Population replacement in Neolithic Spain?

By Razib Khan | June 28, 2012 1:16 pm

There’s a new ancient DNA paper out which examines the maternal lineage and the autosomal background of two individuals extracted from a Spanish site dated to 7,000 years before the present. That is, during the European Mesolithic. In other words, these are the last wave of Iberian hunter-gatherers before agriculture. I have placed the PCA, with some informative labels, to illustrate the peculiarity of these samples. Here’s the abstract:

The genetic background of the European Mesolithic and the extent of population replacement during the Neolithic…is poorly understood, both due to the scarcity of human remains from that period…The mitochondria of both individuals are assigned to U5b2c1, a haplotype common among the small number of other previously studied Mesolithic individuals from Northern and Central Europe. This suggests a remarkable genetic uniformity and little phylogeographic structure over a large geographic area of the pre-Neolithic populations. Using Approximate Bayesian Computation, a model of genetic continuity from Mesolithic to Neolithic populations is poorly supported. Furthermore, analyses of 1.34% and 0.53% of their nuclear genomes, containing about 50,000 and 20,000 ancestry informative SNPs, respectively, show that these two Mesolithic individuals are not related to current populations from either the Iberian Peninsula or Southern Europe.

Here’s another PCA showing one individual on a more fine-grained representation of European populations:


Perhaps the most interesting aspect here is something Dienekes pointed out from the supplements: these two individuals are not only outside of the range of extant European populations in their positioning on a global PCA plot, but they are shifted toward East Asians.

The fact that these two individuals, who really come close to being one data point because they are likely rather closely related, are outside of the modern European population distribution isn’t too surprising. 7,000 years is a long time, and we can’t assume that ancient populations can be recomposed as combinations of the variation of modern populations. But, the shift toward East Asians is surprising to me, because they are Iberian individuals. In PCA and model-based (e.g., ADMIXTURE) clustering frameworks modern Iberians, and populations from Southwest Europe in general, are the most distant from East Eurasians of all West Eurasians. Dienekes opines:

It now appears clear that the Mesolithic substratum in Europe was:

1. Well outside the modern range, contributing a little to extant populations
2. Its contribution in northern populations was higher than in southern ones
3. It may be responsible for the pattern of Asian-shift observed for non-Mediterranean European populations

…It seems that this was the composition of the pre-Neolithic population of Europe that was later supplanted first by the “Mediterranean”/”Southern” components during the early Neolithic, and later by the “West_Asian”/”Caucasus”/”Gedrosia” components, perhaps during the Copper Age. We’ll see whether my prediction pans out soon enough.

On the broadest level I think Dienekes model is entirely possible. I’d give it the highest probability of the range of options, though I have a high uncertainty. The question is the weight of the contributions. Let’s rename the various groups A, B, and C, for the three waves in chronology (hunter-gatherers, first wave farmers, and second wave farmers). Then any European population is:

xA + yB + zC, where x + y + z = 1

About 10 years ago B + C would be one class, so B, and you had a pan-European estimate of:

0.75A + 0.25B

Now some scholars are trying to revise that, and reduce the weight for A. I think we need to be very careful, because we’ve already been burned by overly elegant and simplistic models of the settlement of Europe. For example, here are my made-up estimates quantitatively from everything I’d read so far:

Finns = 0.90A + 0.05B + 0.05C

Lithuanians = 0.85 A + 0.05B + 0.10C

Irish = 0.60 A + 0.30B + 0.10C

Germans = 0.60 A + 0.20B + 0.20C

Basques = 0.40 A + 0.60B

North Italians = 0.40 A + 0.30B + 0.30C

The exercise above was not to give you accurate numbers that I’m sure of, but to give you numbers instead of verbal labels which are imprecise. I’m quite willing to “update” my estimate in the future, and expect to. In the paper the authors highlight that the mtDNA lineages that the two individuals carry is modal in the Sami of northern Finland. Genetically the Sami seem more Finnish that the Finns. But, they also clearly have an “eastern” affinity (diminished, but still present in Finns as well). This dovetails with linguistic connections to North-Central Eurasia, and the margins of Western Siberia. What’s going on here? One hypothesis has been that the Finnic languages (and the Sami) are culturally intrusive, and the Siberian genetic affinity is a signal of this ancient core admixed with the local substrate, which resembles that of Scandinavians.

These results make me update my assessment, and increase my own probability that the Finnic people have deep cultural roots in Northeast Europe. This was already my hunch based on model-based clustering which seemed to show that there were “southern” modal elements present in Scandinavians lacking in Finns or Sami. If Finns or Sami were relatively late arrivals, I would have expected to be more diverse in the complement of ancestral elements, not less. Now the set of results form Mesolithic European genomes indicate to me that what we may be seeing in the Siberian affinity of the Sami (and to a lesser extent the Finns) are the echos of a post-Ice Age expansion of Palearctic peoples from the center of Eurasia which ranged west toward the Atlantic and east toward the Sea of Japan. I suspect that this Palearctic population did not move into a totally empty landscape, so the Mesolithic peoples which the West Asian and East Asia farmers displaced or assimilated were not entirely similar. Rather, they were themselves syntheses between hunter-gatherer groups.

Obviously this is an interesting time to moot these questions. The major issue we need to keep in mind is not to move from one enthusiasm to another. A model of preponderant biological continuity between European hunter-gatherers and early agriculturalists is now in serious doubt, especially for southern Europe. But that does not mean that we should move to a model where the hunter-gatherers were replaced in totality by agriculturalists. This does not seem to have happened in northeast Europe, and the model of replacement itself is probably more complex than a single-step transition.

Citation: Current Biology, 28 June 2012 doi: 10.1016/j.cub.2012.06.00

CATEGORIZED UNDER: Agriculture, Anthroplogy
  • Halvorson

    I actually think the trend in Uralic linguistics over the past couple of years has been to revise the age of Proto-Uralic in a younger direction, so that the consensus now seems to be that it was spoken as late as 3000-2000 B.C., with proto-Finnic spoken around 500 B.C in what is today Estonia.

    A couple years back there was a paper published which showed that hunter gatherers of the late (3200-2300 B.C.) Swedish coastal Pitted Ware Culture belonged predominantly to haplogroups U4 and U5a, as opposed to the V and U5b1b that you see in the Saami.

    http://blogs.discovermagazine.com/gnxp/2009/09/saami-not-descended-from-swedish-hunter-gather

    The big conclusion everyone drew from this was that old Scandinavian hunter gatherers were most like today’s Baltic peoples, where U4 is most common. The possibility that no one at the time seemed to consider was that maybe the Pitted Ware people, whose cultural artifacts resemble that of the much larger NE European Comb Ceramic complex, were themselves intrusive to Scandinavia! In most people’s mental models of the Neolithic transition in Europe, the farmers keep moving and moving while hunter-gatherers sit around and wait to die. But HGs migrate all the time: think of the mass migration of Amerindian tribes in North America set in motion by the arrival of Europeans. It’s possible that at the same time farmers were pushing into the domains of foragers, different cultures of foragers were doing the same thing to each other. There’s evidence of this in Scandinavia: Ante Akio showed back in 2004 that Sami has a very large substratum unrelated to any other European language and which presumably comes from the Paleolithic Euros the Sami absorbed when they arrived from Finland.The big point here is that just because the Finnic peoples have a lot of old European blood in them doesn’t really have that much bearing on whether or not their language is old European (or at least indigenous to Finland/Estonia): the linguistic consensus is strong that it’s not.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    The possibility that no one at the time seemed to consider was that maybe the Pitted Ware people, whose cultural artifacts resemble that of the much larger NE European Comb Ceramic complex, were themselves intrusive to Scandinavia!

    some people did bring this up on the blogs and what not. i remember the discussion at the time….

  • http://dienekes.blogspot.com Dienekes

    The Comb Ceramic people were partially Mongoloid, and the Ajv Neolithic hunter-gatherers of the PWC do not appear to have that type of admixture.

    I disagree with Razib’s levels of “A”. mtDNA haplogroup U appears to be the best marker of Mesolithic Europeans and it reaches levels of ~10-20% in most of Europe. On the Y-chromosome side, pretty much the only game left in town is Y-haplogroup I as a marker of Mesolithic Europeans; this has been affected by recent expansions in the Balkans and Scandinavia, but it also reaches low overall frequencies in most places.

    So, I would say that the pre-Neolithic Europeans did leave their genes to modern Europeans, but that is a minority element everywhere.

  • Halvorson

    @3 You’re right about CC, at least according to my copy of The Races of Europe.

    One thing I never understood about I1 was how it could be a Mesolithic marker if it exists at double plus the frequency of mtDNA U in Norway and Sweden. Wouldn’t that imply that it was the foragers who were absorbing the farmers and not the other way around? And if R1a and R1b mark the arrival of Indo-Europeans, what Y haplogroups did the first farmers in Northern Europe possess? G and I2? How could their frequency have fallen so low while I1 stayed so high?

  • http://dienekes.blogspot.com Dienekes

    One thing I never understood about I1 was how it could be a Mesolithic marker if it exists at double plus the frequency of mtDNA U in Norway and Sweden.

    Y chromosome lineages mostly appear to coalesce to recent (late Neolithic or Bronze Age) ancestors. They represent successful founder lineages. The various U sublineages have pre-Neolithic antiquity in Europe and do not appear to have been affected by the “super-mother” effect that would have substantially increased their frequencies.

    what Y haplogroups did the first farmers in Northern Europe possess? G and I2?

    Definitely at least G. There were others detected that may not have been placed fully in the phylogeny because of the limited number of SNPs tested.

  • Karl Zimmerman

    Arrg. I had a long reply written, and accidental use of the back button deleted it.

    To recap, I’ve noticed there’s a striking correlation geographically speaking with Mesolithic ancestry in Europe and blond hair and blue eyes, with the Saami the only exception.

    While I know there have been hypotheses that both only expanded post-agriculture, to the best of my knowledge blond hair is now agreed to be a Mesolithic mutation. Blue eyes are thought to be more recent, but it’s difficult to see how the populations with the least genetic replacement would gain the highest incidence of blue eyes by far, particularly because unlike light skin, there’s no apparent fitness adaptation to light hair or eyes (although light eyes are probably negatively selected for in places with brighter sunlight to some degree). Sexual selection alone can only work as an explanation if you are dealing with a founder effect, or an isolated sub-population – not spread across an entire set of established populations, as local standards of beauty will vary with time and geography.

    While I wouldn’t hypothesize that everyone in Europe looked like an Aryan before agriculture (after all look at the Saami), I’d guess that the trait itself is basically 1 for 1 linked with this population substratum, and later expansions of it into Central Asia (Iranians, Tochaharians, etc.) were because some Indo-European speaking peoples became hybrids of Mesolithic/first wave agriculturalists and the IE-source population fairly early on.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    particularly because unlike light skin, there’s no apparent fitness adaptation to light hair or eyes (although light eyes are probably negatively selected for in places with brighter sunlight to some degree)

    just to be clear, several of the loci implicated in skin color have correlated responses with hair and skin color. so, for example, herc2-oca2 locus which is responsible for 75 percent of the blue vs. non-blue eye color variance in europeans is a generalized pigmentation locus (oca2 is a gene which has an albinism mutation).

    Sexual selection alone can only work as an explanation if you are dealing with a founder effect, or an isolated sub-population – not spread across an entire set of established populations, as local standards of beauty will vary with time and geography.

    hm. depends on the type of sexual selection. e.g., if it’s ‘sensory bias,’ then it could be species wide.

    and later expansions of it into Central Asia (Iranians, Tochaharians, etc.) were because some Indo-European speaking peoples became hybrids of Mesolithic/first wave agriculturalists and the IE-source population fairly early on.

    my hunch too. the ‘west asian’ minor element which is lacking in the basques, sardinians, and finns is suspicious too me. so i think that’s the ‘original’ PIE population. otoh, the european-like populations in western mongolia seem to have conventional northern european phenotypes (blue eyes). this both from depictions in paintings, and ancient genetics.

  • Karl Zimmerman

    Razib -

    I didn’t know that hair, skin, and eye color were so tightly linked genetically, although it makes sense, given the close linkage of the traits even in areas like South Asia where at least some fair component hybridized into the population. It seems like light eyes with dark skin are much easier to link than the European blondism mutation and dark skin however. So even if these were novel mutations in the post-agricultural era, presuming they further lightened the skin, and allowed for more Vitamin D production, they could be strongly selected for. That said, there still would be the issue that both traits seem to follow a Mesolithic/Neolithic cline more closely than a north/south cline, with more “Mediterranean” groups like the Irish, Welsh, and Northern French a bit darker despite a similar latitude to the Baltic.

    I think I get what you mean for sensory bias – the same sort of thing which makes a peacock’s tail feathers as long as possible, and a deer’s antlers (in some species) as big as possible. But while it seems that pale skin is found attractive on women across many cultures (even in China, where IE-influence can’t be claimed) I’m not aware of any study which has found blond hair similarly universally appealing. Indeed, I would presume in the absence of any other form of selection, it would form an equilibrium with darker hair, as those with the rarer phenotype would become desired as “exotic” until they became commonplace enough to be ho-hum, at which point it would become a matter of individual taste, and simply not important for many at all.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    It seems like light eyes with dark skin are much easier to link than the European blondism mutation and dark skin however.</i.

    i think that's because no one gene has such a strong link to hair color as herc2-oca2 has to eye color. hair color is more like skin color in its genetic architecture, mildly polygenic.

    also:

    http://essays.backintyme.com/item/4

    re: sexual selection. all talk, really difficult to confirm anything. if some neandertals had red hair, and some melanesians have blonde hair, i think there's a high chance it has adaptive directly or is a side effect.

  • Lassi Hippeläinen

    Razib: “I’m quite willing to “update” my estimate in the future, and expect to.”

    Good. You could start by tweaking the coefficients for Finns and Lithuanians so that they add up to one.

    Halvorson: “I actually think the trend in Uralic linguistics over the past couple of years has been to revise the age of Proto-Uralic in a younger direction, so that the consensus now seems to be that it was spoken as late as 3000-2000 B.C., with proto-Finnic spoken around 500 B.C in what is today Estonia.”

    I’m always a bit skeptical about linquistic datings, especially if they are used to argue about the history of peoples. A recent DNA study showed that Gotland had Finnish HG inhabitants at least 5000 years ago. At the same time a farmer on the nearby Swedish continent was Mediterranean.
    http://www.nature.com/news/ancient-swedish-farmer-came-from-the-mediterranean-1.10541

  • http://dienekes.blogspot.com Dienekes

    While I wouldn’t hypothesize that everyone in Europe looked like an Aryan before agriculture (after all look at the Saami), I’d guess that the trait itself is basically 1 for 1 linked with this population substratum, and later expansions of it into Central Asia (Iranians, Tochaharians, etc.) were because some Indo-European speaking peoples became hybrids of Mesolithic/first wave agriculturalists and the IE-source population fairly early on.

    There seem to be no migrations out of Europe in the time frame in question, and in relation to the Indo-Europeans, for several reasons:

    1. Lack of Y-haplogroup I in South Asia
    2. Lack of “Mediterranean”/”Southern” component of first agriculturalists in South Asia

    and also:

    http://dienekes.blogspot.com/2012/06/smbe-2012-abstracts-part-ii.html

    “Using an approximate Bayesian framework, we find that present patterns of genetic diversity in Central Asia may be best explained by a demographic history which combines long-term presence of some ethnic groups (Indo-Iranians) with a more recent admixed origin of other groups (Turco-Mongols). Interestingly, the results also provide indications that this region might have genetically influenced Western European populations, rather than vice versa. A further evaluation in MCMC-based Bayesian analyses of isolation-with-migration models confirms the different times of establishment of ethnic groups, and suggests gene flow into Central Asia from the east. The results from the approximate Bayesian and full Bayesian analyses are thus largely congruent. In conclusion, these analyses illustrate the power of Bayesian inference on genetic data and suggest that the high genetic diversity in Central Asia reflects both long-term presence and admixture in more recent historical times. ”

    In the same link as above:

    “Using logistic regression model we found that SLC24A5 functional SNP, rs1426654, is strongly associated with pigmentation in our sample and explains alone more than half of the skin colour difference between the light and the dark group of individuals. Conversely, the other tested SNPs fail to show any significance”

    Note that SLC24A5 is the general fixed “Caucasoid” light pigmentation locus, rather the more specific loci differentiating Northern from Southern Europeans.

    As for the “Northern” phenotypes going all the way to Mongolia, it seems that the area of U dominance extended in a very long “boreal” zone, not just from Spain to Eastern Europe, but from Eastern Europe to Lake Baikal as well:

    http://dienekes.blogspot.com/2005/12/mtdna-of-neolithic-siberians-from-lake.html
    http://dienekes.blogspot.com/2005/04/mtdna-of-lokomotiv-siberians-from-lake.html

    This area was initially mostly a destination rather than a source of human settlement, in which the early Proto-Europoid/mtDNA U bearing population was mostly replaced during the Neolithic and later.

  • http://dienekes.blogspot.com Dienekes

    A recent DNA study showed that Gotland had Finnish HG inhabitants at least 5000 years ago. At the same time a farmer on the nearby Swedish continent was Mediterranean.
    http://www.nature.com/news/ancient-swedish-farmer-came-from-the-mediterranean-1.10541

    Those hunter-gatherers were not “Finnish”; they resembled populations from Northeastern Europe (and hence Finns), but they spoke an unknown language, and were lacking in East Eurasian signals of ancestry which tie Finns together with their eastern linguistic brethren. Hence, in all likelihood they were not Finnish speakers.

    Personally, I can see an arrival of Finns at any time between the establishment of Comb Ceramic and the Seima Turbino phenomenon. Thankfully it’s cold enough there that we ought to be able ancient DNA from different periods/cultures and establish the sequence of events.

  • http://dienekes.blogspot.com Dienekes

    This ought to be of interest re: movements from Europe to Asia:

    http://dienekes.blogspot.com/2012/06/smbe-2012-abstracts-part-ii.html

    “Using an approximate Bayesian framework, we find that present patterns of genetic diversity in Central Asia may be best explained by a demographic history which combines long-term presence of some ethnic groups (Indo-Iranians) with a more recent admixed origin of other groups (Turco-Mongols). Interestingly, the results also provide indications that this region might have genetically influenced Western European populations, rather than vice versa.”

    I will also add that the region of mtDNA-U dominance did not stretch simply from Spain to Eastern Europe, but deep into Asia as well:

    http://dienekes.blogspot.com/2005/04/mtdna-of-lokomotiv-siberians-from-lake.html
    http://dienekes.blogspot.com/2005/12/mtdna-of-neolithic-siberians-from-lake.html

    It appears that this U-dominant population was largely replaced first by groups of Caucasoids from the south, and later (at the eastern end) by groups of Mongoloids from the East.

  • Justin Giancola

    “increase my own probability that the Finnic people have deep cultural roots in Northeast Europe.”

    that’s what I’m saying!

    3.”The Comb Ceramic people were partially Mongoloid”

    Have we considered that the similarities in phenotype arise from the conditions of the environment, rather than archetypal groups and admixtures? Not trying to overt snottiness but sometimes your positions can seem extreme to this degree.

    I have been to Finnic lands in Russia so I’m not talking out my butt toward the bias I’ve expressed.

  • Karl Zimmerman

    Razib,

    I’ve read that article before, and it makes a convincing case in some manners that the ultra-light European skin, along with light hair (and eyes presumably, although not discussed) could be a post-agricultural adaption.

    That said, while it’s plausible if you presume there was near-total replacement of the Northern European Mesolithic population by first-wave agriculturalists (or maybe the second wave if they also wiped out most of the first genetically), I cannot see how such these traits could become predominant throughout a mainly indigenous population.

    As an example, consider a farming group of fair people in northern Europe, next to swarthy hunter-gatherers. There is limited genetic exchange, and the hunter-gatherers pick up farming, along with some genes for blondism. However, as they switch to farming, except for the blonds in the population, they tend to suffer from rickets, reducing their fitness. However, the neighboring population of “pure stock” farmers is almost entirely blond, and more fit as an entire population. The “pure stock” farmers would thus have a population growth rate much higher than their neighbors until selection brought fair features to parity in the former hunter gatherers – meaning a model closer to population replacement than incorporation of a few advantageous traits into a mainly indigenous population.

    Of course, lactose tolerance fairly closely aligns with the Baltic in Europe, and must have been a post-agricultural mutation. It’s clear that the first people in Europe to develop lactose tolerance in Europe did not roll over all the remaining hunter-gatherers and ethnically replace them. So I suppose the northern European hunter-gatherers had a way of evening the score somewhat compared to groups that modern East Asians and Bantu displaced – probably adoption of a semi-pastoral lifestyle, as there could have been workarounds for lactose intolerance originally, and it would have allowed higher densities and lessened chance of rickets while the population adjusted to a diet nearly as plant rich as southern Europe.

  • Raimo Kangasniemi

    Re: “Personally, I can see an arrival of Finns at any time between the establishment of Comb Ceramic and the Seima Turbino phenomenon. Thankfully it’s cold enough there that we ought to be able ancient DNA from different periods/cultures and establish the sequence of events.”

    The soil is so acidic in Finland that human remains are very rare beyond Iron Age (starting about 300 BCE in Finland) when you start to get charred pieces of bone thanks to changes in funeral habits. A typical Finnish burial from Neolithic for example consists just of change of colour in the soil, showing that there was once a body – often covered in ochre – there. All organic material has gone.

    It’s a bit better in the Åland islands, but they were extremely sparsely settled before the arrival of colonists coming from current Sweden in about 400 CE, who completely replaced the probably extremely small previous population.

    So, not much chance of getting ancient DNA from Finland.

    When it comes to the Sami being more Finnish than the Finns, the reason probably lies in the fact that the Finnish tribes had, through their living areas including most of the Baltic sea coast of current Finland, far more outside contact. The very name of Sami is speculated to be a Baltic loan word, perhaps once sounding something like Zäme. One of the names of the original three Finnish tribes, the Tavastians(Hämäläiset in Finnish), is probably the same word. The gist is that it is speculated to mean “Inland” and to be a very old loan word, I’ve seen 4500 years mentioned. So, the coastal Finns had outside contacts, the inland Sami had mostly contacts with those Finnish populations who were the most inland and had thus smaller number of genetic inheritance derived from outside contacts.

    When it comes to Uralic languages and “arrival” of Finns, the archaeological and linguistic work done in recent decades in Finland has largely led to the conclusion that any “arrival” of Finns is a myth, the Finns coming into existence in situ through the last 10 000 years and that proto-Uralian languages could have been spoken in current Finland and neighbouring areas from the earliest of times.

  • Onur

    When it comes to Uralic languages and “arrival” of Finns, the archaeological and linguistic work done in recent decades in Finland has largely led to the conclusion that any “arrival” of Finns is a myth, the Finns coming into existence in situ through the last 10 000 years and that proto-Uralian languages could have been spoken in current Finland and neighbouring areas from the earliest of times.

    These are very bold claims. Can you substantiate them?

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    #17, he’s right. the reality is it turns out that finns were created by Väinämöinen!

  • Justin Giancola

    ^More like Inmar!

    And Kuaz reminds me of something curiously similar…

  • Onur

    #18,

    As far as I know, the current consensus in Finnic linguistics, and Uralic linguistics in general, favors younger ages, as Halvorson pointed out, and a Proto-Uralic urheimat in northwestern Siberia or the northern Ural Mountains area.

  • Dm

    The soil is so acidic in Finland that human remains are very rare beyond Iron Age
    Isn’t the soil acidity of peat bogs and conifer woods just as common in Sweden? What helped the preservation of the human remains there?

    Since Finland is but a border province of the semi-extinct past Finnic continuum spanning the Northern 2/3rds of today’s European Russia, the relevant DNA may be searched for in far larger territory than today’s Finland. Some of the areas there have alkaline soils, for example limestones of Kargopol Dryland, not a hundred miles from the areas populated by today’s Finnic Karelians and Veps.

  • Raimo Kangasniemi

    It’s not just peat bogs or woods, the reasons for the acidic soil in Finland are deeper and tied to the basic geology and geologic history which make the soil in Finland the most acidic in Europe. The soils in Sweden are not as acidic, and it’s geology tends to be more varied. As far as I recall, the few bog body finds in Sweden still come from the south of the country and are much younger than those on Denmark.

    I can’t say anything really on chances of finding ancient DNA tied to Finnic peoples from northern Russia. I would guess that it wouldn’t be terribly high on anyone’s list and would carry political baggage that might make it unwanted in Russia. The Russians want themselves neither reminded of the extent of past Finno-Ugric populations in Russia nor of the significant contributions from those populations to current Russians. The Russians have their own, Slavic, national project to protect. ;)

  • Raimo Kangasniemi

    I had a large post written about other points on the population history of Finland and the Uralic languages, submitted it, but it didn’t went through for some reason, yet when I tried to post it again the system claimed I was trying to post a duplicate.

    To repeat a few points from it:

    - The “arrival” of Finns to Finland at some point in the past is just part of a 19th century project to create a proper history for people who never had one before. We had to have a migration because everybody else in Europe tended to have a migration, and originally the “arrival” of Finns was tied to the migrations of the Germanic and Slavic peoples during and after the fall of the Western Roman Empire. It was not based on any evidence and no evidence have been found.
    - If there would have been a non- Finno Ugric language spoken in Finland before Finno-Ugric languages, then it is unlikely that it would have been totally replaced without leaving any traces when there’s no evidence of a population change, especially if it would have happened in a more recent past like the last 3000 years. At least few traces or anomalies would remain in some of the extant languages of our neighbourhood.

  • Raimo Kangasniemi

    To recover more briefly part of that lost post.

    Onur, part of the “arrival” story in Finland were supposed gaps in the occupation of Finland. Those gaps have been removed. The supposed migration of Finns to Finland receded ever farther into the past and finally had no place to where to seek refuge.

    There has been a continued occupation for the last 11 000 years. No abrupt changes in culture, just small groups of migrants now and then bringing new genes and new innovations, adding to the existing population.

    We are not talking of large populations here, the population of Finland when it came to be drawn to the Swedish sphere in the 12th century could have been as low as 20 000 and a very crude estimate of the population before Bronze Age is 3000-5000. The migrant groups in archaeological record would probably have been few hundreds at most at any given time and that is a high estimate.

    Based on current evidence, neither the Finns or the Sami appeared fully formed from somewhere else, but very slowly developed on site in Finland, not migrating from elsewhere, but emerging from existing populations which went through slow changes thanks to the small scale migration and cultural innovations. The coastal populations were probably always higher in number and succeeded in assimilating newcomers, the small number of Sami who adapted agriculture from their Finnish neighbors got assimilated in a very drawn out process.

    One has to also point out that Finnish and Sami identities didn’t really fully emerge until very late and that happened well during the span of recorded history in Finland. Both remained works in progress well into the 20th century.

  • Raimo Kangasniemi

    The “urheimat” of Uralic languages, the last time I came to face that discussion in Finland, was still supposed to be basically the area of current Tatarstan. The evidence being that only there you could find all the things described by the shared words of Uralic languages. I just point this out.

  • Halvorson

    @24

    Sami was spoken in Finland before Finnish and there is evidence of this in Finnish place names.

    See here
    http://mnytud.arts.klte.hu/onomural/kotetek/ou4/08aikio.pdf

    Sami itself derives around 25 percent of its vocabulary from an unknown source unrelated to any living language and which is probably Paleo-European. Some of these mystery words are common to all Sami languages and others only to a few, suggesting that they picked some up in Finland when the languages were still united and others as they plowed forward into Scandinavia. The last speakers of Paleo-European languages were probably culturally extinguished and absorbed in the years after the Sami entered Scandinavia c. 650 B.C.

    Finns are autosomally around 10 percent Siberian/East Asian. If that 10 percent was a leftover from some extremely distant mixture 10,000+ years in the past it wouldn’t show up in ADMIXTURE runs, but it does. That 10 percent plus the Middle Eastern 10 percent you see in ADMIXTURE is the lower bound for how Finnish ancestry is “non-native” but is almost certainly much lower than the real total. By the time the Finns got to Finland they’d had thousands of years to mix with Balts and other North Europeans they’d met on the journey from somewhere far to the east. It’s possible that modern Finns trace a majority of their ancestry to people who didn’t enter Finland until the post-Jesus era.

    The easiest to understand reason why the Finns aren’t native to Finland is that if they’d been there since the end of the Ice Age we should expect to see much more linguistic diversity than currently exists. Hunter gatherers can’t sustain a speech community over an area the size of Finland for very long without fragmenting into many other languages. On New Guinea there are something like a thousand different of these languages spoken; the Sami, who moved into Scandinavia relatively recently, have already split into 9. The fact that different dialects of only one language are spoken across such a large area of Finland (and that that language is so closely related to Estonian) is pretty good evidence that it hasn’t been around long enough for such a split to happen.

  • Justin Giancola

    25. Udmurts! Have you studied?

  • Lassi Hippeläinen

    @26:

    “Hunter gatherers can’t sustain a speech community over an area the size of Finland for very long without fragmenting into many other languages.”

    You obviously don’t understand anything about Finnish geography. It is very easy to travel around, because there are no mountains, but lots of rivers and lakes. It was quite common to go to hunting expeditions several hundreds of kilometers away from home. The HGs were the ones who needed and sustained a common language in a wide area.

    “the Sami, who moved into Scandinavia relatively recently”

    Citation, please.

  • Raimo Kangasniemi

    Halvorson, the Sami like the Finns for all we know based on archaeology have come to existence in the places where they now live and the areas they inhabited before. Certainly the Sami were spread on a far larger area in Finland in the past than they are now, and indeed there are hints of now extinct Sami dialects in some place names. But, the Sami seem to have been, in the area of current Finland, descendants of the people of who lived inland and Finns those of the more coastal populations.

    Some recent theories propose that the Sami are direct descendants of the first people to live on the coasts of modern Norway after they opened at the last stages of the last ice age. They could have followed the retreating ice on the opening coasts from the current western Europe, and there are some hints of connections to peoples on these coastal fringes of western Europe. When the ice would have retreated enough in the interior of Scandinavia, the ancestral population to Sami would have moved inland. There are some recent archaeological discoveries in northern Finland, up to 9000 years old, that could result of such population movement from the north, instead of the south.

    The problem with the theories of migration of people more or less fully formed is that it basically just removes the problem of how they came into existence in somewhere else, and in this case, to some unknown territory in Northern Russia. And we don’t have evidence of large scale migrations. We have evidence of movements of small number of people that, for all evidence, didn’t replace the pre-existing populations. If the Finns, Sami and the Finno-Ugric languages are late arrivals, it has been something of a stealth arrival.

  • Onur

    If the Finns, Sami and the Finno-Ugric languages are late arrivals, it has been something of a stealth arrival.

    It might have been a stealth arrival at least in the case of the arrivals of what are today geographically western Uralic languages such as Hungarian, Finnish, Estonian, etc. except Saami to their current locations, when we consider that current western Uralic speakers such as Hungarians, Estonians, Finns, etc. except Saamis have little genetic connection to the Uralic speakers of the likely Proto-Uralic urheimat areas. Only clear Uralic genetic markers in western Uralic speakers seem to be their minority Mongoloid autosomal components and at least part of the eastern haplogroups such as the Y-DNA haplogroup N they carry.

  • Karl Zimmerman

    Bridging the gap for a second -

    The lack of linguistic diversity of the Finns does not mean that the peoples themselves are not essentially indigenous. Finnic could have been part of a much more diverse branch of Uralic (of which all but Saami is now extinct). However, whatever lucky Finnic group managed to absorb agricultural knowledge (presumably from Indo-European groups, as the second millennium BCE is probably too recent for first-wavers) without either huge intermarriage or linguistic domination would have had a huge advantage over their cousins. They could have either rolled over them through language/culture or simple population replacement. IIRC, genetic studies of Finland suggest the current population has unusually low genetic diversity (suggesting growth from a very small base in the past), so I’d lean towards the latter, though comparisons with other Finnic groups like Estonians would be helpful.

  • Raimo Kangasniemi

    There are genetic connections between for example patrilinear lineages among Finnish and Sami and those of more eastern Uralic language speakers up to Samoyeds. When it comes to Hungarians, their limited genetic connections are explained by the fact that the Hungarians who migrated to Pannonia in 896 culturally assimilated a larger existing population, just like was the case with the Turkish in Anatolia.

    In the 10th to 12th centuries the archaeological evidence in Finland shows clearly the impact the Scandinavian expansion had to Finnish populations in the coastal areas: Abandonement of settlements along the Viking eastern sea route in the Gulf of Finland, increased hoards, more hilltop forts. And the new Swedish settlements on the coasts and the material evidence of this Viking era. The arrival of the Vikings and the Swedes was anything but stealth, and previous cases of “takeover” in the past should have left similar signs.

    I looked at few linguistic theories of the last years based on Halverson’s comments, and I think they are mainly revisionistic fancy. The theory of abandonment of Finland by existing population just before the onset of the Bronze Age and the “arrival” of the Finns is brought up in one, but although naturally the specifics of the linguistic theories are beyond me (I’m a historian by training, and not specialized in Finnish history) the claim on the population history seems curious at the least, like embers from a long since faded fire being used again. I don’t think the evidence stands.

    There’s little reason to expect the proto-Uralic population to have been very homogenous, especially if instead of a small urheimat the proto-Uralic developed on a vastly larger geographical area between the Ural mountains and the Baltic, if not the Atlantic.

    The genetic differences even inside the population of Finland itself – west/east divide – are greater than those between such Germanic peoples like the English and the Germans based on resent genetic studies. This itself speaks of a population with a long history, not of a relatively recent foundation, which would have led to a more homogenous population.

  • Halvorson

    Archaeology is useful Raimo, but you seem to be using it as some sort of trump card that has the power to overrule other findings in linguistics and genetics. In several recent historical disputes population genetics has proved the archaeological consensus to be completely false. Take for instance the co-existing TRB farming and Pitted Ware cultures of Neolithic Scandinavia. The archaeological consensus around 2008 was that the TRB people were native hunter gatherers who had culturally taken up farming without any outsiders moving in; the Pitted Ware people were former members of the TRB who had given up farming and reverted back to a foraging way of life. This was in sharp contrast to views held by early 20th century physical anthropologists that the TRB people were Mediterraneans who had invaded Scandinavia and that the Pitted Ware folk were Europe’s aboriginal inhabitants who had been hunting since the beginning of time.

    What genetics showed was that the modern population that the TRBers most closely resembled was Greeks. Pitted Ware hunters fell outside the variation of any modern population, but overwhelming carried mtDNA haplogroup U, which has been found in a majority of pre-Neolithic European skeletons and which is much rarer in today’s Europeans. (~12 percent).
    http://dienekes.blogspot.com/2012/04/ancient-dna-from-neolithic-sweden.html

    This is only one case, but on virtually every issue where genetics has been able to give us a definitive answer archaeologists have been wrong: for example, the LBK people weren’t acculturated natives or even a native-immigrant hybrid like archaeologists thought, they were just immigrants. On the most important question in European pre-history: Farming: Acculturation or Migration?, archaeologists have been wrong. At this stage it looks like the Europeans natives were replaced by migrants in all of southern and western Europe and yet archaeologists somehow didn’t notice this. This isn’t a minor point: it would be as if the Indo-European language family turned out to have been an invention and linguists had to start everything from scratch. For reasons I don’t understand, archaeologists have a strong preference for a history that’s as boring as possible: all modern populations have been living in their homelands since the beginning of time and nobody ever moved anywhere, in total contrast to the observed behavior of human beings since the invention of writing. Archaeology can used to supplement findings in other fields, but it can’t settle debates by itself.

    The recent fanciful linguistic papers you alluded to aren’t arguing against an old consensus that Finnish has been in Finland since the end of the Ice Age. The old consensus was that Uralic languages arrived with the huge Comb Ceramic culture c. 4000 B.C. and it’s that view that’s currently being challenged, with good reason. Uralic languages contain vocabulary borrowed from Indo-Iranian, which is definitely not a stone age language family (I’m getting this from location 1193 on my Kindle version of The Horse, the Wheel, and Language, sorry I can’t find the page).

    The genetic variation of the Finns isn’t that great of an argument for their antiquity: American whites are more genetically diverse than the British, but that doesn’t mean they’ve been around longer. Finns are distinct from each other because, like American whites, they arose from a mixture of genetically distinct populations: Siberians, Swedes, and (the largest element) northeast Europeans. Different Finnish populations contain different mixtures of these elements which accounts for the large regional variation.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    Finns are autosomally around 10 percent Siberian/East Asian. If that 10 percent was a leftover from some extremely distant mixture 10,000+ years in the past it wouldn’t show up in ADMIXTURE runs, but it does.

    this is not implausible, but not necessarily true. you shouldn’t speak so definitively if you don’t have much personal experience with ADMIXTURE, i do. if the ancestral reference population isn’t found in your panel of reference populations then the “nearest closest” populations will shake out. the “nearness” here can be rather distant. for example, south asians can come out as 20-30% east asian in admixture runs. but the last common ancestor with east asians was on the order of 20-30 thousand years ago. this isn’t to dismiss your concern, but it doesn’t have quite the definitive force you seem to think it does (if you respond to me at length, you better bring some serious arguments about ADMIXTURE, because i’ve run the program hundreds of times at this point and do have some intuition).

    On New Guinea there are something like a thousand different of these languages spoken;

    new guinea is characterized by gardening agriculture. it is not HG. you should use australia as your example. this sort of mistake reduces the plausibility of your whole argument! (i.e., if i can pick our errors in what i do know, i wonder if you make errors in what i presume you know)

    If the Finns, Sami and the Finno-Ugric languages are late arrivals, it has been something of a stealth arrival.

    archaeology has a very mixed record, to say the least, in supporting or rejecting hypotheses of migration.

    IIRC, genetic studies of Finland suggest the current population has unusually low genetic diversity (suggesting growth from a very small base in the past)

    yes. the samples from eastern and inland finland show evidence of radical recent population bottleneck (probably in historical time).

    When it comes to Hungarians, their limited genetic connections are explained by the fact that the Hungarians who migrated to Pannonia in 896 culturally assimilated a larger existing population, just like was the case with the Turkish in Anatolia.

    1) medieval ancient DNA does show connection between hungarians and eastern populations. some hypothesize that the hungarian nobility was decimated by mongols and turks in turn, reducing that footprint

    2) turkish anatolia has a non-trivial mongoloid element, on the order of 10%. that’s between the ‘eastern’ element in finns and samis. it was assimilation, but the genetic impact was significant.

    The genetic differences even inside the population of Finland itself – west/east divide – are greater than those between such Germanic peoples like the English and the Germans based on resent genetic studies.

    which studies? i’m pretty sure you are confusing long branches/greater phylogenetic distance due to massive genetic drift in eastern finland with genuine high diversity. but i’m willing to be convinced.

    In several recent historical disputes population genetics has proved the archaeological consensus to be completely false.

    yes, this too shifts my priors a bit.

    i am enjoying this discussion. but everyone: please don’t try to bullshit me on the genetics. i know this area quite well and don’t take well to attempts to bluff me if i call you out on something. also, more citations would be helpful for those of us not familiar with the field to follow up.

    thanks.

  • Onur

    2) turkish anatolia has a non-trivial mongoloid element, on the order of 10%. that’s between the ‘eastern’ element in finns and samis. it was assimilation, but the genetic impact was significant.

    The Mongoloid element average in Anatolian Turks (thus excluding Balkan Turks, Aegean Islander Turks and Cypriot Turks) seems to be 6 or 7%, thus more or less the same as the Mongoloid element average of the Adyghe and Finns. Saamis, on the other hand, seem to have a Mongoloid element average that is more than double that of the Adyghe, Anatolian Turks and Finns.

  • Raimo Kangasniemi

    When it comes to bottleneck in eastern Finland etc, yes, that’s pretty much happened during historical times and is tied to population expansion during the Swedish era. The population of Finland was estimated to be 20 000 – 50 000 in the 12th century and in the first Swedish census in the 1740s it was somewhat over 400 000. The famine of 1696-97 which struck especially hard the eastern regions, killing perhaps a third of the population of the time in the whole area of Finland, would have played a part too.

    There was also some movement from the Karelians living under Novgorod to the mainly Sami inhabited areas of Bothnia and Savonia when those areas were officially under Novgorod’s rule after 1323. This latter is largely a forgotten episode in the population history of Finland.

    When it comes to Finnish vs English and Germans, it’s a study of individuals from four years ago. This is not the best article about it, but there’s clear difference between eastern and western Finns, when the English and Germans pretty much cluster together in the visualization at the end of the article:

    http://ki.se/ki/jsp/polopoly.jsp?l=en&d=130&a=63734

    When it comes to Hungarian nobility etc, the nobles tend to be more cosmopolitan the population at large, from the very beginning of the Hungarian kingdom. Even just looking at the ruling Hungarian royal families shows that. There’s also the case that by 1800 the Hungarian towns had populations to a significant degree populated by if not fully ethnically, then linguistically German. A decimation of a supposedly “pure” Hungarian nobility by Turks and Mongols is a suspiciously nationalistic explanation.

  • Raimo Kangasniemi

    Halvorson, archaeologists are perhaps supporters of the more boring explanations, because they had learned through the hard way that the less boring and more “heroic” explanations perhaps had less to do with facts at hand than with using myths as explanations (like the population history of Greece, connecting the destruction of the Mycenean civilization to the movement of Dorians to Greece and using Greek myths as a support for that). It’s like a bit how the “catastrophe” theories fell out of favour in geology (and have a bit resurrected in recent decades, like as explanations for the extinction of the dinosaurs).

    The problem with some of the linguistic theories that I have as a historian is with how closely they seem to follow some old theories in history. In history younger historians sometimes have a habit of going back to old theories and resurrecting them in a new form as a part of deliberate act of revisionism, a kind of challenge. Like in recent years the flood of more positive portayals of crusades among the young Anglo-Saxon historians. What I got from the linguistic theories is a feeling of this kind of deliberate revisionism, especially as for example in Uralic language studies the time for the proto-Uralic have been all around the place, not just 4000 BCE, but ranging from 7000-6000 BCE to 2000 BCE. To an outsider it seems to be very hard to pin the changes in the languages down that well and then the dates for the Indo-European languages – to whom the study of Uralic languages seem to be wedded – are greatly different between scholars. That’s why I prefer archaeology. But then I do understand it much better than linguistics too.

  • https://plus.google.com/109962494182694679780/posts Razib Khan

    The Mongoloid element average in Anatolian Turks (thus excluding Balkan Turks, Aegean Islander Turks and Cypriot Turks) seems to be 6 or 7%, thus more or less the same as the Mongoloid element average of the Adyghe and Finns

    1) that’s why i said on the order of. i’m not totally confident in differentiating between 5 and 10 percent with the representativeness we have.

    2) re: finns, are you sure that the ‘mongoloid’ element is analogous to that in turks? ie, sometimes an ‘eastern’ element in a europe context is actually more ‘western’ as mongoloid sub-elements go.

    . The population of Finland was estimated to be 20 000 – 50 000 in the 12th century and in the first Swedish census in the 1740s it was somewhat over 400 000. The famine of 1696-97 which struck especially hard the eastern regions, killing perhaps a third of the population of the time in the whole area of Finland, would have played a part too.

    much of this is not relevant to what i’m talking about. the census sizes are too large. rather, what is more plausible is that you have REALLY SMALL FOUNDING populations for colonization events. i’m talking on the order of dozens to show up in a statistica l genetic sense.

    When it comes to Finnish vs English and Germans, it’s a study of individuals from four years ago. This is not the best article about it, but there’s clear difference between eastern and western Finns, when the English and Germans pretty much cluster together in the visualization at the end of the article:

    i have blogged the article you link to, so i’m familiar with it. again, my main issue, which i pointed out to you earlier, is that population bottlenecks can distort genetic distance measures. to be colloquial, if you push a population through a bottleneck it can “evolve very fast” and the genetic measure you’re using is going to give you a big distance. this distance is real, but it’s not an accurate measure of the historical processes you’re trying to infer.

    A decimation of a supposedly“pure” Hungarian nobility by Turks and Mongols is a suspiciously nationalistic explanation.

    i didn’t say anything about ‘pure.’ if you put words into my mouth again i’ll ban you with no further warning. first rule of thumb with discussing things with me on my blog: argue with what i say, not your interpretation of what i say. second, you may or may not believe it, but to reiterate: ancient DNA from then nobles of medieval hungary show an ‘eastern’ genetic affinity which is almost impossible to find in modern hungary. this may be due to some nationalistic bias, i have no idea. but be clear what you’re implying. we’re not talking interpretations of old texts.

  • Onur

    2) re: finns, are you sure that the ‘mongoloid’ element is analogous to that in turks? ie, sometimes an ‘eastern’ element in a europe context is actually more ‘western’ as mongoloid sub-elements go.

    By Mongoloid element, I am referring to an undifferentiated Mongoloid component (i.e., the only Mongoloid component in its K level) and making my average Mongoloid element comparisons according to that. So the effects of the different Mongoloid coefficients of different Mongoloid sub-elements is not an issue.

    Having said that, the Mongoloid element average in the Adyghe seems to be a bit lower than that of Anatolian Turks. But Finns (and also Vologda Russians) do not seem to have lower Mongoloid element average than Anatolian Turks have, so Anatolian Turks might on average be less Mongoloid-admixed than Finns (and also Vologda Russians).

  • Raimo Kangasniemi

    First of all, when it comes to the “pure” Hungarian nobility, I didn’t take the whole claim about Hungarian nobility and the Turks and the Mongols as your own personal claim, but something you passed on from a source. If I was wrong on that, I was wrong.

    When it comes to loss of some genetic materials among Hungarian aristocracy, aristocratic families tend to be short-lived and need to be replenished from below or abroad. 95 percent of current European aristocratic families were not “noble” in the male line before 1600. We can look at it straight from the fate of the Arpad dynasty itself and it’s replacement with Italo-French, German, Polish-Lithuanian and Austrian dynasties with only marginal links to the Arpads at the end. Already the second king of Hungary was the son of the doge of Venice.

    Upholding cultural and class identity over time doesn’t have to mean upholding biological unity.

  • Halvorson

    I don’t know anything about ADMIXTURE other than what I’ve scavenged from the Dienekes-sphere, but I think you can make a decent non-technical argument that the Siberian component in Finns isn’t a holdover from an Asian influenced Paleo-European group. Given the really short divergence dates (22,000 years) some papers have given for the European-East Asian split, it makes sense that old Europeans might be tilted genetically in an Asian direction. But if the North European component in ADMIXTURE really is Asian-influenced hunter gatherer, shouldn’t you see roughly the same Siberian appendage in populations with the same amount of N. Euro? For example, looking through one of Dienekes’ K12 spreadsheets, Finns are roughly 75% N. Euro, with Lithuanians slightly edging them out with around 77%. But the Lithuanians have zero Siberian compared to the Finns’ 6.7 percent. Isn’t the more parsimonious explanation here that Finns probably received that 6.7 percent dose with the arrival of Uralic peoples?

  • Raimo Kangasniemi

    If you connect that 6.7 percent to Uralic languages, the problem is why the Lithuanians – living so close on to Uralic language speakers (not just Estonia, Latvia being to a significant degree Uralic speaking until the Crusades and still having a lingering Livonian presence) – would not have gotten any Siberian genetic heritage from their Uralic neighbours?

  • Halvorson

    It does seem strange at first, but there’s a similar situation going on with the Basques: normal French samples are around 8 percent West Asian, but French Basques only score 0.1 percent of the same component. It’s apparently possible for geographically close populations to live in near total isolation from each other for quite a while.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »