

On the other hand, the paper has also shown a potential movement of Corded Ware-derived peoples, if the change from LBA to IA samples is meaningful; in fact, even more Corded Ware-like than Baltic and Estonian BA populations. The exact origin of that movement is difficult to pinpoint, and it may not be related to the arrival of Akozino warrior-traders from the south-east, since theirs seems to be a minor impact proper of elites in a chiefdom system around the Baltic.

Distribution of fortified settlements (filled circles) and other hilltop sites (empty circles) of the Late Bronze Age and Pre-Roman Iron Ages in the East Baltic region. Tentative area of most intensive contacts between Baltic and Balto-Finnic communities marked with a dashed line. Image modified from (Lang 2016).

Also suggesting a potential movement is the ‘southern’ shift observed in the West and East Baltic areas, likely showing the arrival of Proto-East Baltic speakers (such as the Trzciniec outlier), as we have already discussed in this blog. The unexpected increase in Corded Ware-like ancestry in the Eastern Baltic, coupled with the expected large continuity of hg. R1a-Z283 in the homeland of Balto-Finnic expansions, gives even more support to the known complex system of exogamy along the Baltic coasts, and offers another potential reason for the rise of Baltic-speaking territories in the West Baltic: elite domination.

It is nevertheless important to understand that, even among the most “genetic continuous” regions like Estonia, not a single population in Europe is heir of some ancestral, immutable people. Not in terms of haplogroups, and not in terms of admixture. Balto-Finnic speakers, however continuous they might seem (e.g. in Southern Estonians) aren’t an exception.

After all, this blog was (re)born to fight the currently prevalent sheer stupidity surrounding the simplistic “R1a/steppe ancestry=Indo-European” association, so I wouldn’t like to see it replaced with some other stupid continuity or purity ideas within 10 to 20 years…

Late Uralic stems from East Corded Ware groups

With the currently available tools – linguistics, archaeology, and now genetics -, I don’t think there is any argument to date to question the direct connection of the Late Proto-Uralic expansion with all Eastern Corded Ware groups (i.e. Battle Axe, Fatyanovo-Balanovo, and Abashevo), and thus at least with the unifying A-horizon of Corded Ware and the bottlenecks under R1a-Z645.

NOTE. The only out-group among Corded Ware cultures is the Single Grave culture. It appears to be an early Corded Ware offshoot, reflected in their non-unitary cultural traits (distinct from later unifying waves), in their varied patrilineal clans, and in the short-lasting cultural effect in northern Europe before their complete demise under pressure of expanding Yamna/Bell Beaker peoples from the Danube. The culture’s minimal (if any) effects on succeeding peoples might be seen mostly in the (mainly phonetic) Uralic substrate found in Balto-Slavic – although this may also stem from a more eastern influence, close to the Baltic – and in the contacts of Celtic with Uralic. The huge time depth between this early hypothetic Uralic layer in northern Europe and the emergence of peoples inhabiting these territories in recorded history have no doubt been erroneously interpreted as a lack of Uralic presence in the area.

1) That connection was evident in the Yamna – CWC differences in archaeology, and especially later, with at least Fatyanovo-Balanovo and Abashevo representing the obvious replacement of the Volosovo culture before further expansions of CWC-related groups west and east of the Urals.

The mythical millennia-long continuity of Volosovo hunter-gatherers, including centuries among Corded Ware peoples, as expected lately by the Copenhagen group (and anyone who doesn’t want to question the 1960s association of Indo-European with CWC) must be rejected today in population genomics, as the recent studies of ancient and modern populations show, and as ancient DNA from the region will confirm.

2) In linguistics, the survival of Volosovo as The Uralic-speaking culture was also hardly believable. From Kallio (2015):

While we can say at least something about Uralic substrates in Northeastern Europe, non-Uralic substrates cannot at all easily be identified, because of multiple language shifts, viz. first from non-Uralic to Uralic and then from Uralic to Russian. Yet the Soviet Uralicist Boris Serebrennikov (1956, 1959) argued that there are some non-Uralic substrate toponyms in the Volga-Oka region, but his idea was never taken seriously in the west (cf. Sauvageot 1958), and it pretty soon also sank into oblivion in Russia, even though it can still occasionally pop up there in non-onomastic circles (cf. Napolskikh 1995: 18–19). However, not all the hypotheses on non-Uralic substrates in Northeastern Europe should be rejected (see e.g. Helimski 2001b).

Tentative map of the distribution of known languages in Eastern Europe during the Early Bronze Age. See full map.

Helimski (2001) argues for a non-Uralic topo-hydronomy in Northern Russia, whose population may have kept their languages up to the Common Era despite the Corded Ware expansion, which is in line with the survival of some non-Indo-European languages everywhere in Europe after the expansion of Yamna and its offshoots:

It should be borne in mind that these [Uralic] hydronyms reached us mainly through Northern Russian and, accordingly, with a tendency to phonetic-morphological adaptation and unification (for river names it is “natural” to be, like the word ‘river’ itself, feminine and to end in -a). Taking into account this circumstance, it may turn out to be non-useless for etymological identification of at least some of the hydronyms on the Finno-Ugric basis.

On the other hand, I wouldn’t exclude the possibility that some parts of this large geographical area were never (completely) Finno-Ugric. The population that created the most important part of the hydronymy of the Russian North could be finally pushed aside or assimilated only at the end of the 1st – beginning of the 2nd millennium AD, during the Russian colonization, retaining the memory of the White-Eyed Chude in its own memory.

NOTE. For more on this non-IE substrate in (especially West) Uralic, see e.g. Zhivlov (2015),

The same non-Uralic substrate is most likely behind most of the shared traits by Mordvinic and Balto-Finnic (see below).

3) In genetics, I don’t think the picture could get any clearer. I don’t know what “Steppe ancestry = Indo-European” proponents expected from 2019, if they expected anything at all (I haven’t seen any coherent model, proposal, or prediction for a long time now), but I doubt the recent results are compatible with any of their implied expectations.

Detail of the PCA of the Corded Ware expansion. See full PCA and more related files.

Notice, from the PCA above, how this Baltic Late Neolithic group shows actually a shift from Sredni Stog (see PCA with Sredni Stog) towards typical Khvalynsk-Urals-related ancestry, i.e. populations from eastern European forested regions, derived from hunter-gatherer pottery groups, as I have proposed for a very long time, since the first time a Baltic LN “outlier” appeared. It’s amazing how some amateurs can find 0.1% of any Siberian outlier’s ancestry among Uralians 4,000 years later, but fail to see the direct connection here. The esoteric uses of qpAdm, I guess…

Especially noticeable is the extra WHG-like ancestry and corresponding shift, seen especially marked in late Polish CWC samples, but also in Baltic CWC and especially in one Sweden Battle Axe sample, all of them shifting apparently closer to Pitted Ware and SHG. While that may have been interpreted as an in situ admixture in Scandinavia before, the late Polish CWC samples show likely a resurgence of local populations, so we can assume that both shifts (to SHG- and EHG-like populations) of available CWC samples around the Baltic are clearly part of the WHG:EHG continuum that will be found in the eastern European sub-Neolithic cultures, from Narva to Volosovo.

This WHG-related ancestry is clearly predominant in groups with which Battle Axe peoples admixed, based on the shift towards Pitted Ware, which – I can only guess based on modern Volga Finns – is different from the shift we will see in Netted Ware, more towards the Khvalynsk-Urals cluster. This is in line with the expansion of Battle Axe eastward through coastal areas (West to East Baltic and Finland into Sweden), while Fatyanovo peoples probably emerged from a slightly different route, but also a northern one, if one is to follow archaological similarities and their chronology.

Detail of the PCA of European Bronze Age populations. See full PCA and more related files.

During the Iron Age, the only peoples that probably shifted strongly (based on modern populations) are West Baltic ones, getting closer to the available Late Trzciniec samples, and even closer to the Trzciniec outlier, i.e. away from the earlier Eastern Corded Ware cluster, and towards Central European groups like Czech EBA or Poland EBA, both of them clearly derived from Bell Beakers, but also admixed with (and thus shifted toward) CW-like populations.

If one looks carefully at the previous PCA on Bronze Age populations, and the next one on Iron Age clusters, it is evident that adding the Swedish LN outlier to East Baltic BA (both strongly related to Battle Axe populations) essentially gives us the continuity of East Baltic BA into the Iron Age. This cluster is continued also in two outliers from Sigtuna, a Viking town close to the Gulf of Finland, known to be an important trading site, 1,500 years later. Not much of a change around the Gulf of Finland, then:

Detail of the PCA of East and North European Iron Age populations. See full PCA and more related files.

Based on the two simplistic Uralic clines one might see described (among the many that certainly existed, from Corded Ware to different Eurasian populations), and just like BOO was for some months fashionable as “Samic”, some may be tempted to say that certain Sintashta or Srubna outliers close to the Urals mark the True Uralic™ peoples. Because, of course they do. Ghost haplogroup N and stuff. And Corded Ware never ever Uralic. Because Gimbutas, and my IE R1a grandfather.

NOTE. Funny thing here: there might be Corded Ware, Iranian, Slavic, Germanic, etc… outliers or out-groups, and they might form the widest genetic clusters ever seen, but they are all of one language, because archaeology and linguistics; however, one “outlier” (also, put your own definition of “outlier” here, let’s say 1% of whatever, and strontium isotope potentially from 100 km away) ca. 600 BC in the Baltic who (surprise!) happens to show hg. N, and he signals the first incoming True Uralic™ speaker from wherever… It won’t be the first or the last time some people resort to “the complexity of Uralic-speaking peoples” in ancestry, just to look for “hg. N = Uralic” like crazy. You only need common sense to understand that this is not how this works. Amateur genomics can’t get more embarrassing than the current “let’s look for ‘Siberian ancestry’ in every individual of haplogroup N” trend. Or maybe it can, and it will, but I can’t see it yet.

If one were to insist on looking for ‘foreign’ contributions among Iron Age Estonians, though, I think one should also check out first archaeology, and then the PC3 (or, more graphically, a 3D plot), to understand what might be happening with the many Uralic clines derived from Corded Ware, before starting to play around with bioinformatic tools to discover a teeny tiny 1% admixture of the wrong population, and rushing to build far-fetched narratives. Apparently, one of the different clines formed roughly between southern (steppe – forest-steppe) and northern (tundra-taiga) populations in Uralians is also seen in some Iron Age Estonian individuals – especially in some late samples from Ingria…This is not my main interest, so I will leave this here for others to keep wasting their time chasing the white whale of the 0.5% of True Uralic™ ancestry in ancient Baltic samples of hg. N.

Still images of the 3D plot of Eurasian samples. Typical PC1 vs. PC2 visualization to the left, and shift of the view to PC3 on the right image. See full PCA and more related files.

An exclusive Volga-Kama homeland for Disintegrating Uralic?

Since I don’t believe in macro-regions of largely continuous ethnolinguistic communities, as I have often said about Slavic (naively associated with prehistoric tribes of Eastern Europe) or Germanic (absurdly considered to be represented by Battle Axe), it is difficult for me to believe that Battle Axe-derived cultures remained of the same Finno-Samic dialects since the Corded Ware expansion…unless we live in Westeros, where everything happens “for thousands of years”.

I have to admit, then, that the now prevalent identification among Uralicists has become quite attractive:

  • Fatyanovo-Balanovo as Finno-Permic:
    • Fatyanovo/Netted Ware with West Uralic (also called Finno-Mordvinic).
    • Balanovo/Chirkovo-Kazan with Central Uralic (Mari-Permic).
  • Abashevo, into the Andronovo-like Horizon through the Seima-Turbino phenomenon, with East Uralic (also Ugro-Samoyedic).

Exactly like the identification of Yamna Hungary – Bell Beaker transition as the North-West Indo-European homeland, it gives us simplicity and small and late ethnolinguistic communities, away from the traditionally overused big and early language territories.

This late homeland would be supported, among others, by:

  • The presence of Indo-Iranian loanwords in Finno-Permic and Ugric (probably also in Samoyedic, either lost, or – much more likely – underresearched), compatible with the immediate contact between Abashevo – Sintashta-Potapovka-Filatovka and Fatyanovo-Balanovo.
  • The supposed expansion of Netted Ware from Fatyanovo to the north-west, which may be explained as the split and expansion of Balto-Finnic and Samic ca. 1900 BC.
  • A longer-lasting Finno-Permic (West+Central Uralic) community contrasting with the early separation of East Uralic.
  • The compatibility of this late expansion with the late expansion of Pre-Germanic from Denmark with the Dagger Period, and of Balto-Slavic with Trzciniec, which puts all three dialects reaching the Baltic Sea in the EBA.

NOTE. I meant to update the linguistic text to include the most recently favoured phylogenetic tree of Uralic languages after Häkkinen (2007, 2009, 2014), which has very quickly become the new normal among Uralicists, but I don’t think I will have enough time to review the necessary papers for that. I am rushing to publish a printed edition, so the text will wind up being a mixture of “traditional” (meaning, basically, pre-2010s) description of Uralic dialects but using modern divisions; say, “West Uralic” instead of “Finno-Samic”. By the way, I am still amazed that none of my reader-haters (or any online user discussing Uralic migrations, for that matter) have come up with the questions that the new division pose, and it supports my suspicion about the complete lack of interest in linguistics of most (a)DNA fans, except for the occasional use of old and free PDFs Googled to support new narratives invented expressly for some qpAdm results…

Textile ceramic styles and influence of Bronze Age cultures divided in clusters.

Problems with this Parpola-Carpelan’s (2012-2018) interpretation include:

  • The differentiation between Fennoscandian Textile Ceramics vs. Netted Ware, which is not warranted in archaeology. The assumption that Netted Ware expanded to the Baltic Sea (as Kallio does, following the traditional view) is thus weak, and it was probably a question of cultural contacts coupled with short-distance population movements/exchange in both directions (from the Baltic to the Volga and vice versa). In fact, the culture division relies on some fairly common and technically simple ornamentation patterns, widespread all over northern Europe, even before the Corded Ware expansion, and it is very difficult to separate certain neighboring Textile Ceramics from Netted Ware groups in southern Finland (i.e. Sarsa-Tomitsa groups).
  • The strict and radical direction described for the Netted Ware by Carpelan, as an eastward and northward expansion, within a very short time frame (ca. 1900-1800 BC), based on few radiocarbon dates, which seems to me like a very risky assumption. We know how this kind of descriptions of direction of culture expansion based on radiocarbon dates has turned out in much more complex “packages”, like the Bell Beaker culture… In fact, the earliest dates for Textile Ware are from the East Baltic, earlier than those of Netted Ware.
  • The assumption that Balto-Finnic traits shared with Mordvinic are a) late and b) meaningful for dialectalization of two closely related dialects, when it is clear that both dialects separated quite early. Phonologically Finnic is more conservative, morphologically less so, and the shared traits include a handful of non-Uralic substrate words which can’t be traced to a single common source, hence they were adopted when both languages had already separated… All in all, Finnic – Mordvinic correspondances are not even close to Italo-Celtic ones, which is clearly fully incompatible with a proposal of a Finnic separation from Mordvinic coinciding with the LBA-IA transition.

Especially problematic for Parpola’s model is the lack of genetic impact in Bronze Age or Iron Age Estonians, not reaching a significant level under any possible statistical threshold – which I am sure was quite disappointing for some of my readers -, but is in line with major archaeological continuity of groups the from region, only disturbed in cultural (and Y-chromosome) terms by the expansion of Akozino warrior-traders all over the Baltic Sea. Any proposed population movement will be very difficult to support in genetics, given the Corded Ware-derived populations that we will see in both regions, and the continued Baltic-Volga contacts since the Corded Ware expansion.

Problems with an interpretation of such a small impact in population genomics includes the similarly weak impacts and haplogroup infiltrations that can be seen among populations basically everywhere in Eurasia, during any given period, and much greater genetic impacts that are supposed to be (or that were certainly) followed by ethnolinguistic continuity.

Distribution of the Akozino-Mälar axes according to Sergej V. Kuz’minykh (1996: 8, Abb. 2).

The Battle Axe question

From Kallio (2015), about choosing a tentative homeland for Proto-Uralic:

(…) linguistically uniform Proto-Uralic would have been spoken in the Volga-Oka region until the mid-third millennium BC when the Proto-Uralic-speaking area would have expanded to the Volga-Kama region as well. By the end of the same millennium, this expansion would have led to the earliest dialectal splits within Uralic into Finno-Mordvin, Mari-Permic, and Ugro-Samoyed. The splitting up of these three soon followed during the early second millennium BC when the Uralic-speaking area finally stretched from the Baltic Sea in the west to the Altai mountains in the east. Indeed, no matter where Proto-Uralic was spoken, the branching into the nine well-attested subgroups (viz. Finnic, Saami, Mordvin, Mari, Permic, Hungarian, Mansi, Khanty, and Samoyed) must have taken less than a millennium, because their shared phonological and morphosyntactic isoglosses are rather limited (see Salminen 2002). The traditional view that all this branching would have taken several millennia violates everything linguistic typology teaches us about the rate of language change.

The basic problem of this identification of Fatyanovo-Balanovo as West-Central Uralic and Abashevo as East Uralic is the nature of the Battle Axe culture, including the Bronze Age East Baltic and Gulf of Finland area. Even if it is accepted that Fatyanovo-Balanovo represented all Western groups, Battle Axe must have represented West Uralic-like dialects.

The ethnolinguistic identification of Battle Axe depends ultimately on the nature of contacts of Fatyanovo/Netted Ware with Battle Axe/Textile Ceramics. If both groups were close and interacted profusely, as it seems, it doesn’t seem granted that we will be able to distinguish a close Para-West Uralic dialect of Scandinavia from the actual expanding Balto-Finnic and Samic dialects, if they were actually linked to the Netted Ware expansion. Also from Kallio (2015):

No doubt the most convincing substrate theory has recently been put forward by the Saami Uralicist Ante Aikio (2004), who has not only rehabilitated but also improved the old idea of a non-Uralic substrate in Saami. His study shows that there were still non-Uralic languages spoken in Northern Fennoscandia as recently as the first millennium AD. Most of all, they were not only genetically non-Uralic but also typologically non-Uralic-looking, bearing a closer resemblance to the so-called Palaeo-European substrates (for which see e.g. Schrijver 2001; Vennemann 2003).

In comparison, the case of Finnic is much more difficult. The fact that Proto-Uralic was not spoken in the East Baltic region means that this area must have originally been non-Uralic-speaking, but so far the evidence for a non-Uralic substrate in Finnic has consisted of appellatives and proper names with no etymology (cf. Ariste 1971; Saarikivi 2004a). Contrary to the proposed substrate words in Saami, those in Finnic show no structural non-Uralisms, as if they had indeed been borrowed from some genetically related or at least typologically similar languages, as I suggested above. Also none of them is more recent than the Middle Proto-Finnic stage, which makes them at least two millennia old. All this agrees with archaeological evidence discussed earlier that the Uralicization of the East Baltic region occurred during the Bronze Age (ca. 1900–500 BC).

The discussion of the paper continues with an unsuccessful attempt to find a hypothetical ancient Indo-European substrate that Kallio believes must be associated with the expansion of Corded Ware, in line with the traditional belief. For example, the often mentioned – almost folk etymology-like, unsurprisingly popular among amateurs – ‘Neva’ as derived from IE “young” is logically rejected…Unlike Parpola, Kallio’s view seems to be confident that Netted Ware (as Textile Ware) expanded into the East Baltic, on both sides of the Gulf of Finland, already during the Bronze Age.

As it has become apparent in population genomics, none of them was right, and Textile Ceramics will essentially show – like Netted Ware – a large genetic continuity of Corded Ware peoples in the whole north-eastern Europe – despite small regional population movements, obviously -, which necessarily implies that the whole Corded Ware culture – and not only Fatyanovo-Balanovo and Abashevo – were Uralic-speaking territories.

The similarities in terms of culture and Y-DNA bottlenecks between Battle Axe and Fatyanovo-Balanovo also imply that the linguistic differences between these groups were probably not many, and became strongly divided only after their territorial division. Continued contacts between Battle Axe- and Fatyanovo-derived groups can explain the proposed contacts (Finnic with Samic, Finnic with Mordvinic) after their linguistic-but-not-physical separation.

East European movement directions (arrows) of the representatives of the Central European Corded Ware Culture (according to I.I. Artemenko).

Battle Axe spoke “Para-Balto-Finnic”?

The Balto-Finnic-speaking nature of Battle Axe is thus supported by:

  • The lack of non-Uralic substrates in Balto-Finnic territory (Kallio 2015).
  • The early separation of Samic and Finnic from Mordvinic, and the virtual identity of Proto-West-Uralic and Proto-Uralic, which suggests that Proto-Uralic spread fast (Parpola 2012).
  • The scarce non-Uralic topo-hydronymy in the East Baltic and around the Gulf of Finland (Saarikivi 2004), comparable to that on the Upper Volga region.
  • The strong influence of a Balto-Finnic-like substrate on Pre-Germanic (or, in Kallio’s opinion, the same Scandinavian substrate influencing both Germanic and Balto-Finnic at the same time), and the continued influence of Balto-Finnic on Proto-Baltic and Proto-Slavic.
  • The continued influence of Corded Ware-derived groups in central-east Sweden in Finland and the East Baltic in terms of agricultural innovations appearing in the LBA, compatible with Schrijver’s proposal of intermediate Germanic-shifted Balto-Finnic groups and Balto-Finnic groups influenced by their pronunciation.
  • The intense Palaeo-Germanic and late Balto-Slavic / early Proto-Baltic superstrate on Balto-Finnic, which place all three dialects around the Baltic Sea since the Early Bronze Age.
  • The easy replacement of a hypothetic Para-Balto-Finnic dialect by incoming Proto-Balto-Finnic-speaking peoples (say, with textile ceramics), without much linguistic impact.

In fact, the continuous contacts of the East Baltic with the Volga, and especially the close interaction with Akozino warrior-traders just before the Tarand-grave period, could be the actual origin of the recent (if any) Finnic-Mordvinic connections that need to be traced back to the LBA-IA (maybe here the number ‘ten’), since most of them can be related to a Pit-Comb Ware culture substrate and earlier contacts through the forest zone, which Samic (due to its early split and presence to the north of the Gulf of Finland during the BA) does not share. In fact, some of them can be traced back to Balto-Finnic first

These are the most often mentioned, in order of descending relevance for a shared ancient community:

  • Noun paradigms and the form and function of individual cases.
  • The geminate *mm (foreign to Proto-Uralic before the development of Fennic under Germanic influence) and other non-Uralic consonant clusters.
  • The change of numeral *luka ‘ten’ with (non-Uralic) *kümmen.
  • The presence of loanwords of non-Uralic origin, related to farming and trees, potentially Palaeo-European in nature.

It’s not only a question of quantity. Are these shared Mordvinic – Balto-Finnic traits really more relevant than, say, those between Italo-Celtic, which are supposed to have formed a community for a very short period at the end of the 3rd millennium around the Alps? Are these traits even sufficient to propose a common early Mordvinic-Finnic group within West Uralic, rather than loose Mordvinic – Balto-Finnic contacts, i.e. contacts between East Baltic (Textile Ceramics) and Volga-Kama (Netted Ware)?

Based on the alternative (Kallio’s) view of continued contacts between Textile Ceramics groups, even without knowing anything about linguistics, you can guess that Parpola is spinning very thin when assuming that these changes suggest that Balto-Finnic may have expanded with Akozino warrior-traders, separating thus ca. 800 BC from Mordvinic…

Genetic findings now clearly help dismiss any meaningful population impact in the LBA-IA transition, although any linguist can obviously argue for linguistic change in spite of major genetic continuity. But then we are stuck in the pre-ancient DNA era, so what’s ancient DNA for.

Middle Bronze Age cultures of Eastern Europe.

Genetic continuity = language continuity?

In the end, it’s very difficult to say how much language continuity there is around Estonia since the arrival of Corded Ware peoples. Looking at Modern Estonians, they have been clearly influenced by recent contacts with Baltic- and Germanic-speaking peoples clustering to the south-west in the PCA. They seem to have also received contacts from north(-east)ern peoples, likely from Finland, evidenced by their shifts toward the modern Estonian cluster during and after the Middle Ages, with a slight increase in Siberian ancestry and N1c subclades associated with Lovozero Ware. How much language change did these contacts bring? Maybe an expansion of Gulf of Finland Finnic (Northern Estonian) over Inland Finnic (Southern Estonian) and Gulf of Riga Finnic (Livonian)? Difficult to know, exactly, but, in the traditional view of Balto-Finnic dialectal distribution among Uralicists like Kallio, possibly no change at all.

So, if the obvious changes in the Estonia_MA cluster relative to Estonia_IA cluster and Estonia_Modern relative to Estonia_MA do not represent radical language change…Why would Estonia_IA represent a change relative to Estonia_BA, when it is statistically basically the same? Or Estonia_BA relative to CWC_Baltic? Because of the infiltration of haplogroup N1c around the whole Baltic? Because of the occasional 1% “Siberian” ancestry in some non-locals of varied haplogroups across the whole Baltic area?

In spite of all this, the amount of special pleading we are seeing among openly Nordicist amateurs when discussing the Uralic homeland relative to the Indo-European question in genetics has become a matter of plain willful ignorance. Like the living corpses of the Anatolian homeland, the Armenian homeland, the OIT proponents, or the nativist Basque R1b association, the personal involvement in the revival of “R1a=Indo-European” and “N=Uralic” trends is just painful to watch.

[Next post in this line, if I manage to make time for it: “Genetic (dis)continuity in Central Europe“. Let’s see if early Balts and early Slavs, as well as Germanic peoples, show a cluster closer to Danubian EBA (viz. Maros), Hungary-Balkans BA, and Urnfield-related samples than their predecessors in their areas, i.e. away from East Corded Ware groups… If you want, you can enjoy for the moment the new PCAs I could get done and the tentative map of languages in the Early Bronze Age, that will probably give you the right idea about early Indo-European and Uralic population movements]

European Early Bronze Age: tentative language map based on linguistics, archaeology, and genetics. See full map.


Genetic landscapes showing human genetic diversity aligning with geography


New preprint at BioRxiv, Genetic landscapes reveal how human genetic diversity aligns with geography, by Peter, Petkova, and Novembre (2017).


Summarizing spatial patterns in human genetic diversity to understand population history has been a persistent goal for human geneticists. Here, we use a recently developed spatially explicit method to estimate “effective migration” surfaces to visualize how human genetic diversity is geographically structured (the EEMS method). The resulting surfaces are “rugged”, which indicates the relationship between genetic and geographic distance is heterogenous and distorted as a rule. Most prominently, topographic and marine features regularly align with increased genetic differentiation (e.g. the Sahara desert, Mediterranean Sea or Himalaya at large scales; the Adriatic, inter-island straits in near Oceania at smaller scales). We also see traces of historical migrations and boundaries of language families. These results provide visualizations of human genetic diversity that reveal local patterns of differentiation in detail and emphasize that while genetic similarity generally decays with geographic distance, there have regularly been factors that subtly distort the underlying relationship across space observed today. The fine-scale population structure depicted here is relevant to understanding complex processes of human population history and may provide insights for geographic patterning in rare variants and heritable disease risk.

Regional patterns of genetic diversity. a: scale bar for relative effective migration rate. Posterior effective migration surfaces for b: Western Eurasia (WEA) e: Central/Eastern Eurasia (CEA) g: Africa (AFR) h Southern African hunter-gatherers (SAHG) k: and Southeast Asian (SEA) analysis panels. ‘X’ marks locations of samples noted as displaced or recently admixed, ‘H’ denotes Hunter-Gatherer populations (both ‘X’ and ‘H’ samples are omitted from the EEMS model fit); in panel g, red circles indicate Nilo-Saharan speakers and in panel h, ‘B’ denotes Bantu-speaking populations. Approximate location of troughs are shown with dashed lines (see Extended Data Figure 4). PCA plots: c: WEA d:Europeans in WEA f: CEA i: SAHG j: AFR l: SEA. Individuals are displayed as grey dots. Large dots reflect median PC position for a sample; with colors reflecting geography matched to the corresponding EEMS figure. In the EEMS plots, approximate sample locations are annotated. For exact locations, see annotated Extended Data Figure 4 and Table S1. Features discussed in the main text and supplement are labeled. FST values per panelemphasize the low absolute levels of differentiation.”

Among ‘effective migration surfaces‘ (or potential past migration routes), the Pontic-Caspian steppe and its most direct connection with the Carpathian basin, the Danubian plains, appear maybe paradoxically as a constant ‘trough’ (below average migration rate) in all maps.

After all, we could have agreed that this region should be a priori thought as the route of many migrations from the steppe and Asia into Central Europe (and thus of ‘effective migration’) in prehistoric, proto-historic and historic times, such as Suvorovo-Novodanilovka (Pre-Anatolian), Yamna (Late Indo-European), probably Srubna, Scythian-Cimmerian, Sarmatian, Huns, Goths, Avars, Slavs, Mongols

It most likely (at least partially) represents a rather recent historical barrier to admixture, involving successive Byzantine, South Slavic, and Ottoman spheres of influence positioned against Balto-Slavic societies of Eastern Europe.

Location of troughs in West Eurasia (below average migration rate in more than 95% of MCMC iterations) are given in brown. Sample locations and EEMS grid are displayed for the West Eurasian analysis panel. FST values are provided per panel to emphasize the low absolute levels of differentiation.

Featured image, from the article: “Large-scale patterns of population structure. a: EEMS posterior mean effective migration surface for Afro-Eurasia (AEA) panel. ‘X’ marks locations of samples excluded as displaced or recently admixed. ‘H marks locations of excluded hunter-gatherer populations. Regions and features discussed in the main text are labeled. Approximate locations of troughs are annotated with dashed lines (see Extended Data Figure 4). b: PCA plot of AEA panel: Individuals are displayed as grey dots, colored dots reflect median of sample locations; with colors reflecting geography and matching with the EEMS plot. Locations displayed in the EEMS plot reflect the position of populations after alignment to grid vertices used in the model (see methods).”

Images and text available under a CC-BY-NC-ND 4.0 International License.

Discovered via Razib Khan’s blog.


Königsberg (AKA Kaliningrad) under international law: Russian, German, Polish, Lithuanian, or simply Prussian?

The progress of the ‘star wars’ (AKA missile shield) affair, which Russia seemed willing to aggravate by talking about plans to station missiles in Kaliningrad, without any concerns whatsoever for the welfare of Kaliningraders and Europeans, should make the European Union reexamine its current policy under the Kaliningrad Strategy, of collaborating with Russia by facilitating the transit of goods and persons and helping its socio-economic development.

Instead of just hearing what Russians have to claim before the international community, the EU should ask the international community by which right keeps the Russian Federation hold on Königsberg territory, and should demand from Russia a date for devolution, no matter how hard Russian media propaganda tries to avoid the question:

Although disputes over the status of Russia’s westernmost exclave of Kaliningrad have practically ceased, this should be regarded as a signal that all the parties concerned are aware of the serious repercussions that instability in that region could cause.

Geopolitical Stability has been by far the most repeated pro-Russian argument since the 90’s, also in official European Union forums (see Freedom to Kaliningrad thread); it is easily summed up into a “let’s maintain the statu quo to avoid destabilizing the region”. The murmuring of those plans to use Kaliningrad as missile base made by Russian military officials to the press, to escalate tensions in the missile shield affair, has shown how the Russian Federation respects the will of Europeans for stability in the region. Not to talk about Russia’s lack of respect for the lives of thousands of European citizens in this winter’s gas disputes, or its lack of respect for Estonian democratic decisions, or its support for the authoritarian Belarusian regime of Lukashenko

Other great arguments made by pro-Russians include “Nazi Germany”, “World War II” and “Mother Russia”, and are easily read elsewhere in Russian media and blogs when the Kaliningrad question is mentioned. Nevertheless, most Kaliningraders – whether ethnic Russians or not – show often an open mind about the return options. And even official Russian media like Russia Today recognize still in 2009 (only in English texts for outsiders) the Lithuanian claims to the territory and its return; East German rights are still taboo in Russian ‘free’ media, while Polish claims are probably too weak to be worth mentioning:

The region became an administrative unit of Russia [sic] in 1946 after the Potsdam conference and the partition of Germany. Although it solidified as an administrative entity, with the collapse of the Soviet Union, the issue of reassimilating the Kaliningrad region into its historic entity of Lithuania arose.

According to a thorough study on the question (The Kaliningrad Challenge, 2003) Russia has been always concerned about the risk of separatism in Kaliningrad, which might be greater than expected if the European Report The EU and Kaliningrad (2002) is correct in assessing that Kaliningrad’s level of development is closer to Lithuania and Latvia than previously thought. In that sense, ethnic Russian Kaliningraders see Kaliningrad in the future as another Baltic Republic, either still somehow federated to Russia with great autonomy or fully independent. Moreover,

There are opinion polls – now more frequently held within blogs and forums – which show that Kaliningraders occasionally imagine their future not so much as a fourth Baltic Republic, but as part of a return to Germany

As it has been already argued on the situation of Königsberg/Kaliningrad region and the Northern Territories/Southern Kuril Islands under international law:

In a similar way, the Soviets also refused to discuss the final peace settlement in Europe after the Second World War. It is important to emphasize that neither the United States nor Britain agreed at Potsdam or anywhere else to the transfer of East Prussia or part of the Königsberg Region to the Soviet Union. Thus, although the Kaliningrad Region is currently administered by Russia, it is not a legal part of Russia.

Stalin was seeking a deal on East Prussia at the Tehran conference in 1943, drawing a line in red pencil on the map “to illustrate the fact that, if part of eastern Prussia, including the ports of Könisberg and Tilsit, were given to the Soviet Union, he would be prepared to accept the Curzon line […] as the frontier between the Soviet Union and Poland.”

This line goes roughly along the current border between the Kaliningrad Region and Poland, but Stalin’s red line on the map went virtually through the cities of Königsberg and Insterburg (see the Map). Charles E. Bolen, the interpreter for the American delegation, says in his memoirs that during their discussion, Stalin and Churchill virtually agreed on the future borders of Poland, but the official American record of the conversation says that “although nothing was stated, it was apparent that the British were going to take this suggestion back to London to the Poles.”

On February 11, 1945, at the Crimea (Yalta) Conference, the Big Three agreed on the Curzon Line as the boundary between Poland and the USSR. However, the archival material clearly shows that there had not been any legally binding agreement made between the allies about the transfer of the Königsberg Region to the Soviet Union at any of the Second World War conferences. This is why Stalin attempted to secure his gains at the Potsdam conference in Berlin, which took place from July 17 to August 2, 1945.

After the end of the Second World War, the Kaliningrad question began by Stalin’s personal will of revenge against Germany:

Königsberg was neither appended outright to the Soviet Union nor was it to be considered part of the Soviet zone of occupation, which had been outlined earlier in the agreement.

[The Soviet Union] acted decisively to completely eradicate the German presence in Königsberg and replace it with Soviet presence. This began even before the end of hostilities with the Reich:

Königsberg was destroyed in the last weeks of the war when there was no real reason to assault it. When the soldiers of the Byelorussian front were dying in its streets in the first week of April, 1945, the rest of the Red Army was already besieging Berlin. Seven centuries of history went up in smoke in one week of shelling and bombing. By then, the decision to annihilate East Prussia and grant Königsberg to the Soviet Union had already been taken, so the reason for its destruction remains a mystery. Did Stalin take the decision in a fit of war revenge? Did he think that the setting of an ancient bourgeois city would hamper the development of the new Soviet city he wanted to build in its place? Or did he fear that, unless turned into a pile of ruins, Königsberg might not be conceded to him by the Allies after all? Pictures and models in the bunker-cum-museum where the capitulation of the city was signed are revealing. Most of the destruction was done after-wards, when the victors took to the task of building a new city on the ruins of the old…

While the destruction of the city’s infrastructure was underway, an equally brutal purge of its population through gang rapes and indiscriminate crimes was carried out:

The demography of that part of Lithuania Minor which is under direct Soviet administration, the “Oblast,” has changed in the most radical way in all its history. The original population of the area — German as well as Lithuanian — has disappeared completely. Many had fled before the Soviet armed forces invaded the area in 1945; those who remained — several hundred thousand — either perished from hunger or disease or were deported to Siberia; the others were expelled to Germany in 1949. They all — about 1,200,000 before World War II — were replaced by about 600,000 settlers from the northern and central parts of Russia. The administration and economy of the “Oblast” has been reorganized to conform with Soviet models and practices. It has been fortified to serve the strategic aims of the Soviet Union.

Modern Claims in Europe

After the fall of the Soviet Union, there were 4 main alternatives for the future of Kaliningrad, following Raymond A. Smith’s article The Status of The Kaliningrad Oblast Under International Law (1992), which argues in favour of the Lithuanian claim, but which also addresses some historical and political questions:

From the historical [point of view] sovereignty over the territory of the Kaliningrad Oblast passed over the course of centuries from the the indigenous Old Prussian population, to the Teutonic Order, to the Kingdom of Poland, to the Kingdom of Prussia (later the German Empire) and finally, perhaps, to the USSR/RSFSR. It is not surprising, then, to find that each of these entities (with the exception, of course, of the Teutonic Order) has a conceivable claim to this territory. This section examines the legal basis, or lack thereof, of the actual or potential claim of each entity, as well as the potential claim of the indigenous population.

  • The German Claim: Some Germans challenge the validity of both the Final Settlement and the original “dismemberment” of the German Reich.
    Their arguments are complex but can be reduced in essence to two claims:

    1. the Allies had no power to allow German territory to be annexed by other countries
    2. the West Germany and even the modern Federal Republic of Germany are not coextensive with the German Reich and are therefore not competent to speak for it in its entirety

    The first proposition is supported by numerous charges: that the guarantees of self-determination in the Atlantic Charter, the UN Charter, and the Vienna Convention on the Law of Treaties were ignored; that the Ancient Roman principle of ex injuria non oritur jus prohibits punishing Germany by unilateral confiscation of its territory; that the powers of the Allies as occupiers were strictly curtailed by the Hague Laws of War of 1907; that use of German lands as “compensation” to Poland for lands lost to the Soviet Union has no basis in international law; and many others.

  • The Russian Claim: As the historical overview recounted, the working premise of the Potsdam Conference was that the Soviets would receive the Oblast at the final peace conference. The Allies specifically committed themselves to supporting the Soviet claim in the Final Settlement, but when that settlement was finally signed in 1990, specific title was not transferred. Why the Final Settlement did not include a specific statement of transfer is unclear. The seemingly most probable reason is that the transfer of Kaliningrad to the Soviet Union is considered a fait accompli and that the legal niceties of including a specific mention of transfer were outweighed by potential political embarassment such a mention might have caused the Kohl government. Such a position assumes that the tranfer has already taken place, an assertion which rests on shaky ground.

    Similarly, the Act of Military Surrender specifically indicates that the occupation itself did not effect the annexation of Germany. Thus, although Germany surrendered unconditionally, none of its territories were automatically annexed to any other state. Such annexation would have to be made explicit in a legally binding document. Only “administration” was established by the Potsdam Agreement, however, and “administration” is definitely not the same as “annexation” under international law.

    Rather than present arguments based on international law, Stalin advanced the law of revenge. ‘The Russians had suffered so much and lost so much blood, they were anxious to have some small satisfaction to [sic] tens of millions of their inhabitants who had suffered in the war,” Stalin said at Potsdam.

    In the absence of ethnic and historical claims to shore up their questionable legal claim, then, the only argument which the Soviet Union can depend upon is the principle of prescriptive claim. This principle transfers title to land when a country has held it for a long period of time without protest by the land’s original owners or by the international community at large. No specific time frame is suggested for the acquisition of prescriptive claim. Grotius suggested 100 years, a figure which the Permanent Court of International Justice endorsed in 1933. The International Court of Justice, on the other hand, said that fifty years had been long enough for a boundary between Venezuela and British Guiana to have legal effect.

  • The Polish Claim: Poland has no ethnic claim to the Oblast. Although the southern half of East Prussia was occupied mainly by Polish Masurians, they had almost no presence in the northern part.

    Poland’s historic claim is only marginally stronger. For some two centuries, Prussia was a fief of the Polish King, but during that period the area remained firmly under German control. In any case, title was decisively transferred by the Treaty of Wehlau in 1657. During World War II many Poles operated under the belief that all of East Prussia would become theirs, but they were never legally promised the territory in its entirety.

  • Lithuanian Claim: The claim of the Lithuanian state could rely upon both ethnic and historical grounds.
    1. The Lithuanians may argue that
      the first peoples to hold sovereignty over the region were ethnic Lithuanians and closely related Old Prussians, and
    2. the pre-1945 population outside the cities of the Oblast was largely of Lithuanian origin. If the status of the Oblast were to be altered in the future, then, the Lithuanian state could have a strong argument for assimilating this remainder of Lithuania Minor.

    The idea of unifying the Oblast with the rest of Lithuania has strong historical precedents. Lithuanian assemblies met in Chicago and New York in 1914, The Hague in 1916 and Berne in 1917 to demand an independent Lithuania including all of Lithuania Minor. An assembly in Vilnius in 1917 restated the problem to define the new Lithuania within its “ethnographic borders,” a concept endorsed by a later assembly in Voronezh the same year.
    Finally, on November 30, 1918, the National Council of Prussian Lithuania issued the Declaration of Tilsit:

    Taking into account that everything that exists has a right to continue existing and that we, Lithuanians who live here in Prussian Lithuania, are the majority of the population of this land, we demand, on the basis of Wilson’s right of national self-determination, that Lithuania Minor be joined to Lithuania Major

    The clearest catch here is that any annexation of the Oblast by Lithuania might hinge upon the democratic decision of an indigenous Lithuanian majority to authorize such an annexation. And, as we have seen, virtually none of the indigenous Lithuanian population remains in the Oblast, having fled or been killed or exiled after World War II. This raises the final claim to be discussed — that of the indigenous population.

  • The Claim of the Native Population: The right to national self-determination is one of the main cornerstones of the contemporary international legal order. Eight of Wilson’s Fourteen Points refer to such concerns. The Atlantic Charter’s third and fourth principles call for self-determination in matters of both boundaries and choice of government. The Charter of the United Nations calls for colonial powers to foster self-determination in “non-self governing territories”. That right might be interpreted as concerning:
    1. The Oblast’s postwar ethnic Russian settlers – as opposed to central Soviet or Russian authorities.
    2. the traditional population which was decimated or expelled en masse after World War II, which is defended on the grounds that forcible deportations of native populations is clearly in violation of international law – native Königsbergers expelled after World War II, then, have a right under international law to choose to return to their native land.

    On that question, there is the precedent of United Nations action regarding the settlement of Gibraltar:

    As in the case of the Oblast, the key issue was whether the rightful native population of the Rock should be considered to be the contemporary residents or an earlier population who had been compelled to depart in 1704. The British argued that over the centuries since 1704 a permanent and authentic population had been developed on the Rock, which now had the right to determine their own fate. The Spanish countered that the post-1704 population were “pseudo-Gibraltarians” and that the rightful rulers of Gibraltar Rock were the descendants of Spaniards who had resettled, for the most part, in the nearby city of San Roque.

    Under pressure from the United Nations to end its colonial occupation of Gibraltar and in an attempt to settle the status of the Rock once and for all, the British government conducted a plebiscite in 1967. The choices were stark — full political affiliation with either Great Britain or with Spain — and the result was unequivocal: 12,138 to 44 in favor of Great Britain. Nonetheless, the U.N. General Assembly once again condemned British occupation of Gibraltar, this time in the strongest language yet. It, in essence, declared the plebiscite null, accused the British of resisting decolonization, and called once again for immediate negotiations between Great Britain and Spain for a transfer of sovereignty.

    Whatever the merits of the Gibraltar case, the precedent for the Oblast is clear. If the rights of native populations can stretch back to 1704, then surely the postwar expellees from the Oblast would have an unambiguous right to return to their homeland and choose its political fate — be that choice in-dependence or association with another state. The current population of the Oblast would presumably have no say in the territory’s political future.

    The key difference between Gibraltar and the Oblast is that in the former case, there actually is a population in San Roque able and willing to resettle the Rock. No analagous “population-in-exile” exists in the case of the Oblast. Rather, much of the population of Königsberg was killed or died in exile. Those who were deported to Germany (and their descendants) in all likelihood now enjoy a standard of living which is, at least quantitatively, many times better than any which would be possible in the backward conditions of the Oblast. Further, most — although far from all — Germans seem to have accepted the loss of the prewar lands; the idea of reclaiming part of East Prussia would not necessarily resonate with much of the population. It seems extremely unlikely, then, that more than a handful of such native German Königsbergers would wish to uproot and resettle in the Oblast.

Even with German and Lithuanian strong claims about the Soviet colony of Königsberg opposing the legality of Stalin’s annexation, Russia did in the 90’s what it was used to in such cases when the Soviet Union was still a Great Power: they took the easy way, and annexed the territory to Russia, expecting the international community to accept it. Which is nice, because the EU as a Great Power will therefore be entitled to follow the same principle in the future…

In my personal opinion, the European Union faces today 3 alternatives, given Russia’s will to retain Stalin’s European exclave no matter how illegal or illegitimate it is from an international point of view:

  1. Support modern Kaliningraders in their demands of greater autonomy within the Russian Federation – and maybe a future separation from it -, which is the fairest position under modern international law, which demands non-belligerant positions (against Russia in this case) and respect for human rights – Russian settlers and their families. This is certainly the option of most Kaliningraders of Russian ethnicity, as well as most EU-politicians.
  2. Support Germany’s or Lithuania’s claims (or both), seeking to integrate Kaliningrad within the European Union, maybe as a sort of a Baltic territory co-administered by both Germany and Lithuania, financing the return of (families of) expellees to Königsberg, and the return of (willing) families of Russian settlers to Russia. This is the option preferred by many Germans and (I guess) most Lithuanians.
  3. Support the creation of a modern Baltic Prussian State (Prusa), which could help unite the Pro-Baltic (and Pro-European) attitude of Russian Kaliningraders, the will of native peoples and their families to return to East Prussia, as well as claims of EU member states to integrate Königsberg in Europe, by embracing Old Prussian history of the territory and its peoples. Modern organizations supporting the revival of the Old Prussian language would probably support its revitalization in Königsberg include the future Research Institute of Prussology and the Prussian language organization in Poland.

The third is my preferred option, not because I am some kind of language revival freak (what I possibly am, given that I also support Old Prussian language revival), but because what many (want to) regard simply as ethnic German and ethnic Lithuanian inhabitants of East Prussia in 1945 were in fact descendants of Old Prussians who had lost their language in favour of either German or Lithuanian languages, depending on the territories they dwelled when they ceased to speak Prussian. Given that historical, cultural and linguistic background of the Königsberg (or East Prussian) territory, the European Union should take action supporting the return of those expelled peoples and their families to their ancient territory, which they were forced to leave half a century ago.

There is therefore no need to support the adscription of East Prussia to modern countries or peoples, be it Russia, Germany, Poland or Lithuania. And the only alternative to modern peoples, cultures and states is to support a linguistic and cultural revival of a Prussian people and language that should have never disappeared.