Amerindians have both Mongoloid and Caucasoid physical features

It’s amazing that so many people were so credulous and so stupid that they actually believed David Reich et al. when they proclaimed that Nordics are Caucasoid-Mongoloid hybrids. All of the big names in the “HBD community” were among the duped. Witness the stupidity on display in the following posts from last year. The titles alone are incredibly stupid:

There’s Razib Khan’s post “Across the sea of grass: how Northern Europeans got to be ~10% Northeast Asian“. In this post Razib Khan tells all of his Nordic readers that “You’re Asian. Yes, you!”. He speculates in bold that “perhaps the Indo-Europeans were mongrels!”. He exclaims in bold that “it turns out that white people were always 10% non-white! We just didn’t know”.

Dienekes is a Mediterraneanist and was therefore thrilled to hear the claim by Reich et al. that Nordics are mongrels. He did a whole series of posts on the supposed racial impurity of Nordics, all of them chock-full of stupidity:

Hints of East/Central Asian admixture in Northern Europe

Estimating admixture proportions and dates with ADMIXTOOLS (Patterson et al. 2012)

The tangled web of humanity

ADMIXTURE tracks Amerindian-like admixture in northern Europe

Then there’s Greg Cochran’s post entitled “Injun-Europeans“. In that post Cochran says that “The French (and the Germans, and the Dutch and Irish) are closer to pure-blooded South American Indians than Italians are”. Reich et al. told him that was because Nordics are part Amerindian, and he believed them. It never occurred to him what an absolutely absurd notion that is. Later he observes that “Amerindians have the East Asian skin color genes, along with some of their own – as far as I know, you don’t see these at all in Europeans”, and that “there is zero overlap between Amerindian mtDNA haplotypes and those found in ye olde hunter-gatherers in Europe”. Doesn’t quite add up, does it Cochran? Then he notes the presence of the European mitochondrial haplogroup X and European Y-chromosome haplogroups in Amerindians, but fails to draw the obvious conclusion. Toward the end Cochran repeats the slur made by his good buddy Razib Khan, saying “maybe the Indo-Europeans were a mixed people”.

There are also these posts by Davidski:

They had blond hair and light eyes, and came from the north…but they were racially impure

Europeans are a three-way mix, with a whopping 20-40% ancient North Asian ancestry

Later, when a paper came out showing that the Mongoloid EDAR mutation is completely absent in European populations, Davidski began to dimly perceive that something wasn’t quite right about the picture that Reich et al. had painted, and posted “Ancient Amerindian-like admixture in Europe – something doesn’t add up“. He still didn’t get what was going on though, so the post contains a lot of convoluted nonsense attempting to explain the facts.

What makes the credulity and stupidity of all these people so amazing is that all anyone has to do to see that the claims made by Reich et al. are false is to simply look at the people they are making claims about.

There is nothing at all about the physical features of Nordics that is even remotely Mongoloid. Indeed, Nordics are the most ultra-Caucasoid of all Caucasoids. Their phenotype is more derived, more divergent from that of their Veddoid ancestors, than that of all other Caucasoids.

On the other hand, it’s obvious that Amerindians don’t look like pure Mongoloids. It’s obvious that they all look like Mongoloid-Caucasoid hybrids.

Though simply looking at people is sufficient to expose the falsity of the claims made by Reich et al., there are plenty of physical anthropological data that expose it even further.

An important paper containing such data is “Observations on the face and teeth of the North American Indians“, published in 1931 in the Anthropological Papers of The American Museum of Natural History. This paper contains qualitative data on the physical features of full-blood Indians, mixed-blood Indians, Eskimos, American Whites, North Europeans, South Europeans, Hawaiians, Asiatics, Mexicans, and Negroes. The eight physical features studied were intercanthus distance (distance between the corners of the eyes), nasal bridge, the epicanthic eyefold, the palpebral fissure (the opening between the eyelids), the enamel rim, Carabelli’s cusps, the number of molar cusps, and skin color.

The most important pattern that emerges from the data is that Indians are intermediate between Whites and Asiatics in the frequencies of the traits studied.

Eskimos are much more Mongoloid than Indians, and Indians have higher frequencies of narrow intercanthus distance than Eskimos. Full-blood Indians have lower frequencies of narrow intercanthus distance than mixed-blood Indians.

Whites have high frequencies of high nasal bridge, Asiatics have low frequencies of high nasal bridge, and Indians have intermediate frequencies of high nasal bridge.

Asiatics have about a 100% frequency of the epicanthic eyefold, Europeans have about a 0% frequency of the epicanthic eyefold, and Indians have intermediate frequencies of the epicanthic eyefold.

The enamel rim is a feature of the shoveled incisors characteristic of Sinodonty. Sinodonty is one of the pleiotropic effects of the Mongoloid EDAR mutation. Asiatics have high frequencies of the enamel rim, Whites have low frequencies of the enamel rim, and Indians have intermediate frequencies of the enamel rim.

If the claims made by Reich et al. were true, then we would expect the trait frequencies of Asiatics and Indians to be the same and at one extreme, the trait frequencies of South Europeans to be at the other extreme, and the trait frequencies of North Europeans to be intermediate between those of Asiatics and Indians and those of South Europeans. But of course the claims made by Reich et al. aren’t true, and of course that is not at all the pattern that we observe. What we observe is that the trait frequencies of Asiatics are at one extreme, the trait frequencies of North Europeans and South Europeans are at the other extreme, and the trait frequencies of Indians are intermediate between those of Asiatics and those of Europeans. And in fact, the trait frequencies of South Europeans are slightly closer to those of Asiatics than the trait frequencies of North Europeans are. South Europeans have a slightly lower frequency of high nasal bridge than North Europeans, and a slightly higher frequency of the epicanthic eyefold than North Europeans.

A second pattern that emerges is that some groups of Indians have trait frequencies closer to those of Whites than other groups of Indians, across multiple traits. The final remark of the paper is “One curious relation observed is that if the European is taken as the basis of comparison, the village tribes are nearer the whites than are the other Indians, or perhaps one should say the Mexicans and mixed-Indians”. The “village tribes” are the Pueblo Indians of New Mexico and Arizona. Compared to the other Indians, the Pueblo Indians tend to have lower frequencies of wide intercanthus distance, higher frequencies of high nasal bridge, lower frequencies of the epicanthic eyefold, lower frequencies of the enamel rim, and higher frequencies of light skin color.

Among the Pueblo Indians, the Zuni tribe in particular stands out from the rest. Of all the Indian tribes, the Zuni have the lowest frequency of wide intercanthus distance, the highest frequency of high nasal bridge, the lowest frequency of the epicanthic eyefold, and among the highest frequencies of light skin color.

The following photographs of Zuni Indians show that many of them have a very Caucasoid appearance. (Click to enlarge.)

zuni-1

zuni-2

zuni-3

zuni-4

zuni-5

zuni-6

zuni-7

zuni-8

zuni-9

zuni-10

zuni-11

zuni-12

zuni-13

zuni-14

zuni-15

zuni-16

In the photograph of a group of Zuni Indians below, the boy on the right in the foreground and the girl at the upper-right look especially Caucasoid.

zuni-17

And in the photograph of Zuni Indians below, the boy and the man holding the boy look especially Caucasoid.

zuni-18

Given that the Zuni tribe stands out from the other Pueblo Indian tribes as being especially Caucasoid, it is interesting that, unlike the languages of the other Pueblo Indian tribes, the Zuni language does not belong to any language family, but is a language isolate.

After the Pueblo Indians, the tribe with the trait frequencies closest to those of Whites is the Navajo tribe. Based on the findings of the K = 17 admixture analysis of Amerindians and the K = 16 admixture analysis of Mongoloids, this is likely the result of a greater amount of Y C3-M217 proto-Caucasoid admixture in the Navajo.

A third pattern in the data is that while Indians have trait frequencies that are intermediate between those of Whites and those of Asiatics, Hawaiians have trait frequencies that are intermediate between those of Whites and those of Indians. This pattern is seen for frequencies of high nasal bridge, the epicanthic eyefold, and the enamel rim. For Carabelli’s cusps, Asiatics and Indians have frequencies that are about the same, while Hawaiians have a frequency that is intermediate between the frequencies for Asiatics and Indians and the frequency for Whites. These observations are due in part to Hawaiians being primarily Polynesoid and not Mongoloid, but they are also due in part to Caucasoid admixture in Hawaiians.

In the White Gods post I pointed out that the Wikipedia article on mummies mentions the Andeans, the Egyptians, and the Guanches as the only three ancient peoples that practiced artificial mummification. But in his 1971 article “The Bearded Gods Speak”, Thor Heyerdahl noted a fourth ancient people that practiced artificial mummification: the Polynesians. Here is an excerpt from the article:

This means that modern archaeologists have direct evidence of the remarkable fact that true mummification was practiced by the very founders of the earliest pre-Inca civilization in Peru. In fact, true mummification—with evisceration through the anus and rubbing with resinous and oily preservatives—was common both to Peru and adjacent Polynesia, while it was totally unknown in Indonesia. But whereas hundreds of actual mummies are still available from the desert region of Peru, we have mainly the written records of early voyagers to attest to the wide distribution of the practice of mummifying royal persons throughout the far-flung islands of Polynesia—from Easter Island in the East, to Hawaii in the north, to New Zealand in the southwest. The very widespread occurrence of this elaborate practice in a tropical island area whose damp climate prevents lasting success shows that it must have spread from a common cultural source outside the island area. Since mummification cannot have reached the islands from Southeast Asia, it is all the more noteworthy that two most elaborate royal mummy-bundles recently brought from a cave in Hawaii to the Bishop Museum in Honolulu correspond in striking detail with the sophisticated mummy-bundles of the pre-Inca Tiahuanaco culture.

You can read the entire article, and see a photograph of a Hawaiian mummy casket, on this page.

The K = 5 craniometric admixture analysis showed that the Polynesoid component is modal for the skulls from the Mokapu Peninsula of the Hawaiian island of Oahu, and that many of those skulls have significant amounts of the Mongoloid component, but it also showed that many of those skulls have significant amounts of the Caucasoid component. The Caucasoid component also appears in the other two corners of Polynesia, in some of the skulls of the Māori of New Zealand and the Moriori of the Chatham Islands, and in a couple of the skulls of the Easter Islanders.

A fourth pattern in the data is that the frequencies of several traits that characterize Mongoloids decrease with age, while the frequencies of the opposite traits, which characterize Caucasoids, increase with age. Specifically, the frequencies of wide intercanthus distance, flat nasal bridge, and the epicanthic eyefold, all characteristic of Mongoloids, decrease with age. The frequencies of narrow intercanthus distance, high nasal bridge, and the absence of the epicanthic eyefold, all characteristic of Caucasoids, increase with age.

This confirms the observation that has been made by many anthropologists many times before, that Mongoloids are the most pedomorphic of the three major races. Negroids are the second most pedomorphic, and Caucasoids are the least pedomorphic. Among Caucasoids, Mediterraneans are more pedomorphic than Nordics. Veddoids are more pedomorphic than Mediterraneans.

A fifth pattern in the data is that the frequencies of wide intercanthus distance and flat nasal bridge, both characteristic of Mongoloids, tend to be higher in females than in males. This is consistent with the idea that Mongoloids are the most feminized of the three major races.

Both Negroids and Mongoloids have low nasal bridges, while a high nasal bridge is a characteristically Caucasoid trait. Some people with a high nasal bridge are said to have an aquiline nose. According to the Wikipedia article on aquiline noses, Samuel George Morton wrote in his 1839 work Crania Americana that the aquiline nose is primarily found among Mediterraneans, Middle Easterners, and North Africans. The article also says that William Z. Ripley wrote in his 1899 work The Races of Europe that an aquiline nose is a characteristic of the Teutonic (Nordic) race. It’s clear, then, that an aquiline nose is a characteristically Caucasoid trait. On the Wikipedia page you can see a large gallery of famous Caucasoids with aquiline noses. There are no non-Caucasoids in that gallery. But as the article notes, Morton also wrote in Crania Americana that Amerindians are “marked by a brown complexion; long, black, lank hair; and deficient beard. The eyes are black and deep set, the brow low, the cheekbones high, the nose large and aquiline, the mouth large, and the lips tumid and compressed”. The aquiline nose of Amerindians is the strongest sign of their obvious Caucasoid admixture.

Posted in Uncategorized
52 comments on “Amerindians have both Mongoloid and Caucasoid physical features
  1. Jackson Devoni says:

    Very interesting reading. Would you ever be able to run my 23andme data through an ADMIXTURE calculator to see my proportions of ancient Caucasoid and Mongoloid ancestry? I am 25% Italian, 25% Finnish and 50% Irish/British. I think my Finnish side may harbour some distant Mongoloid ancestry.

  2. genetiker says:

    I would just use Dienekes’ globe13 calculator. The ADMIXTURE analysis on which it’s based shows Finns as having ~6% of the Mongoloid Siberian component and ~1% of the Mongoloid Arctic component. It would be able to detect the same components in your data.

  3. Jackson Devoni says:

    Okay well this is easy as I am already a member of the Dodecad Project so Dienekes ran my results for me when he did the original globe13 analysis. I am DOD018 and I score 1.6% in the Siberian component and 0.9% in the Arctic component. I don’t know if these results are noise or real though since they are so small. I mean 1.6% Siberian is not very much. What do you think?

  4. genetiker says:

    I think that ADMIXTURE and DIYDodecad give accurate estimates of even very small levels of admixture. I think that very little of what they detect is noise.

    Since you’re 25% Finnish, you would be expected to have ~1.5% of the Siberian component and ~0.25% of the Arctic component. Your admixture estimates are in the ballpark of what would be expected.

  5. Jackson Devoni says:

    And here are my Dodecad globe4 ADMIXTURE results in which Dienekes thought he could track the Amerindian/Northeast Asian results that Patterson found in their study using the f4 analysis.

    90.08% European
    1.13% Asian
    0.22% African
    8.57% Amerindian

    As you can see I score 8.57% in the Amerindian component although I have no known Amerindian ancestry. As I mentioned above I am 25% Italian, 25% Finnish and 50% Irish/British by known ancestry. So does my 8.57% Amerindian score from the Dodecad globe4 calculator above just indicate very ancient shared ancestry with Amerindian peoples?

  6. genetiker says:

    You’re actually 98.65% Caucasoid. The 8.57% of your genome that ADMIXTURE is identifying as Amerindian is actually Nordic Caucasoid.

    The Amerindian component in the K = 4 ADMIXTURE analysis is a composite of Mongoloid and Nordic Caucasoid genetic elements. As a result Nordics show up as having some of the Amerindian component.

    In other words, the Amerindian admixture that Nordics appear to have in ADMIXTURE analyses (and in other analyses) is spurious.

  7. Jackson Devoni says:

    Okay so this 8.57% Amerindian/Nordic Caucasoid score that I get in the globe4 analysis comes form a very ancient Northern European population that shares some ancestry with present day Amerindians then?

  8. genetiker says:

    The 8.57% is just Nordic. There’s no reason to think it’s any different from the rest of your Nordic genetic material that happens not to be identified as Amerindian by ADMIXTURE.

    Nordics crossed the Atlantic long before the Vikings and contributed part of the genetic material of present-day Amerindians. Part of this genetic material is from the Solutreans and part of it is from the White Gods. And there may have been other pre-Viking Nordic waves besides those two.

  9. Jackson Devoni says:

    Okay and so back to the Dodecad globe13 analysis then. What type of old genetic ancestry does the North_European component represent do you think?

  10. genetiker says:

    The “North European” component is the Nordic component.

    Races are biological entities. The whole idea of referring to biological entities with geographical terminology is idiotic. It’s enforced by political correctness because political correctness denies the existence of races.

    Nordics are Nordics whether they’re in Northern Europe or anywhere else on earth. And there are now a lot of people in Northern Europe who aren’t Nordics.

    So for example, the components in a K = 3 ADMIXTURE analysis should be called Caucasoid, Mongoloid, and Negroid. But Dienekes calls them “European”, “Asian”, and “African”, because Dienekes conforms to the idiotic demands of political correctness.

  11. Jackson Devoni says:

    Okay thanks yet again. I am finding this conversation very interesting actually. Here are my Dodecad globe13 results below.

    Siberian: 1.6%
    Amerindian: 0.8%
    West_African: 0%
    Palaeo_African: 0%
    Southwest_Asian: 6.4%
    East_Asian: 0.2%
    Mediterranean: 28.1%
    Australasian: 0.5%
    Arctic: 0.9%
    West_Asian: 10.6%
    North_European: 50.8%
    South_Asian: 0.1%
    East_African: 0%

    So based on these results then the majority of my genes are Nordic correct?

  12. Jackson Devoni says:

    And what would you call the Mediterranean and West_Asian components from this globe13 analysis? Like you call North_European Nordic what would you call these other two components?

  13. genetiker says:

    The “Mediterranean” component corresponds to the Mediterranean race. In this case the geographical term and the biological term happen to coincide. There was a massive influx of Mediterraneans from the Near East into Europe during the Neolithic, and this influx was primarily associated with the E1b1b Y haplogroup.

    The “West Asian” component corresponds to the race that has been called the Alpine or Alpinid race in Europe and the Assyroid or Armenid race in the Middle East. It’s the same race in both places, but there’s a European variety and a Middle Eastern variety. This race entered Europe from Asia Minor during the Neolithic. It’s associated primarily with the G Y haplogroup, but also with the J2 Y haplogroup.

  14. Jackson Devoni says:

    Okay that does make some sense. And when do you think the North_European Nordic component entered Europe?

  15. genetiker says:

    Nordics never entered Europe. They’re indigenous to Europe. They evolved from Mediterraneans in Northern Europe between 40,000 and 25,000 years ago.

  16. Jackson Devoni says:

    So the North_European component represents autosomal genetic ancestry inherited from the Upper Paleolithic and Mesolithic hunter-gatherers of Europe right?

  17. Jackson Devoni says:

    Polako and Dienekes seem to think that the North_European component is a composite mixed component made up of European Mesolithic hunter-gatherer alleles and some early Neolithic farmer alleles. What do you think?

  18. genetiker says:

    Nonsense. Nordics were the native hunter-gatherers and Mediterraneans and Alpines were the immigrant farmers. ADMIXTURE has no trouble identifying the genetic signatures of these different races. The components that emerge from the analysis correspond exactly to the races that had been identified by physical anthropologists over 100 years ago.

  19. Jackson Devoni says:

    Very interesting stuff for sure. So then basically the North_European component descends pretty much fully from the aboriginal Mesolithic hunter-gatherers of Europe and both the Mediterranean and West_Asian components came later to Europe totally with waves of Neolithic farmers originating in the Near East correct?

  20. genetiker says:

    Basically correct. There was some Mediterranean presence in Europe prior to the Neolithic, but it was much smaller than what it became in the Neolithic, and it was mainly in Southeastern Europe, in the Balkan peninsula. It’s doubtful whether it extended to Italy. It didn’t extend to Iberia. Today Iberia is the most racially Mediterranean place in Europe behind Sardinia, but prior to the Neolithic Iberia was Nordic. It was flooded with Mediterraneans from North Africa during the Neolithic.

    Nordics evolved in Northern Europe, and that was their original homeland, but when the Last Glacial Maximum set in around 25,000 years ago, much of Northern Europe became too cold to inhabit, and the Nordics went south to Southern Europe. The Nordics who went south into Iberia continued further south into North Africa. Prior to the Neolithic, all of North Africa, from Morocco to Egypt, was Nordic.

    It was during the Last Glacial Maximum that the Nordic Solutreans in Iberia and France crossed the Atlantic and populated the Americas.

  21. Jackson Devoni says:

    Okay well that would explain then why the ancient Mesolithic samples that were recovered from La Brana in Spain come out to be majority North_European in the Dodecad globe13 analysis.

    So then there may have been some small presence of the Mediterranean component in Southeastern Europe during the Mesolithic? But it spread out mostly later during the Neolithic with farming from both the Balkans and the Near East?

  22. genetiker says:

    The Mediterraneans who were in the Balkans prior to the Neolithic didn’t play any significant part in what happened in the Neolithic. Agriculture began in the Near East 2000 years before it started in the Balkans, and once it got underway the Mediterranean population of the Near East exploded. Waves of Mediterraneans from the Near East then spread out over Nordic North Africa and Nordic Europe.

    And yes, the La Brana samples confirm what I’m saying.

  23. Jackson Devoni says:

    If the North_European component is of 100% European Mesolithic hunter-gatherer background and the West_Asian component is of 100% Neolithic West Asian/Near Eastern background then how are they or how can they be so close to each other when it comes to FST distances?

  24. genetiker says:

    Alpines inhabited the parts of the Middle East closest to Europe, Asia Minor and Transcaucasia. There was gene flow between Nordic hunter-gatherers and Alpine hunter-gatherers during the Paleolithic and Mesolithic.

  25. Jackson Devoni says:

    Ahhh I see okay then. But the West_Asian component itself then only made it to Europe during the Neolithic and later correct?

  26. genetiker says:

    Based on all of the available evidence (that I’m aware of), that is correct.

  27. Jackson Devoni says:

    Okay I think one of the main reasons that Dienekes and Davidski think the North_European component is a hybrid component made up of Mesolithic hunter-gatherer and early Neolithic farmer alleles is because the La Brana Mesolithic hunter-gatherers and Swedish Neolithic hunter-gatherer samples/specimens that have been tested for autosomal DNA have been found by the scientists who tested them initially to be outside of the range of modern variation. On the other hand the Neolithic farmer remains that have been tested such as GOK4 and Oetzi have been found to be within the modern range of variation. What do you think about this?

  28. genetiker says:

    None of that makes any sense.

    First, ADMIXTURE doesn’t produce “hybrid components” for admixture events that took place in the past few thousand years, as is the case with Nordics and Mediterraneans. That’s the whole point of ADMIXTURE: to resolve the elements that mixed.

    Second, Europeans are more mixed today than they were thousands of years ago, so a mixed sample would be more likely to fall within the range of modern variation, not less.

    Third, of course the La Brana hunter-gatherers would fall further outside the range of modern variation than the others: they’re 2000 years older.

    Fourth, of course the Swedish hunter-gatherers would fall further outside the range of modern variation than the Swedish farmer: the Swedish farmer is more mixed than the Swedish hunter-gatherers.

    Fifth, the claim that Oetzi is within the modern range of variation is false.

    Sixth, if there’s a PCA plot that includes the La Brana hunter-gatherers, the Swedish hunter-gatherers, the Swedish farmer, and Oetzi, I haven’t seen it. I don’t know how anyone makes valid comparisons between all of them without one.

  29. Jackson Devoni says:

    Have you seen the work that Vadim Verenich has done with these ancient samples? He apparently found an ADMIXTURE component which he calls North-European-Mesolithic that was modal in the La Brana samples ad today only found in high frequencies in the Sami people of Scandinavia and then Finns. It is only found in very small percentages on all other European populations. The La Brana samples score I would say over 80% in this component. It seems very different from the North_European component that Dienekes and Davidski have found in their different ADMIXTURE analysis. In Vadim Verenich’s MDLP-World 22 analysis the component that matches the North_European component found by both Dienekes and Davidski in their various analysis is called North-East-European and is quite different from the North-European-Mesolithic component which dominates the La Brana and Swedish hunter-gatherer samples. They hardly show and of the North-East-European component. Here is a link to that analysis below. Thoughts?

    http://magnusducatus.blogspot.ca/2012/09/behind-curtains-mdlp-world-22-showcase.html

  30. genetiker says:

    The “North-European-Mesolithic” component is the Cro-Magnon Nordic component.

    The Cro-Magnon Y haplogroup is I.

    The Wikipedia article on Finns notes that “Of modern nationalities, Finns are closest to Cro-Magnons in terms of anthropological measurements”.

    If you look at the photograph of the skull at the top of the Wikipedia Cro-Magnon article, and then compare it with the photographs at the top of the Wikipedia article on the Sami, you’ll notice the obvious similarities. Ole Henrik Magga especially looks like a Cro-Magnon. Helga Pedersen and Renée Zellweger also have the Cro-Magnon look. The wide jaw is characteristic.

    The “North-East-European” component is the Aryan Nordic component. The Aryans were the Nordics who lived on the Pontic-Caspian steppe and spoke Proto-Indo-European.

    The Aryan Y haplogroups are R1a and R1b.

    The majority of the DNA of present-day Nordics is from Aryans. A minority is from Cro-Magnons.

    The “North_European” component encompasses both the Cro-Magnon Nordic component and the Aryan Nordic component.

    One of the commenters said he is pure Irish, and that he has 48.89% of the “North-East-European” component and 6.41% of the “North-European-Mesolithic” component. Irish have on average ~59% of the “North_European” component, so it’s the sum of the “North-East-European” component and the “North-European-Mesolithic” component that is equivalent to the “North_European” component.

  31. Jackson Devoni says:

    Hmmm okay i see how that can work. So the North-East-European ”Aryan” Nordic component from this MDLP analysis is also descended from European Mesolithic hunter-gatherers correct? It seems to be very similar to the North_European component that both Dienekes and Davidski have found in their various analysis.

  32. genetiker says:

    Both Cro-Magnon Nordics and Aryan Nordics evolved in Northern Europe during the Paleolithic. They’ve been in Europe ever since.

    The “North_European” component corresponds to the general Nordic race. The Cro-Magnon race and the Aryan race are the two major subraces of the Nordic race, and separate components emerge for each of them in higher K ADMIXTURE analyses.

    At the present day the Nordic component is dominated by the Aryan subcomponent, but it does nevertheless also include the Cro-Magnon subcomponent.

  33. […] May 28: Amerindians have both Mongoloid and Caucasoid physical features […]

    • Genetiker said: There was a massive influx of Mediterraneans from the Near East into Europe during the Neolithic, and this influx was primarily associated with the E1b1b Y haplogroup.

      This claim is in no way supported by ancient DNA. The Cardium pottery culture (the south European stream of early famers) and the cultures derived from it were clearly dominated by G2a (26 specimens so far). In comparison E-V13 was only found in 1 specimen. And the Linear Pottery culture (the Danubian stream of early farmers) so far yielded only 2 F and 1 G2a.

      Genetiker said: The “West Asian” component corresponds to the race that has been called the Alpine or Alpinid race in Europe and the Assyroid or Armenid race in the Middle East. It’s the same race in both places, but there’s a European variety and a Middle Eastern variety. This race entered Europe from Asia Minor during the Neolithic.

      It’s absolutely nonsensical to equate the West Asian component with Alpinids. First of all the West Asian component is very strong in Iran and Pakistan, two rather leptodolichomorphic countries. Second, if you look at the most recent reconstruction of Ötzi, you’ll note that he looked rather Alpine. Although he possibly wasn’t brachycephalic (?), facially he looks more Alpinid than Mediterranid. And the same can be observed on many Sardinians, his closest living relatives. Such Alpinid-like, but longheaded Sardinians are usually classified as Berid. Third, France, which does have some relatively Alpinid areas is comparatively little West Asian admixed. It would make sense to regard the Alpinids partly as brachycephalized Berids. Fourth, while there are certainly many Alpinid looking people in the Caucasus, in Turkey and southeastern Europe, you can also find many Dinarid looking people there. They look very different from Alpinids, but their distribution in Europe also correlates with West Asian admixture. Now you could conclude that West Asian is a conglomerate of brachycephalic Alpinids and Dinarids (or Armenids), just as you suggested. But then I point out again that the West Asian component is very strong in all of Iran, not just the brachycephalic northern fringe, and Pakistan. And that during the Iron Age and the Roman age the Near and Middle East, also the parts that are now brachycephalic, was dominated by longheaded people. And I would also point out that in Italy the South and Sicily are much more West Asian than the North, and yet the South and Sicily are at the same time more longheaded than the North, the latter being rather brachycephalic. Yet it does make sense that the South is more West Asian, you can see it in the facial features. So I conclude that the South was influenced by a rather longheaded West Asian influence.

      Genetiker said: Today Iberia is the most racially Mediterranean place in Europe behind Sardinia, but prior to the Neolithic Iberia was Nordic. It was flooded with Mediterraneans from North Africa during the Neolithic.

      According to the dominant view among modern archaeologists, the Iberian Neolithic stemmed from the Cardium Pottery culture which came from the east. An influence from northern Africa is controversial.

      Genetiker said: Alpines inhabited the parts of the Middle East closest to Europe, Asia Minor and Transcaucasia. There was gene flow between Nordic hunter-gatherers and Alpine hunter-gatherers during the Paleolithic and Mesolithic.

      Is there any cranial evidence for Alpines in the Middle East during the Paleolithic and Mesolithic?

      Genetiker said: First, ADMIXTURE doesn’t produce “hybrid components” for admixture events that took place in the past few thousand years, as is the case with Nordics and Mediterraneans. That’s the whole point of ADMIXTURE: to resolve the elements that mixed.

      Yet, in the case of the Amerindian component you claim that it is hybrid…

      Genetiker said: The “North-European-Mesolithic” component is the Cro-Magnon Nordic component.
      The Cro-Magnon Y haplogroup is I.
      The “North-East-European” component is the Aryan Nordic component. The Aryans were the Nordics who lived on the Pontic-Caspian steppe and spoke Proto-Indo-European.

      The suggestion that the North-European-Mesolithic component is Cro-Magnon related and the North-East-European component Aurignacoid does appear to make some sense. Baltic hunter-gatherers were particularly long-faced, and the North-East-European component peaks in the Baltic. And while western European hunter-gatherers tended to be more Cromagnoid, there have been Aurignacoids in western Europe since the Paleolithic, so the results of your analyses of the La Brana genomes are in line with this. However, note that the Swedish hunter-gatherers tested so far all had y-haplogroup I. So it does not appear to have been exclusively a Cro-Magnon haplogroup. No R has been found in Europe prior to the Chalcolithic. Also I have to point out that it’s not at all proven that the Proto-Indo-Europeans were Aurignacoid Nordics. Moreover, the hunter-gatherers and pastoralists of the Pontic-Caspian steppe were not Aurignacoid. The hunter-gatherers were strongly Cromagnoid, the pastoralists were partly mesomorphic, but according to a recent paper rather dark pigmented…

      • Joss Beaumont said: The suggestion that the North-European-Mesolithic component is Cro-Magnon related and the North-East-European component Aurignacoid does appear to make some sense. Baltic hunter-gatherers were particularly long-faced, and the North-East-European component peaks in the Baltic. 

        On a second thought, this doesn’t make sense either. Because even though the Baltic hunter-gatherers were particularly longfaced, the modern Baltic populations are not. This is because of the westward migrations of the eastern Baltic Latgalians and Lithuanians who assimilated the western Baltic Curonians, Semigallians and Selonians. Hence, the modern Latvians and Lithuanians are not overly Skando-Nordid, although that type is certainly still present there; but it’s mixed with a more broad-faced Cromagnoid eastern type. More generally I think it’s quite obvious that northeastern Europe isn’t the peak of the Skando-Nordid type. So, what is this MDLP World-22 North European-Mesolithic component? As your analysis of the complete La Brana genomes has shown, it isn’t the legacy of the western European hunter-gatherers, in contrast to later „Aryan“ expansions either, because the western European hunter-gatherers were predominantly Northeast European, too. All we can say so far, is that it is something Sami-related.

        The Dodecad Globe10 analysis showed:
        The Northern component peaks in Lithuanians, Finns, Swedes and Norwegians. Hence it is East Baltid and Skando-Nordid, but also Fälid, which is apparently related with the East Baltid type.
        The Southern component peaks in Saudis, Mozabites, Moroccans and Bedouins. Hence it is Arabid and South Mediterranid.
        The West Asian component peaks in Georgians, Abkhaz, Brahui, Makrani and Balochi. Hence it is Mtebid and Irano-Afghan. Related is the Dinarid type.

        If you now look at the K12b analysis, it becomes apparent that:
        The North European component peaks in Lithuanians and Finns, followed by Belorussians and Russians. It is just East Baltid.
        The Atlantic-Med component peaks in Basques and Sardinians, followed by Spaniards. It is Mediterranean, but also strong in the entire West of Europe. Hence H.F.K. Günther’s habit of calling the Mediterranid type „Westisch“ makes sense.
        The Northwest African component peaks in Mozabites, followed by Moroccans. It is Saharid.
        The Southwest Asian component peaks in Saudis, followed by Bedouins. It is Arabid.
        The Caucasus component peaks in Georgians, followed by Abkhaz. It’s Mtebid.
        The Gedrosia component peaks in Brahui, Makrani and Balochi. It’s Irano-Afghan.

        But where in this K12b analysis are the Skando-Nordids? They’ve disappeared. Obviously, in this analysis they are a blend of the North European and the Atlantic-Med components.

        If you now think the K12b analysis is obsolete, then take the Globe13 analysis instead. It’s very similar. Norway is 28,9% Mediterranean in Globe13. Sweden is 26,1% Mediterranean. Lithuania is 13 – 14,4% Mediterranean, Finland 9,8 – 10,2%. That makes the difference between the more East Baltid Finns and Lithuanians and the more Skando-Nordid Swedes and Norwegians.

  34. Genetiker said: First, ADMIXTURE doesn’t produce “hybrid components” for admixture events that took place in the past few thousand years, as is the case with Nordics and Mediterraneans. That’s the whole point of ADMIXTURE: to resolve the elements that mixed.

    In fact, the Globe13 Mediterranean component is a hybrid component: It is mixed of the Globe10 Southern and Northern („Atlantic-Baltic“) components.

    This can be seen in the fact that the Basques in Globe13 are 59,5% Mediterranean, but only 21,8% Southern in Globe10. So a large part of their Globe13 Mediterranean component is actually Northern.

    Similarly, the Sardinians are 71% Mediterranean in Globe13, but only 39% Southern in Globe10.

    And this leads us to the real reason why the La Brana samples had some Mediterranean admixture in Globe13. The reason is not that they were mixed with incoming Mediterraneans. Because in Globe10, La Brana 1 had 0% of the Southern component and 96,5% of the Northern (Atlantic-Baltic) component. The real reason for the Mediterranean component in La Brana is that these alleles were incorporated into the Mediterranean component. So these are alleles of northern, European hunter-gatherer origin which became part of the Mediterranean component.

  35. genetiker says:

    I acknowledged that G is associated with the Mediterranean component in this comment.

    I have been aware for a long time that some of my associations of phenotype-based classifications with autosomal components were problematic. Your criticism forcefully reminded me of the problems, and has led me to decide to abandon them.

    Associating the West Asian component with Alpines wasn’t completely nonsensical.

    If you look at this map of the MDLP World-22 West Asian component and compare it with the inset of this map from Madison Grant’s The Passing of the Great Race (which was based on William Z. Ripley’s tripartite system), you’ll see that all the populations for which the West Asian component is modal were classified as Alpine.

    According to the Wikipedia article on the Alpine race, “Ripley argued that the Alpines had originated in Asia, and had spread westwards along with the emergence and expansion of agriculture, which they established in Europe”. And this map from Grant’s book shows Alpines entering Europe from Asia Minor and Transcaucasia during the Bronze Age. These ideas are consistent with the Neolithic or Bronze Age introduction of the West Asian component to Europe.

    Armenians have one of highest levels of the West Asian component, and in The Passing of the Great Race Grant wrote “The Armenoids constitute an Alpine subdivision and may possibly represent the ancestral type of this race which remained in the mountains and high plateaux of Anatolia and western Asia”.

    Egon von Eickstedt considered the Alpinid race to include West-Alpinids and Lappids, and he classified Alpinids along with Dinarids, Armenids, and Turanids as all belonging to the larger race Homo sapiens albi brachimorphi, inhabiting what he called the mountain race belt. You can see that on his map of European races here.

    But as you correctly pointed out in the case of France, the West Asian component is not modal for any of the European populations that were always classified as Alpine.

    And as you also correctly pointed out in the case of Iran and Pakistan, there are populations for which the West Asian component is modal that cannot be classified as Alpine given the fact that brachycephaly was the main defining characteristic of the Alpine race.

    One has to conclude that the cephalic index was simply never a good criterion to use for racial classification, because brachycephaly has evolved independently in so many different places and at so many different times.

    I know that the Cardium pottery culture entered Iberia from the east, but it’s hard to imagine the earlier La Almagra pottery culture of Andalusia coming from anywhere but North Africa.

    Note that I said that ADMIXTURE doesn’t produce hybrid components for admixture events that took place in the past few thousand years. Actually, for a sufficiently high K value, ADMIXTURE will produce a hybrid component for any mixed population, but the further back in time the admixture event that produced that population occurred, and the more genetic drift there was after that admixture event, the lower the K value required. The admixture event that produced Amerindians occurred a very long time ago (over 15 ka), and there was a lot of genetic drift in Amerindians after that event, so it only takes a K value of 4 for ADMIXTURE to produce a hybrid component for Amerindians. The mixing of hunter-gatherers and farmers in Europe took place much more recently, and it took place during and after the Neolithic, when population sizes were much larger, and when there was therefore much less genetic drift. So it will take a significantly higher K value for ADMIXTURE to produce hybrid components for Europeans.

    I had never even seen the term “Aurignacoid” before reading your comments. It’s a horrible term for a racial type that is supposed to be contrasted with Cro-Magnons, because the Cro-Magnons were the Aurignacians. The first page that a search for the term turns up is this one. It says that the “Aurignacoid” type is based on the Combe Capelle skull. First, the belief that the Combe Capelle skull was from the Aurignacian time period was shown to be erroneous when accelerator mass spectrometry was used in 2011 to date it to 7575 BC. Second, the Comb Capelle skull is one of the most freakish, alien-looking things I have ever seen. The idea that anyone would use it as the holotype for a race is crazy.

    Just to be clear, I think that the Cro-Magnons of the Aurignacian were I, and that the Gravettians were R1 Indo-Europeans. So I think that starting with the Gravettian there was a mix of I and R1 throughout Europe. The finding that some of the hunter-gatherers in Sweden 8,000 years ago were I does not contradict this scenario.

    You say that no R has been found in Europe before the Copper Age. The Bell Beaker site at Kromsdorf where R1b was found and the Corded Ware site at Eulau where R1a was found were actually both late Neolithic sites.

    We only have six Y haplotypes from Mesolithic Europe, and they have already shown heterogeneity (I and C), and we don’t have any Y haplotypes from Paleolithic Europe, so it’s way too premature to be suggesting that there was no R in Europe during these time periods.

    And we now know that the mammoth hunters of Siberia 24,000 years ago were R. The idea that there was no R in Europe during the Paleolithic requires us to believe that although mammoths ranged from Iberia to Siberia, the R men who hunted those mammoths did not, but instead confined themselves to Siberia.

    I don’t believe that the Western European hunter-gatherers of 7,000 years ago were distinct from Indo-Europeans. I think that they were autosomally predominantly Indo-European.

    The North European Mesolithic component is obviously related to the Sami, and to a lesser extent Finns and other Finnic peoples. But as I pointed out in the comments above, and also in this post, the Sami have cranial features which are similar to those of Cro-Magnons. And the analysis of craniometric data in this paper showed that of all the populations analyzed, Sami and Finns had skulls that were the most similar to those of Cro-Magnons.

    Also, if the North European Mesolithic component is just a Sami component, then why does La Braña 1, a hunter-gatherer from Iberia 7,000 years ago, have more of that component than any modern European population except for Sami, Finns, and other Finnic peoples?

    In the past I have interpreted the globe13 North European component as being associated with the European hunter-gatherers, and the globe13 Mediterranean component as being associated with the European farmers.

    You’re saying that it’s the globe10 Atlantic Baltic component that’s associated with the hunter-gatherers, and the globe10 Southern component that’s associated with the farmers, and that the globe13 Mediterranean component is a hunter-gatherer and farmer hybrid component.

    Your explanation of the Mediterranean component in La Braña 1 had occurred to me before, and it is true that there is no way that it can be ruled out. But nor can my explanation of it as being genuine farmer admixture be ruled out, because it is a fact that the farmers of the La Almagra pottery culture had entered Iberia long before the time of La Braña 1.

    If we had DNA from Iberian hunter-gatherers from before 8,000 years ago, and it showed some of the Mediterranean component, then that would be compelling evidence for your interpretation. If such DNA didn’t show any of the Mediterranean component, it would be compelling evidence for my interpretation. Without any such evidence, both interpretations can only be considered conjectures.

    And I actually now think that it’s likely that neither interpretation is really all that accurate.

    I’ve come to the view that many of the subracial autosomal components are reflections of genetic gradients or clines that cannot sensibly be related to discrete, typological, phenotype-based subracial categories.

    Some of the subracial components may genuinely correspond to distinct populations, but these are often populations that existed many thousands of years ago, long before the formation of the historical and present-day European race.

    Associating such components with many of the traditional subraces is impossible, because many of the phenotypic traits that were used to define those subraces have emerged only relatively recently.

    As an example, Nordics were defined in part by their fair skin, while Mediterraneans were defined in part by their darker skin. But we now know that even just the alleles for basic white skin, as opposed to brown or tan skin, were still in the process of rising in frequency to fixation in Europe as late as the Bronze Age.

    Brachycephaly, one of the defining traits of Alpines, is likely another example of a trait that evolved in certain regions of Europe only relatively recently.

    • I don’t know much about the Almagra culture. And certainly the genetic evidence on the Iberian peninsula allows for some North African admixture. But for the European Neolithic in general, Northern Africa wasn’t an important source. The two main waves of the early European Neolithic were the Danubian one, giving rise to the LBK in central Europe, and the Mediterranean wave, represented by the Cardium / Impressa pottery, which can be traced back to Greece.

      I don’t believe you had never seen the term „Aurignacoid“ before, you must have forgotten about it. I agree though, that the name is problematic, considering that the Aurignacian is the early Upper Paleolithic culture of Europe in general. But note that judging from the C14 dating of associated grave goods, the human remains from the site of Cro-Magnon appear to be Gravettian rather than Aurignacian.

      Indeed, the Combe Capelle skull turned out to be Mesolithic, which explains why it looks relatively modern, compared to early UP finds.

      When I used the term Aurignacoid I was of course thinking of more leptomorphic types, in contrast to the very eurymorphic Cro-Magnon I skull. Albeit Combe Capelle is now out of the question, there are other more leptomorphic finds of comparable age as those from Cro-Magnon; Predmost 3 or Sungir 1 for instance:
      Predmost 3, Frontal View
      http://s1.zetaboards.com/anthroscape/topic/1019403/1/

      Just to make it clear: The Copper Age is a subsection of the Neolithic; it coincides with the late to final Neolithic (in central Europe). So yes, in the late Neolithic, R1a and R1b were present in central Europe. But before that, there is no shred of evidence of them, in all the samples from the Mesolithic to the earlier Neolithic, so far. Two mesolithic hunter-gatherers from western Europe were I2a and C; four mesolithic hunter-gatherers from Scandinavia were I or I2. Apparently I was rather common among mesolithic European hunter-gatherers, but there is also the surprising finding of the nowadays very uncommon C. If there was any R1 present, which is hard to rule out, it surely wasn’t common. In my view R1a and R1b were more eastern, they may have been present in eastern European hunter-gatherers. The recent study by Lazaridis et al. that you despise so much, did in fact show that there was a post-mesolithic and post-early neolithic admixture event with a (or several) population(s) that was more Ancient North Eurasian-admixed than the western and north European hunter-gatherers and the early European farmers. It would make sense if this was associated with an influx of R1a and R1b, because Ancient North Eurasian means Mal’ta 1-like, and as you say, he was R.

      As for the Indo-Europeans, that’s a linguistic, and to some extent, cultural category. Biologically, Indo-Europeans vary a lot. The Irish, the Norwegians, the Latvians, the Spaniards are all Indo-Europeans, as are the Greeks and the Armenians, the Pathans and the Persians, the Singhalese etc., etc. On the other hand, non-IE Basques are not far away genetically from IE western Europeans, the non-IE Estonians are not far at all from IE Latvians, etc. What’s true is that the Proto-Indo-Europeans must have been a homogenous population, but it’s not quite clear where they lived and what they were like. Some respected linguists have presented good linguistic evidence for an IE homeland in western Asia. And the Dodecad experiments have shown that IE Europeans, even in the far west and north do have some seizable West Asian admixture, whereas the non-IE Basques and Finns don’t. And recently the Corded people, regarded by many scholars as early IEs, proved to have mt-DNA with links to the Caucasus and trans-Caucasus. The custom of erecting Kurgans arose south of the Caucasus anyway. So even if you’re not convinced of a West Asian homeland of the Proto-IEs, you can’t dismiss the possibility offhandedly.

      And I have to say, the association of R1b with IE, at least as far as western Europe is concerned, is illusory. In southwestern Europe, R1b is actually stronger in areas where IE was introduced late, by the Romans (or not at all, in the case of the Basques), than in the old Celtic and „para-Celtic“ areas. That’s exactly the opposite of what we would expect if R1b was introduced by the IEs. From northwestern Europe we don’t have any non-IE languages, because writing was introduced too late. But I noted that in northwestern Europe I2b, a central European variant of I2, must have arrived after R1b, because they have complementary distributions, with R1b being strongest in refuge areas. And I2b doesn’t exclusively correlate with Germanic influence, but is found at similar frequencies in Celtic areas where Germanic influence was minimal. And since I2b came from central Europe, to me it’s clear that it was brought to northwestern Europe with the Celtic invasions – so they arrived there when R1b was already in place, and the whole association of Italo-Celtic with R1b is spurious. And in other IE populations, R1b isn’t that predominant.

      The paper that is supposed to show the similarity between Cro-Magnons and Sami and Finns uses the term Cro-Magnon in a very wide sense, so that it actually means nothing more than European hunter-gatherers. If you check the skulls that the author included into his Cro-Magnon sample, these were (with ages in brackets):

      Cro-Magnon 1 (27,7k), Predmost III (26 – 27k), Pavlov I (25,5 – 27k), Obercassel I (14k), Chancelade I (17k), and Cheddar I (7k).

      So the study demonstrated that Sami and Finns are close to European hunter-gatherers, but that’s rather uncontroversial.

      Finally, if you look at the aDNA samples with Globe10:
      Ötzi is 42,2% Southern.
      Gok4 is 33,7% Southern.
      Bra1, Ajv52 and Ajv70 are all 0% Southern.

      The only reasonable conclusion being that the Southern component was brought by the early farmers, the European hunter-gatherers didn’t have it.
      And since Bra1 had 0% Southern he could hardly have any farmer admixture.

  36. genetiker says:

    I think that Sungir 1 was Y R, which leads me to believe that the leptomorphic type of Sungir 1 and Predmost 3 was associated with R and introduced to Europe during the Gravettian, and that the eurymorphic type of Cro-Magnon 1 was associated with I and introduced to Europe during the Aurignacian.

    I is the most common haplogroup in the Sami after N, and I think that Sami crania look like a brachycephalized version of the Cro-Magnon 1 cranium.

    The Copper Age is thought of as part of the Neolithic by some people, but certainly not by everybody. The first paragraph of the Wikipedia article on the Copper Age states

    “The Chalcolithic period or Copper Age, also known as the Eneolithic/Æneolithic (from Latin aeneus ‘of bronze’), is a phase of the Bronze Age before metallurgists discovered that adding tin to copper formed the harder bronze. The Copper Age was originally defined as a transition between the Neolithic and the Bronze Age. However, because it is characterized by the use of metals, the Copper Age is considered a part of the Bronze Age rather than the Stone Age.”

    It goes on to say that

    “In 1884 Gaetano Chierici, perhaps following the lead of Evans, renamed it in Italian as the Eneo-litica, or ‘Bronze-stone’ transition. This phrase was never intended to mean that the period was one in which both bronze and stone were used. The Copper Age features the use of copper, excluding bronze; moreover, stone continued to be used throughout both the Bronze Age and the Iron Age. ‘Litica’ simply names the Stone Age as the point from which the transition began and is not another -lithic age. The Eneolithic was never part of the Stone Age, which ended conclusively the moment the first smelter succeeded in obtaining copper from copper ore for the first time.”

    The Wikipedia article on the Neolithic says that

    “It ended when metal tools became widespread (in the Copper Age or Bronze Age; or, in some geographical regions, in the Iron Age).”

    Later it says

    “The Chalcolithic period began about 4500 BC, then the Bronze Age began about 3500 BC, replacing the Neolithic cultures.”

    It also says

    “The clothing worn in the Neolithic Age might be similar to that worn by Ötzi the Iceman, although he was not Neolithic (since he belonged to the later Copper age).”

    The Copper Age is defined here as

    “a cultural period between the Neolithic and the Bronze ages, marked by the development and use of copper tools.”

    And it’s defined here as

    “a short prehistoric period after the Stone Age and preceding the Bronze Age, when some tools and weapons were made of copper”

    The sole criterion determining whether the Kromsdorf and Eulau sites were Neolithic or Copper Age, as those periods are defined above, is the type of artifacts found in association with the human remains. The artifacts found at those two sites were made out of stone, not copper, so by the above definitions, they were Neolithic sites, not Copper Age sites.

    The fact that R1b and R1a have been found at Neolithic sites is important, because in Marija Gimbutas’ version of events, Copper and Bronze Age IE-speaking pastoralists from the steppes of Eastern Europe invaded and conquered the inhabitants of Central and Western Europe. That idea is undermined by the existence of R1b and R1a (which are associated with IE languages) at sites where there were only stone artifacts and no copper or bronze weapons.

    To see just how absurd Gimbutas’ account is, you can look at this map and this map. The first map shows that the steppes of Eastern Europe, above the Black Sea and the Caspian Sea, were the last places in Europe that agriculture spread to, so the population density of Central and Western Europe was much higher, at a much earlier time, than that of the steppes of Eastern Europe. The second map shows that metallurgy existed along the Danube, in Central Europe, and in parts of Western Europe long before it existed in the steppes of Eastern Europe. The idea that the high population density areas of Central and Western Europe were invaded and conquered by people from Eastern Europe who developed agriculture thousands of years later, and who developed metallurgy hundreds of years later, is ridiculous.

    Not to mention the fact that there’s zero archeological evidence of a Copper or Bronze Age invasion of Western Europe by Eastern Europeans.

    The sample of Y haplotypes of Mesolithic Europeans that we currently have is too small to draw any meaningful conclusions about the presence or prevalence of R1b and R1a during that time period.

    The idea that R and its descendants were confined to Siberia, or to Siberia and Eastern Europe, until the Copper or Bronze Age doesn’t make any sense.

    During the Upper Paleolithic, the steppe-tundra ecoregion stretched from the Atlantic coast of Europe to Lake Baikal. Mammoths roamed the steppes over that entire region. There’s absolutely no reason why the R mammoth hunters at the eastern end of that area wouldn’t have expanded to fill the entire area; indeed, that is exactly what we would expect.

    In addition to the ecological continuity from the Atlantic to Lake Baikal, there was cultural continuity. Venus figurines are found over the entire area.

    And there is also genetic continuity over the entire area. Mitochondrial haplogroup U has been found from one end to the other.

    It doesn’t make any sense to think that there was ecological continuity, cultural continuity, and maternal genetic continuity across the whole steppe-tundra region, but no paternal genetic continuity.

    Also, the distributions of R1b and R1a don’t make any sense if one supposes that they each originated in Asia, or in Eastern Europe, and then spread across Europe. If that were the case, then they would both have similar distributions, with a decreasing gradient from east to west. What we see instead is a partitioning of the two into separate regions, with R1b concentrated in Western Europe, and R1a concentrated in Eastern Europe. The frequency of R1a drops precipitously at around 15°E, as though its further expansion had been repelled, and the frequency of R1b sharply rises at around the same longitude.

    The distributions do however make complete sense if one supposes that R entered Europe at some very remote time in the past, and that R1b then originated somewhere in Western Europe, and R1a somewhere in Eastern Europe. Over time R1b and R1a would have expanded from their points of origin, with the expansion of one acting to repel the expansion of the other at around 15°E. Although it must be noted that R1b was apparently more successful in repelling R1a than vice versa, because the distribution of R1b extends far to the east.

    And this scenario is consistent with what we know from archeology and genetics. It was estimated here that Q and R diverged from each other 33,000 years ago, which is not long before the beginning of the Gravettian. The latest estimate for the divergence of R1b and R1a is 25,000 years ago, which coincides perfectly with the start of the Last Glacial Maximum. The LGM would have split the R1 in Europe into one population occupying a Western European refuge, where R1b would have originated, and another population occupying an Eastern European refuge, where R1a would have originated.

    David Reich and his associates show no understanding of the simple concept that Loschbour and Mal’ta 1 being separated by 16,000 years of time means that they’re separated by 16,000 years of evolution. Mal’ta 1 appears to be a mix of North European Caucasoid, West Asian Caucasoid, and Veddoid components not because he’s the product of interbreeding between multiple populations, but because he’s a representative of an intermediate stage in the process by which some Veddoids evolved into Caucasoids. Mal’ta 1 shows some of the West Asian component because at that point in time the evolutionary divergence of Upper Paleolithic Europeans from Upper Paleolithic West Asians had not progressed to the point that it ultimately would 16,000 years later. Afontova Gora 2 shows more of the North European component and less of the Veddoid South Asian component not because of some admixture event, but because he was 7,000 years further up the path of Upper Paleolithic European evolution.

    It makes absolutely no sense to calculate how much Mal’ta 1 (“ANE”) ancestry there is in a modern-day Caucasoid individual or population, because no modern-day Caucasoid individual or population is directly descended from a representative of such an early stage of Caucasoid evolution. It would be just as nonsensical for a modern-day Mongoloid to calculate how much Tianyuan ancestry he has. Tianyuan shows more of the Veddoid South Asian component than any other component, but most Mongoloid populations today don’t have any Veddoid DNA. But of course the fact that such calculations are utterly nonsensical hasn’t deterred David Reich and his associates or Davidski and his mindless followers from going right on ahead and making them.

    More generally, it doesn’t make any sense to base entire autosomal admixture components on single individuals widely separated in time and space. But that’s exactly what David Reich and his associates idiotically did when they based “ANE” on Mal’ta 1, “WHG” on Loschbour, and “EEF” on Stuttgart.

    The admixture that entered Europe after the early Neolithic that is seemingly related to Mal’ta 1 is nothing but the West Asian component, which has nothing to do with R or Indo-Europeans. The distribution of the West Asian component shows that its presence in Europe is the result of demic diffusion from Asia Minor to Greece and Italy and then into the rest of Europe. Basques don’t have any of the West Asian component because they’ve been isolated since the Neolithic.

    Comparing the distribution of J2 with the distribution of the West Asian component shows that there’s a very strong relationship between the two. No other haplogroup shows such a strong relationship with the West Asian component. R1b and R1a certainly don’t. And it only makes perfect sense that the West Asian component was brought to Europe by J2 men, because J2 is the dominant haplogroup in the same part of West Asia where the West Asian component is the dominant autosomal component.

    If one supposes that the only Y haplogroup that was prevalent among Paleolithic and Mesolithic Europeans was I, and that the only mitochondrial haplogroup that was prevalent among them was U, then the amounts of the North European autosomal component that one finds across Europe don’t make any sense. There’s far too much of it to have been contributed only by I males and U females. Most of the mitochondrial haplogroups in Europe today are from the farmers, with U only being found at low frequencies across most of Europe. And R1b and R1a are the dominant haplogroups in most of the areas of Europe where there’s a lot of the North European component. The haplogroup frequencies and autosomal composition of Europeans make far more sense if R1b and R1a were Paleolithic and Mesolithic European haplogroups.

    The earliest studies on R1b-M269 concluded that it was in Europe during the Paleolithic. In 2010 this paper found that STR haplotype diversity for R1b-M269 decreased from Asia Minor to Western Europe, and claimed that R1b-M269 was therefore brought to Europe during the Neolithic by farmers from the Near East. The findings of this paper were contradicted later in 2010 by this paper, which supported the presence of R1b-M269 in Europe during the Paleolithic. Then in 2011 this paper also contradicted the findings of the first 2010 paper by looking at a much larger sample of R1b-M269 Y chromosomes and finding no geographical trends in haplotype diversity.

    The association of R1b with Indo-European languages is not “illusory”. It’s quite real.

    It’s obvious to everybody that R1a is associated with Indo-European languages. Its distribution is clearly a reflection of the Aryan invasion of India. (Note that the distribution of the North East European component in Asia parallels that of R1a, and that it is unique in doing so.)

    But it is only the satem Indo-European languages that have a distribution that coincides with that of R1a. It’s just as clear that the centum Indo-European languages have a distribution that coincides with that of R1b.

    And it’s not just that the two major divisions of the IE languages correspond to the two major divisions of R1. The subdivisions of these two main divisions of IE correspond to the subdivisions of R1b and R1a.

    The distribution of R1b-S21 corresponds to the distribution of Germanic languages.

    The distribution of R1b-P312 corresponds to the distribution of Italo-Celtic languages.

    The distribution of R1a-Z282 corresponds to the distribution of Balto-Slavic languages.

    The distribution of R1a-Z93 corresponds to the distribution of Indo-Iranian languages.

    And the subdivisions of R1b-P312 correspond to the subdivisions of Italo-Celtic languages.

    The distribution of R1b-L21 corresponds to the distribution of Insular Celtic languages.

    The distribution of R1b-S28 corresponds to the distribution of Gaulish and Italic languages.

    The distribution of R1b-DF27 corresponds to the distribution of Hispano-Celtic languages.

    In Iberia there’s a decreasing gradient of R1b from the northeast to the southwest. The reason for that is that starting in the Neolithic E, G, J, and T entered Iberia from North Africa, displacing the indigenous R1b.

    Basques are unlike other populations having high frequencies of R1b in that they have the second highest level of the Mediterranean autosomal component in all of Europe. The Mediterranean component makes up about 70% of Sardinian DNA and about 60% of Basque DNA. Sardinians spoke a non-IE language up until the arrival of the Romans, and there is evidence that this language was related to the Basque language. Basques represent a unique case of hunter-gatherer men adopting the language of their farmer wives.

    Absolutely nothing can be inferred about the presence or absence of IE or non-IE languages in Northwestern Europe from the late introduction of writing there.

    The distribution of I2a2 (formerly I2b) clearly coincides with the distribution of Germanic languages. More specifically, I2a2a-M223, which makes up over 90% of I2a2, is associated with Germanic languages, while I2a2b-L38 is associated with Alpine Celts. The complementary distributions of R1b and I2a2 in Northwestern Europe is the result of I2a2 making up part of the Germanic lineages that displaced the earlier R1b Celtic lineages.

    The Welsh are about 90% R1b, and the Welsh language is an IE language. No haplogroup but R1b can possibly be associated with the Welsh language.

    Since R1b men were the originators of the centum IE languages, we know that the Kurgan hypothesis is false, because the Bell Beaker people were R1b, and the Bell Beaker culture was definitely not considered to be part of the “Kurgan culture” of Gimbutas’ account.

    The admixture plot here makes it clear that La Braña 1 did not have any farmer admixture.

    But it’s not possible to say that the globe10 Southern component is the farmer component, or that the globe13 Mediterranean component is the farmer component. In the above plot, there are components from K = 9 to K = 18 that are similar to the globe10 Southern component, and there are components at K = 19 and K = 20 that are similar to the globe13 Mediterranean component. All of the hunter-gatherers lack all of these components. Stuttgart has varying levels of them depending on the K value. It’s impossible to single out one of these components as being the farmer component, especially considering the fact that the analysis only includes a single farmer data point from the huge extent of space and time occupied by the European Neolithic.

    • I wouldn’t rely too much on Wikipedia. While it’s generally a valuable source of information, it’s not completely free of errors, especially in rather esoteric topics, where only few people are really informed.

      As far as European prehistory is concerned, the Copper Age doesn’t belong to the fundamental divisions. The sequence is: Paleolithic, Mesolithic, Neolithic, Bronze Age, Iron Age, etc, with further subdivisions respectively. So basically the Neolithic ends just with the beginning of the Bronze Age.

      There are various opinions regarding the terminology and chronological division of the Neolithic subsections. A reasonable division in my view was suggested by J. Lüning. According to him the sequence in central Europe was:
      Early Neolithic 5500 – 5000 BC
      Middle Neolithic 5000 – 4400 BC
      Upper Neolithic 4400 – 3500 BC
      Late Neolithic 3500 – 2800 BC
      Final Neolithic 2800 – 2200 BC
      And then followed the Early Bronze Age.

      So nowhere there is a section called Copper Age. It’s a vague term which roughly encompassed the Upper, the Late and the Final Neolithic, when copper was used as a material for tools and ornaments, but throughout which stone remained by far the more important resource. (Which is why it makes more sense to ascribe it to the Neolithic.)

      Of course these divisions are not absolute, but strongly dependent on the region in question. In northern Germany the sequence was different:
      Early Neolithic 4400 – 3300 BC
      Middle Neolithic 3300 – 2380 BC
      Late Neolithic 2380 – 1800 BC.
      Then followed the Bronze Age.

      Saying that the Bronze Age began about 3500 BC, as the wikipedia article does, is of course complete nonsense. In some areas it started around 3500 BC, in other areas it started long after that. This is a prime example of the misconceptions that half informed people carry around, and it’s a common mistake that even scientists make. The reason is that they mistake these relative divisions for absolute.

      Recently a preprint of a scientific paper called Ötzi even a Bronze Age farmer. This is completely and utterly wrong. He was Neolithic, late Neolithic to be precise, and of course also from the Copper Age.

      Both the Corded Ware culture and the Bell Beaker culture (to which the Eulau and the Kromsdorf findings respectively belong) belong to the Final Neolithic and at the same time to the Copper Age. These cultures definitely made use of copper, there’s no point in questioning that. But as I said, throughout the Copper Age, stone remained much more important.

      You’re really making a twisted, pointless argument when you’re trying to squeeze out any valid conclusions regarding Gimbutas’ theory from picking on that copper-bronze entanglement and from mixing in R1a and R1b. But let me explain step by step:

      Gimbutas wasn’t an idiot, she knew well what technological era the cultures she was speaking about, belonged to. Basically she was speaking of Neolithic cultures. I know that the guy from Eupedia, Maciamo, consistently speaks of the Bronze Age and Bronze Age cultures, he also calls the Corded Ware a Bronze Age culture, which is nonsense. Obviously Maciamo isn’t an archeologist. So don’t confuse Maciamo’s account of the Kurgan theory with Gimbutas’ account. It’s clear that the Yamnaya culture also made some use of copper. Its neighbours to the south and southwest were very advanced with copper metallurgy, and they definitely exerted some influence on the Yamnaya culture.

      Maciamo appears to explain the expansion of the Kurgan cultures, which he equates with the early Indo-Europeans, with their superior metallurgical skills, among other things. However, afaik the classical Kurgan theory propagated by Gimbutas, and more recently by David Anthony, said no such thing. It would be rather silly, in light of the early advanced metallurgical knowledge of the Balkans, which no serious archeologist would deny. I think Gimbutas and Anthony put much more weight on horse domestication and horse riding. That’s a questionable argument too, but at least it makes no reference to bronze technology or copper. In fact, there is no reconstructed Proto-Indo-European word for bronze, that’s why linguists concluded that the proto-language must have split before the Bronze Age.

      No serious Kurganist would claim that the Kurgan people from the Pontic steppe conquered Europe up to the Atlantic in one sweep. Instead they speak of a step-by-step process. Generally they believe that the Celts reached the Atlantic late, in the Bronze Age or even as late as the Iron Age. So there you have the Bronze Age in play, but not in connection with the earliest Kurgan groups from the steppe.

      As I said, there is no argument on whether the Corded Ware or the Bell Beaker culture were Copper Age cultures. They definitely were. The Bell Beaker culture was especially rich in copper, and one of its aims appears to have been the securing of copper trade routes. And even the Funnel Beaker cultures of central-northern Europe predating the Corded Ware had used some copper, though more rarely.

      And as I said, I wouldn’t conclude from the presence of R1b to the presence of Indo-Europeans. But more about that later.

      Actually it’s questionable if the people in the Pontic steppe ever became real farmers, the climate isn’t well suited for farming. It’s been discussed if they once had limited farming, but eventually they preferred to become pastoralists. The farmers from the Cucuteni-Tripolye culture expanded some way eastwards, but not behind the Dnieper and lower Don. But this makes the Steppe folks an excellent candidate for the later North European and Anient North Eurasian admixture that mixed into the Sardinian- and Basque-like Neolithic Europeans. Another important source are probably the more northern hunter-gatherers of eastern Europe that became assimilated by the Corded Ware. Although the Steppe people were not farmers, they were nonetheless quite populous. As Mallory pointed out, at Khvalynsk on the Volga there have been found about 200 burials + cattle and sheep, dating to the middle of the 5th millennium BC.

      But I admit that the Kurgan theory as presented by Gimbutas and Mallory is nonetheless very problematic and has to be evaluated critically. I think while it may contain some good ideas, it’s most probably not entirely right.

      By the way there is an interesting post regarding R1a, with subsequent discussion of Indo-European origins and the Kurgan theory here:
      http://dienekes.blogspot.com/2014/03/major-new-article-on-deep-origins-of-y.html

      Interesting, you simply brush the paleo-genetic evidence aside, with the excuse that the sample is too small. I would give the direct evidence much more weight than any hypothetical constructs. In two very different places of Europe, haplogroup I was found, and almost all the mesolithic haplotypes so far were I. Yet you say it wasn’t predominant?? And do you really believe that, if you pick out one random haplotype from a population (like mesolithic Iberia), it’s plausible to assume that the haplogroup by chance is a very rare one? No, if you pick out just one haplotype, chances are good, that you get one that is quite common. So, it’s reasonable to assume that C wasn’t rare in mesolithic Iberians, it presumably was common. Yes, it isn’t common today, but can’t you imagine that frequencies of haplogroups can change over the millennia? I won’t exclude that there may have been some R in mesolithic western Europe, as I said, it’s hard (or even impossible) to disprove, but so far 2 of 2 mesolithic western Europeans were something else, casting serious doubt on the assumption that R was common at that time and place.

      I still think that R1b originated in the east. The reason why it’s more common in the west may be founder effects and drift. By chance there was a limited period of time when R1b males were far more successful than I males. I think that explanation would suffice. Once R1b was very common, it couldn’t be supplanted by later invasions, and chances that any highly successful clan leader in later times coincidentally happened to be R1b were higher than that he wasn’t, so he would only have increased the incidence of R1b further.
      The reason for the sharp border in R1a/R1b proportions is that to the east and to the west of it are different populations, Germanics and Slavs respectively, with each having their own prehistory and ethnogenesis. Also, the border may have been slightly less sharp before the expulsion of Pomeranian, Silesian and Prussian Germans and the reshaping of the German borders after WW2.

      Well, „Reich and his associates“ had more than just one sample for their concept of ANE, WHG and EEF. They had 3 EEFs, 3 WHGs and 2 ANEs, and these clearly formed clusters.

      You say the Mal’ta 1 related admixture that entered Europe is nothing but the West Asian component. However I note that in the PCA in fig. 1B, many European populations are shifted towards ANE without a clear tendency towards the West Asian cline on the right hand-side of the plot. There is rather a European cline from EEF towards ANE that kind of parallels the cline in the Near East, between the Bedouins and the Caucasus peoples. Only the strongly West Asian admixed populations like the Greeks and South Italians are inbetween these clines, with a clear tendency towards West Asia. And according to the admixture model, the Scottish for instance are among the most strongly ANE admixed populations of Europe, even though their West Asian admixture is very modest. In the updated version of the paper there is also a PCA where the principal components are based on Loschbour, Stuttgart and MA1 and the West Eurasian populations are projected onto it. There are clearly two seperate peaks approaching MA1. One is composed of Mordovians and Russians, the other one of Lezgins.

      I agree that at least R1a was associated with the populations that dramatically increased the North European component in Europeans. Obviously I1 did so as well, since there isn’t that much R1a in Scandinavia.

      As for the origin of R1b-M269 and of R1b in general, I think it’s more instructive to look at the phylogeography than at minor variations in the STR variance. First of all, what are the basic divisions of R1b? There is V88, present in the Levant and in Africa. There is M335, present in Anatolia. And there is P297, which subsequently split into M73, common in Hazaras and Bashkirs, and into M269, which is most common in western Europe. But the M269 in western Europe belongs mostly to further derived variants. We have to look where R1b-M269* is most common, and that is in the Balkans: In Kosovo it reaches a frequency of 7.9%. So this points to an origin of R1b somewhere in West Asia and of M269 in the Balkans. The path to the haplogroups that dominate western Europe goes via L23. L23* has a frequency of 27% in the Valais, in the Swiss Alps. So we can see the path of R1b from the Balkans to central Europe, presumably in the Copper Age. Via some further mutations we reach R1b-P310/L11*. This is most common in central England, eastern Denmark and northern Poland. And this was the direct ancestor of R1b-U106 and P312 that now dominate western Europe. But note that already P310/L11, their direct ancestor, had a distribution and a centre of weight that was clearly more to the east, quite well coinciding with the Bell Beaker East group, in central-northern Europe rather than in the west. And that’s where R1b was first found in the paleogenetic record. I think its explosive expansion occured in the Bell Beaker culture and immediately succeeding cultures.

      I agree that R1a correlates quite well with Indo-European languages. However I don’t agree on the satem/centum dichotomy being paralleled by R1a and R1b respectively. The Armenians are satem, yet they have quite some R1b, and nearly no R1a. The Germanics are centum, yet R1a isn’t uncommon among them, and there are R1a branches that are purely Germanic. Moreover, apart from the predominance of R1b in Celtic and Italic populations, there is no clear association of R1b with centum Indo-European. There is quite some R1b in Anatolia, but so is J2, and so it’s not clear which is from the ancient Anatolian branch of IE. The idea that the Tocharians were R1b is speculation. The Tarim mummies at least are R1a. And Germanics are a mix of I, R1b and R1a.

      I agree that R1b-S21 correlates extremely well with Germanic influence. But how do you know that it doesn’t stem from a formerly non-IE R1b substrate that eventually became IE, and then the S21 mutation arose in a population that had evolved into speaking proto-Germanic?

      R1b-P312 does not correlate with Italo-Celtic. In southwestern Europe it peaks in areas that adopted an Italo-Celtic language late, from the Romans, and in the non-IE Basques. Note that there had been several IE and non-IE languages on the pre-Roman Iberian peninsula. IE were Celt-Iberian and Lusitanian, and they were in regions where the incidence of R1b varies from more than 50% to more than 60%. The definitely non-IE languages were Iberian and Basque. Tartessian was presumably an additional non-IE language, but that is controversial. Now, Iberian was present along the entire eastern coast up to the Pyrenees. The frequencies of R1b in this area vary from more than 50% in the south to more than 80% in Catalonia. In the Basque country it’s also more than 80%. So the peaks are clearly associated with non-IE Iberians and Basques while the IE Celt-Iberian and Lusitanian areas have a somewhat reduced incidence of R1b.

      Yes, R1b-L21 corresponds to Insular Celts, but that’s probably the legacy of the pre-Celtic substrate that became assimilated.

      R1b-S28 is indeed associated with Gaulish, IE Ligurian and Italic, but I see it as the legacy of a non-IE substrate exactly like it’s Germanic cousin R1b-S21.

      As for the Iberian DF27, the Eupedia maps show it strongest in the east and in the north along the Pyrenees, and that’s exactly where pre-IE Iberian and Basque were spoken. The IE languages of Iberia had their focus in the northwest, west and centre, roughly speaking, where the frequency of DF27 is reduced. I would take these hints very seriously. And the early Neolithic Cardium pottery culture came in from the east and it brought along lots of G2a and also some E-V13. I’m not sure about J, at least J2 probably came later, from the east and is strongest in the south.

      It’s extremely unlikely that Proto-Indo-European split up somewhere in the Paleolithic or Mesolithic. The reconstructed vocabulary points clearly to the Copper Age. And even Bayesian approaches calculating with the average pace of word replacement don’t reach an age that high; the latest calculation that I know of arrived at an age of about 8000 years. So Hispano-Celtic cannot be the relic of Iberian hunter-gatherers.

      And by the way, Basque mt-DNA contains a lot of hunter-gatherer haplotypes.

      About the ancient Sardinian language close to nothing is known. Ancient Iberian may have been related with Basque, but even this is controversial.

      I’d say the Basques are like Copper Age Europeans before the arrival of genetically more North European people from the east. They had their R1b from the old Balkan Copper cultures, they were basically Mediterranean like all early European Farmers, and they had acquired some western European hunter-gatherer ancestry evidenced in their mt-DNA and their slight north European component, which makes them more western than many other populations in the PCA, and which differentiates them from the earliest farmers and from the Sardinians.

      Yes, I2a2 is relatively strong in England, likely because of Anglo-Saxon input. But did you see that it’s equally strong in southwestern Scotland, in the Scottish Lowlands, and in the west of Ireland? The Germanic contribution in these areas was very limited and certainly not comparable to the Germanic input in England. So I stand by my interpretation that it was brought by central European early Celts.

      Celtic was presumably brought to Wales by British R1b people who had been Celticized.

      As for your your argument regarding the Bell Beakers, R1b and the Kurgan theory; IMO the Bell Beaker culture wasn’t Indo-European, at least not predominantly and certainly not originally. It originated in the Iberian peninsula, one of the places that became Indo-Europeanized latest. And whatever the language of R1b people originally was, they were assimilated by this Iberian culture and subsequently incorporated into the Basque gene pool. So I’d rather turn the argument the other way round: The fact that Bell Beakers were R1b speaks against R1b being Indo-European. But who knows, maybe they just switched languages.

      With reference to the admixture plot of La Brana 1, I just say: Q.E.D. I’ve never claimed it had farmer admixture. But indeed, the Globe10 Southern component isnt’ the farmer component either, it was just brought in by the farmers. But this doesn’t mean that they were purely southern. Not even Bedouins and Saudis are purely Southern, so why should Early Farmers have been? I hadn’t intended to suggest such a mad thing.

  37. Hang on!
    I’ve just had an enlightenment!

    I think you’re probably right that R1b in western Europe came with the Italo-Celtic branch of IE, and isn’t the legacy of a non-IE substrate.

    But let me elaborate:

    That recent study by Underhill et al. on R1a, which I linked you to, came to the conclusion that R1a originated somewhere near the Turkey/Iran border.
    Presumably its sister branch R1b originated not too far away. And this is indicated by the phylogeography:

    The basic branches of R1b are:
    V88 = Levant, Africa
    M335 = Anatolia
    P297 = ? It split into:
    M73 = Hazaras, Bashkirs, i.e. an eastern branch
    and into M269* = Balkans, Armenia, i.e. a western branch

    Now, interestingly M269* isn’t purely western, but it’s also present in 2.4% of Bashkirs. So it must have been present in the West Asian ancestors of the Bashkirs when they started to migrate eastwards.

    From M269* followed L23*, whose distribution is similar to that of M269*, just a little wider. It isn’t only common in the Valais in Switzerland, but also in Kosovo, Armenia and notabene in Bashkirs! So it must have already existed at the time when the West Asian ancestors of the Bashkirs started to migrate eastwards.

    So this L23* had a very wide belt-shaped distribution, ranging from central Europe, across the Balkans, West Asia and from there probably through central Asia to the Urals, where the Bashkirs live.

    Now I have to make a step to another category of evidence. The Dodecad K12b analysis had two West Asian components: A western, Caucasus component, and an eastern, Gedrosia component which peaked in Balochistan. I’ve always had the impression that the Gedrosia admixture in Europe is linked with R1b. In Europe, Gedrosia is strongest in the Northwest, especially on the British Isles. In the K12b analysis, the European peak of Gedrosia was in Argyll (western Scotland). It’s particularly strong in insular Celtic populations, and on the continent follow, in descending order, the Bretons, Denmark, the Dutch and interestingly the Basques, who have lots of R1b.

    This Gedrosia component isn’t just nonsense, because it’s closely replicated by the MDLP World-22 Indo-Iranian component. This too has a secondary peak in northwestern Europe, particularly in Wales it seems.

    These autosomal components were not present in the old European hunter-gatherers, and neither in the Neolithic farmers. They arrived later. It may seem puzzling that this admixture apparently links northwestern Europe with the eastern part of West Asia. But I think it makes sense. Let me elaborate further:

    It has been observed before by linguists that the Italo-Celtic branch of IE has similarities with Tocharian. From this, some have concluded that the Tocharians must have originated in central Europe. Others have suggested that the similarities are merely the result of the early split of both branches from the IE mainstream. (Linguists agree that Italo-Celtic and Tocharian both split early from the remainder of IE.) But Gamkrelidze and Ivanov suggested that there really once was a common Tocharo-Celto-Italic branch of IE, besides Anatolian and the still undifferentiated rest.

    Now, Tocharian was spoken in the Tarim basin in the west of modern China, hence very far in the east.

    And modern archeology has more and more evidence that the forebears of the Tocharians arrived there from western Asia via the Silk Road, rather than being an eastern extension of Pontic-Caspian Kurgan tribes.

    And although the Tarim mummies were R1a, R1b is also present in the Uighurs who live in the area now. The Tocharian language may have been brought by R1b people from western Asia. It’s noteworthy that Tocharian has a rich agricultural vocabulary which doesn’t fit with a steppe origin.

    To me it seems very plausible that the Tocharo-Celto-Italic branch of IE migrated from western Asia along the Silk Road and was rich in R1b-L23* and the Gedrosia/Indo-Iranian autosomal component.

    The archeologist Stanislav Grigoriev has suggested that the forebears of the Tumulus culture (central Europe, middle Bronze Age) have originated in southern Siberia, between the Urals and the Irtysh. To my knowledge, the origins of the Tumulus culture are not quite certain. So it seems possible that Grigoriev’s interpretation is right. In other words: The Italo-Celtic branch arrived late in central Europe, in the middle Bronze Age, and it came from southern Siberia. From there it brought R1b-L23* and the Gedrosia/Indo-Iranian autosomal admixture.

    It should also be noted that recent paleo-genetic evidence has indicated that western European hunter-gatherers, early European farmers and Kurgan people from the Pontic steppe all were relatively dark pigmented (except for light eyes in western hunter-gatherers). Genetic evidence of Bronze Age South Siberians however showed that they were light-pigmented. So this would also explain the light-mixed pigmentation of Celtic populations.

    And these late immigrants probably alsoo boosted the North European component in northwestern and central Europe.

    I would also point to mt-DNA haplogroup K2b. This was found in Bronze Age south Siberians. Obviously K isn’t an old hunter-gatherer haplogroup but originated probably in western Asia. Most K, at least K1 already entered Europe with the early Farmers. But later Siberian invaders may have brought K2b, which is now common in central, northern and western Europe.

    But there the question arises what about the Bell Beaker people of Kromsdorf? They were already R1b at a much earlier date?! Well, that is correct, but as I had pointed out above, R1b-L23* had a very wide, belt-shaped distribution that included central Europe and the Balkans. So it’s no surprise that it was found in German Bell Beaker people. But there is no evidence that these already spoke Italo-Celtic. And although they may have diffused their R1b somewhat further, there is no evidence that M412, L11 and the subsequent haplogroups descended from them.

    And what about the evidence of the complementary distribution of R1b and I2b in northwestern Europe? Well, according to linguists there were several different waves of Celtic in northwestern Europe, it may be simply the case that a later wave had more I2b than previous ones.

    And what about the negative correlation of R1b with IE languages in southwestern Europe? Well, I’ve once read somewhere that isolated, small populations are more susceptible to drift, and that mutations can get more easily fixed there. So R1b may have drifted more in the isolated Iberians and Basques.

  38. genetiker says:

    A lot of what Maciamo says comes out of David Anthony’s book, and the complete title of that book is “The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World”.

    To say that the horse domestication and horse riding argument is questionable is an understatement. The earliest representations of mounted warriors appear after 1200 BC, which is 2,000 years after Gimbutas’s Kurgan warriors from the eastern steppes are supposed to have invaded and conquered the rest of Europe. And if one defines domestication as human control over breeding, then the first evidence for the domestication of the horse is found at the Csepel-Haros site in Hungary, which was a Bell Beaker settlement.

    The Wikipedia article on the Kurgan hypothesis humorously says that “The question of further Indo-Europeanization of Central and Western Europe, Central Asia and Northern India during the Bronze Age is beyond its scope, and far more uncertain than the events of the Copper Age”, which is just an evasive way of admitting that it completely fails to explain the existence of IE languages in those areas.

    Let’s suppose that I, C, and R1b were equally common in Mesolithic Western Europe. With two samples, the probability of neither being R1b is 2/3 × 2/3 = 4/9 = 44.4%. So no, two samples are not enough to conclude that R1b wasn’t common in Mesolithic Western Europe. And even this calculation assumes that all haplogroups were uniformly distributed over Western Europe. If there was any kind of complexity in the haplogroup distributions, as is almost certainly the case, then we really can’t induce anything from the two samples that we have.

    Would you be willing to assert that I2a1a-M26 wasn’t common in Mesolithic Western Europe, since it didn’t show up in the two samples that we have?

    Would you be willing to assert that I1-M253 wasn’t common in Mesolithic Northern Europe, since it didn’t show up in the four samples that we have?

    All of the “ANE” calculations were based purely on Mal’ta 1. Mal’ta 1 shows North European Caucasoid, West Asian Caucasoid, and Veddoid components, so larger amounts of the North European component will cause an individual or population to appear to be more like Mal’ta 1, and larger amounts of the West Asian component will also cause an individual or population to appear to be more like Mal’ta 1. The same is true of the Veddoid component, but of course modern-day Europeans don’t have any of the Veddoid component, which illustrates how stupid all of the “ANE” calculations are. The North European component has obviously been in Europe since the Paleolithic, and the West Asian component apparently entered Europe after the early Neolithic. Mordovians and Russians appear to be a lot like Mal’ta 1 primarily because they have so much of the North European component, but also because they have some of the West Asian component, while Lezgins appear to be a lot like Mal’ta 1 primarily because they have so much of the West Asian component, but also because they have a significant amount of the North European component. The fact that Estonians and the Scottish ended up showing the most “ANE admixture” should tell everybody what complete garbage such calculations are, because there’s no coherent explanation of how two populations that are so different could have come out on top of the “ANE” ranking.

    If R1b originated in West Asia, then it would be associated with the West Asian component, but we know with certainty that it isn’t associated with the West Asian component at all, because Basques have one of the highest frequencies of R1b, and they don’t have any of the West Asian component. The only two components that R1b can possibly be associated with are the North European and Mediterranean components, and we already have strong evidence that it isn’t associated with the Mediterranean component, because so far the farmers have been G2a, F, and E1b1b, and not R1b.

    If we assume that I was the only European hunter-gatherer Y haplogroup, and that U and V were the only European hunter-gatherer mt haplogroups, then Basque haplogroups are 85% farmer and 15% hunter-gatherer, which is a 50 percentage point discrepancy from their 60% Mediterranean and 40% North European autosomal composition. But if we assume that R1b was also a hunter-gatherer haplogroup, then Basque haplogroups are 43% farmer and 57% hunter-gatherer, which is only a 34 percentage point discrepancy from their autosomal composition.

    The case of the Welsh gives even stronger evidence that R1b was a European hunter-gatherer haplogroup. If we assume that I and R1a were the only European hunter-gatherer Y haplogroups, then Welsh haplogroups are 90% farmer and 10% hunter-gatherer, which is a 105 percentage point discrepancy from the 35% Mediterranean and 60% North European autosomal composition of the British Isles. But if we assume that R1b was also a hunter-gatherer haplogroup, then Welsh haplogroups are 49% farmer and 51% hunter-gatherer, which is only a 23 percentage point discrepancy from the autosomal composition of the British Isles.

    The Wikipedia article on R1b says that R1b-M343* was found in the Kurds of Kazakhstan, but it fails to mention that the same spreadsheet that it cites for this shows that R1b-M343* was also found in West Germany, Slovakia, West Ukraine, and Romania.

    The article mentions that R1b1-P25* was found in Italy, and that R1b1-P25(xM269) was found in Ukraine.

    R1b1c-V88 is found not only in the Levant and in Africa, but also in Southern Europe.

    R1b1a1-M73 was found in some Russians in Europe.

    The Wikipedia article, citing the above spreadsheet, says that R1b1a2-M269* has a frequency of 7.9% in Kosovo, 5.1% in Macedonia, and 4.4% in Serbia, but it fails to mention that the spreadsheet shows that R1b1a2-M269* has a frequency of 5.3% in Germany, and that it was also found in Switzerland, Italy, Romania, Poland, Slovenia, Hungary, Northern Russia, and West Ukraine.

    R1b1a2a-L23* is found all over Europe, including Poland, Italy, Hungary, the Czech Republic, Denmark, Belarus, Romania, Sweden, Ireland, Slovenia, Slovakia, Ukraine, Croatia, France, Russia, Spain, Germany, Serbia, and Estonia.

    The highest frequencies of R1b1a2a1-L51* are found in France and Ireland, and it’s also found in Poland, Italy, Switzerland, Spain, Slovenia, West Germany, Hungary, Slovakia, Crete, and Russia.

    It doesn’t make any sense to say that R1b1a2a1a-L11* has “a distribution and a centre of weight that was clearly more to the east” when its highest frequency occurs in central England. In addition to Denmark and Poland, it’s also found in Switzerland, France, Germany, Spain, Crete, Estonia, Italy, Slovakia, Sweden, and Russia.

    Every step in the growth of the R1b tree, from the R1b-M343 root to the outermost leaves, is represented within Europe.

    The frequencies of the oldest R1b paragroups can give a completely misleading picture of their places of origin, because in a haplogroup’s place of origin new child clades are constantly being generated, which then partially or completely replace the older parent clades. The highest frequencies of the older parent clades can therefore occur outside of their places of origin.

    The first map on this page would lead one to believe that R1a-M420 originated simultaneously in Western Europe and the Middle East, and then spread from both locations to Eastern Europe, where R1a1-SRY10831.2 originated, which is of course absurd. The phenomenon I described above explains what really happened. R1a-M420 originated in Eastern Europe, spread from there to Western Europe and the Middle East, and was then replaced in Eastern Europe by its child clade R1a1-SRY10831.2, which originated in the same place that its parent did.

    It is not at all clear that Armenian was originally a satem language. The Wikipedia article on the centum-satem isogloss states

    “Satem-like features have arisen multiple times during history (e.g. French cent pron. [sã], Spanish ciento). As a result, it is sometimes difficult to firmly establish which languages were part of the original Satem diffusion and which were affected by secondary assibilation in a later time period. While extensive documentation of Latin and Old Swedish shows that the assibilation found in French and Swedish were later developments, there are not enough records of Dacian and Thracian to conclusively settle the issue of when their Satem-like features originated. Extensive lexical borrowing, such as Armenian from Iranian, may also add to the difficulty. The status of Armenian as a Satem language as opposed to a Centum language with secondary assibilation rests on the evidence of a very few words.”

    The idea that J2 men were the originators of the Anatolian languages doesn’t work, because none of the populations with the highest frequencies of J2 speak Indo-European languages.

    I already explained why Basques speak a non-IE language in spite of their high frequency of R1b.

    You can’t draw any conclusions about associations with pre-Roman languages in Iberia from slight variations in the present-day distribution of haplogroups there, because the distribution has changed over the 2,000 years since the Roman conquest. E, G, J, and T have entered Iberia from North Africa since then, and the highest frequencies of those haplogroups are in the areas where IE languages were spoken before the Romans.

    The non-IE languages in Iberia were spoken in exactly the same places where the Cardium Pottery culture had existed, and we know that the men of that culture were G2a and E1b1b, not R1b. The IE languages in pre-Roman Iberia were spoken in the areas that the Cardium Pottery culture did not expand to, and which would therefore have been inhabited by the indigenous hunter-gatherers.

    The time of the differentiation of Proto-Indo-European that has been inferred from the reconstructed vocabulary is erroneous, because the shared words that inference is based on are associated with technological innovations which spread over Europe after the initial differentiation, and which carried their names with them as they spread.

    The time of differentiation that has been calculated based on the rate of language change is also erroneous, because those calculations assume that the rate of language change is constant, which is a false assumption. Languages don’t just change for no reason. They change in response to changes in the lifestyle of those who speak them. The basic lifestyle of the hunter-gatherers of Europe changed very little over the tens of thousands of years of the Upper Paleolithic and Mesolithic, and their languages therefore also changed very little. The Neolithic brought about dramatic changes in the lifestyles of Europeans, and therefore caused European languages to change at a much higher rate than they had before. Change in European lifestyles and the concomitant change in European languages continued to accelerate in the Bronze Age, the Iron Age, and beyond. Calculations based on recent rates of language change will vastly underestimate the age of a proto-language, because rates of language change in the past were much lower than they have been recently.

    Parts of the Paleo-Sardinian language are preserved in Sardinian toponyms, and these toponyms show similarities to the Basque language and the Iberian language. And note that Ötzi, who was autosomally very similar to Sardinians, was G2a, not R1b.

    You can’t turn my argument about Bell Beakers, R1b, and the Kurgan hypothesis around, because in doing so you’re not basing your argument on any evidence, but are simply assuming that the conventional view of IE languages arriving to Iberia late is correct, and then reasoning from that assumption. The conventional view has no evidence to support it. There’s no direct evidence for what language the prehistoric people of Iberia spoke. There’s no direct evidence for what language any prehistoric people spoke, and there never will be any. The languages spoken by prehistoric people can only be determined by inference, by first using data from historical times to establish the relationships between haplogroups and languages, and then applying those relationships to the haplogroups found in prehistoric DNA. The historical data show without doubt that R1b is associated with the centum IE languages, and finding that the DNA of the people of a prehistoric culture, such as the Bell Beaker culture, is R1b allows us to infer that those people spoke languages that were ancestral to the historical centum IE languages.

    You keep saying that the Bashkirs had West Asian ancestors, but according to the Wikipedia article on the Bashkirs, they’re an amalgamation of Turkic Kipchaks, Ugrians, and speakers of Eastern Iranian languages, none of whom are West Asian.

    The analysis that globe13 is based on and this analysis didn’t produce anything like the Gedrosia component, at any K level, and those are superior analyses, because they’re based on larger datasets.

    The MDLP World-22 Indo-Iranian component is basically just a Kalash component. It makes up 84.4% of Kalash DNA, but for the people with the next largest amount of it, the Burusho, it only makes up 8.7%.

    The Gedrosia component indicates some kind of gene flow between Western Europeans and peoples of Pakistan, but it doesn’t tell us in which direction that gene flow occurred. You have assumed that the gene flow went from Pakistan to Western Europe, but there’s no genetic evidence to support that assumption. There is, however, genetic evidence that gene flow occurred in the opposite direction. Balochi, Makrani, and Pathans have some of the highest levels of the Gedrosia component, and the European haplogroup R1b-M269 is found in all of these populations.

    The way that people have interpreted the Gedrosia component in Western Europeans is reminiscent of the way that everybody interpreted the K = 4 Amerindian component in Northern Europeans. Everybody just assumed that it meant that Northern Europeans had Amerindian admixture, when there actually was never any genetic evidence for such an absurd notion. The Amerindian component in Northern Europeans was of course in reality a reflection of the Upper Paleolithic European admixture in Amerindians contributed by Y Q males.

    Your scenarios are contradicted by estimates of the ages of R1b subclades.

    It was estimated here that R1b1a2a1a1c-Z381, a subclade of the Germanic and Northwestern European R1b1a2a1a1-S21, originated 7,000 years ago. That’s long before the Copper or Bronze Ages.

    It was estimated here that the Western European R1b1a2a1a-L11 originated 6,100 years ago. That’s during the Megalithic culture of the Neolithic, long before the Copper Age Bell Beaker culture.

    It’s not accurate to say that the early European farmers were dark pigmented. The depigmentation allele in SLC24A5 was homozygous in Stuttgart. The depigmentation alleles in both SLC24A5 and SLC45A2 were homozygous in Ötzi.

    It’s an oversimplification to just say that the Kurgan people from the Pontic steppe were relatively dark pigmented. The frequencies of the depigmentation alleles were lower there than today, but those alleles were homozygous in some of the people, while being heterozygous or absent in others.

    And Motala 12 had at least one copy of the depigmentation allele in SLC24A5, so we know that allele was present in the hunter-gatherers of Mesolithic Europe.

    The depigmentation of Europeans didn’t proceed through massive population replacements of brown people in some regions by white people invading from other regions. It was a gradual process that took place all over Europe, from probably the Upper Paleolithic to the Bronze Age, in which the depigmentation alleles were driven to high frequency by sexual selection.

    You say that K2b “is now common in central, northern and western Europe”, but Eupedia says only that it “is found mostly in central and north-eastern Europe”.

  39. […] recent activity on this blog has been on this post from last year, in an exchange initiated by this comment. In my comments I have been arguing the […]

  40. True, Anthony’s book has „bronze-age riders“ in its title. If by this he refers to the Proto-Indoeuropeans and the Yamnaya culture, what would make sense, considering the topic of the book, then that phrasing is misleading and wrong. But of course, the steppe pastoralists reached the bronze age, too, eventually, so there is nothing inherently wrong with it.

    There is evidence for early horse domestication in the Botai culture, 3500 BC:
    http://dienekes.blogspot.com/2009/03/earliest-horse-domestication-in.html

    The Yamnaya culture sensu stricto expanded in southeastern Europe along the Danube and into the Carpathian basin, but not any further. It didn’t expand into modern day Poland and Germany. The Corded Ware can be regarded as another Kurgan culture, in the wide sense, and this extended in the west more or less to the Rhine. That’s where the scope of the Kurgan theory indeed ends, because the further expansion, even according to Kurgan theorists, wasn’t carried out by Kurgan cultures. However, I wouldn’t say that the Indo-Europeanization of western Europe during the Bronze Age would pose a particularly tricky problem, if central Europe was already Indo-European at that stage. Furthermore notice that, if you really favour a central / northern / western European homeland of the Indo-Europeans, then you’ll be facing exactly the same problems regarding the Indo-Europeanization of central Asia and northern India, as the Kurgan theory does.
    Let me add though, that I don’t subscribe to the Kurgan theory.

    [i]Let’s suppose that I, C, and R1b were equally common in Mesolithic Western Europe. With two samples, the probability of neither being R1b is 2/3 × 2/3 = 4/9 = 44.4%. So no, two samples are not enough to conclude that R1b wasn’t common in Mesolithic Western Europe. [/i]

    I don’t like this type of empty speculation. If there was at least phylogeographic evidence in favour of this hypothesis, it would make more sense, but the evidence points to a West Asian origin. Further, if R1b was from mesolithic European hunter-gatherers, why wasn’t it found in the four Scandinavian hunter-gatherers either? Why wasn’t it found in the countless early farmers, even though they had absorbed some hunter-gatherer DNA? Why wasn’t it found in two samples of the Megalithic, western European Seine-Oise-Marne culture dated to 2750 BC?

    I don’t know if I1-M253 was common in Mesolithic Northern Europe. It needn’t have been common, if it expanded strongly at a later date. But there can be little doubt that it originated in or somewhere near northern Europe.

    [i]Mal’ta 1 shows North European Caucasoid, West Asian Caucasoid, and Veddoid components, so larger amounts of the North European component will cause an individual or population to appear to be more like Mal’ta 1[/i]

    That is simply not true. La Brana was 72,7% North European in Globe13, yet he is nowhere near the ANE shift of modern Europeans.

    [i]If R1b originated in West Asia, then it would be associated with the West Asian component, but we know with certainty that it isn’t associated with the West Asian component at all, because Basques have one of the highest frequencies of R1b, and they don’t have any of the West Asian component. [/i]

    But the Basques have the MDLP World-22 Indo-Iranian component.
    Furthermore, if R1b wasn’t associated with some West Asian admixture, how do you explain that the Irish have 6,2% West Asian admixture in Globe 13, although only 1% of them have haplogroup J2?

    [i]If we assume that I was the only European hunter-gatherer Y haplogroup, and that U and V were the only European hunter-gatherer mt haplogroups, then Basque haplogroups are 85% farmer and 15% hunter-gatherer, which is a 50 percentage point discrepancy from their 60% Mediterranean and 40% North European autosomal composition. But if we assume that R1b was also a hunter-gatherer haplogroup, then Basque haplogroups are 43% farmer and 57% hunter-gatherer, which is only a 34 percentage point discrepancy from their autosomal composition.[/i]

    First of all, your calculation seems wrong, because mt-haplogroup V was from the farmers, too. H may be a hunter-gatherer halogroup, at least partly, and the Basques have lots of H.
    Second, you can’t calculate like this, because there is no strict link between uniparental markers and the autosomal make-up. If e.g. a group of successful males takes a lot of local women, the autosomal component originally associated with the male haplogroup will quickly be diluted. And that’s what happened in my opinion. R1b-carriers migrated from West Asia through central Asia and southern Siberia to Europe and along the way the West Asian component got diluted, and they picked up lots of North European admixture.

    Then, regarding the topic of phylogeography, of course the question isn’t just where a certain paragroup occurs, it matters also how common it is. Obviously an incidence of 10% cannot be compared to an incidence of 0,1%. Of course carriers of a certain paragroup may diffuse from their point of origin, so that in trace levels they may even be found far away from the point of origin.
    Regarding R1b-M343*, the spreadsheet you cited says that it’s most common in Kurds of Kazakhstan, with 13%. In West Germany it’s found in 1%. Considering the paleolithic age of R1b*, it’s quite possible that it was already present in some western European hunter-gatherers, but in trace levels at most, and it doesn’t seem to have originated there.

    What you failed to mention is the next study cited by the wikipedia article, which says that R1b-M343* was also found in 4,3% of the Persians from Yazd and in 3,2% of the Iranian Azeri Turks. That’s what I call significant.

    You say R1b1-P25* was found in Italy and in the Ukraine. Well, yes, in 0,26% of Italians and in 0,93% of Ukrainians…

    You say that R1b1c-V88 is found not only in the Levant and in Africa, but also in Southern Europe, and that R1b1a1-M73 was found in some Russians in Europe. True, but that doesn’t change the fact that these are essentially non-European branches. Some admixture in the fringe areas of Europe doesn’t surprise anyone. The wikipedia article goes on to say:

    „In 2010, Myres et al. report that out of 193 R-M73 men (…), all except two Russians occurred outside Europe, either in the Caucasus, Turkey, the Circum-Uralic and North Pakistan regions.“

    Regarding R1b1a2-M269*, I stand corrected. An incidence of 5,3% in Germany is indeed significant, and shows that already the precursor of L23* had reached central Europe. But still, it’s more common in Kosovo, with 7,9%, and it’s also found in the Ararat Valley in Armenia with 4,5%.

    The distribution of R1b1a2a-L23* is particularly interesting. It’s found in 67,9% of Bagvalals in the NE Caucasus, followed by the also NE Caucasian Tabassarans with 37,2%. Recently I noted that one of my grandmothers, who is of Swabian and Swiss extraction, has a significant Tabassaranian-like admixture according to the updated Eurogenes EU-test V2. Tabassaranian admixture is indicated both in the Oracle-4 and in the Oracle-X at GedMatch, it seems to be about 19%. Now the southern German/Swiss region also has quite an elevated West Asian admixture, but of the Caucasus-kind, not much Gedrosia. Phenotypically, Dinarid admixture strikes the eye. It’s well known that some of the Bell Beaker groups were strikingly Dinarid, especially in southern Germany, along the Danube, and this type continued to dominate there in the early Bronze Age. The early Bronze Age people of southern Germany were close to the local Bell Beaker people, according to multivariate analysis. Now, Dinarid-like people are common in West Asia, in Anatolia and the Caucasus, but also in Kosovo, of course. And it has been suggested that the domestic pottery of the Bell Beaker East group, along the Danube, is derived from Vucedol pottery, the Vucedol culture having been present in the Carpathian basin and surroundings, northern Croatia, etc. Frankly, I wouldn’t be surprised if the eastern Bell Beaker people were partly descended from Balkanic populations, and if they had carried some R1b and West Asian admixture into central Europe.

    R1b1a2a1a-L11* is most common in central England with 12%, followed by southern Switzerland with 6,3%, and by eastern Denmark and northern Poland, both with 5,9%. Apart from central England, this is rather central and northern Europe than western Europe, and the English are partly descended from Anglo-Saxons. But I’ll admit that its immediate precursor, R1b1a2a1-L51*, is again stronger in western Europe, with peaks in southeastern France and eastern Ireland. So already at this earlier stage R1b was rooted in western Europe.

    You seem to believe that basic haplogoups like R1b, or R1a, or I, etc, represent some kind of special, fundamental entities on their own. But this is not the case; these are defined by nodes in the tree of human y chromosomes which are completely arbitrary. R1b is no more a fundamental entity than, say, R1b1ba1a1, R1a1a1, MP or GHIJK is. For that reason one has to specify what exactly one is talking about when asking the question about its origin. While R1b is most common in western Europe now, this western European R1b belongs in the vast majority to clades that are much derived, while the older variants are found more often to the east, in southeastern Europe and western Asia.

    [i]The frequencies of the oldest R1b paragroups can give a completely misleading picture of their places of origin, because in a haplogroup’s place of origin new child clades are constantly being generated, which then partially or completely replace the older parent clades. [/i]

    This is faulty reasoning. Why should old paragroups of R1b be more endangered from replacement by other R1b clades than by any other haplogroup? Old paragroups of R1b could just as easily be replaced by J2 in the near East, for instance, as they could by R1b-S116.

    I agree that the map showing the presence of R1a-M420* in northwestern Europe is kind of puzzling. I can only surmise that it may be related to sampling bias. The source of these British, Irish and German R1a-M420* seems to be R1a.org. And probably much more northwestern Europeans have purchased sufficiently accurate y-DNA tests than Middle Easterners did. So in any case the many occurences of R1a-M420* in the Near East are worth more than those in northwestern Europe, because they were found with less DNA-testing. But as I admitted in the case of R1b*, considering the age, R1a* also may have been present at trace levels in western European hunter-gatherers.

    Regarding the case of Armenian and Satem, true, many linguists agree that Armenian’s closest relative is Greek, a centum language. In fact, linguists have long discarded the idea that the centum/satem dichotomy reflects a fundamental, deep split within Indo-European. The deepest split is now seen between Anatolian and the rest. But as a consequence it’s also questionable to assume a correlation of the R1a/R1b split with the centum/satem split. There is no way how the satem innovation could be as old as the R1a/R1b split. In my opinion R1a is most of all associated with the Balto-Slavic-Indo-Iranian branch, to which Germanic is partly related too.

    Kurds and Armenians have lots of J2 and are Indo-European speaking; other areas with high incidence of J2 have become Turkic speaking in the middle ages, but J2 certainly didn’t originate with the Proto-Turks in central and eastern Asia, it stems from the pre-Turkic subtrate, which partly used to be Indo-European. On a more fundamental level it has to be said that the idea that one haplogroup can only be associated with one language family is clearly wrong: R1b is common in Basques and in Celts. J1 is common in Semites and in Dagestanic peoples. Linguistically they are unrelated, while being dominated by the same haplogroup. Since this is evidently possible, it’s equally possible that two such linguistically unrelated groups expand and diversify into daughter languages. There is no way how you could exclude that this had once happened. Furthermore it’s also wrong to think that the speakers of a proto-language must have been pure in the y-chromosomes. There is absolutely no reason why they should have been pure. No modern population is pure in the y-chromosomes. And the speakers of a proto-language were not some special kind of people, they were just a group that expanded and subsequently split into several linguistic groups. So the idea that J2 must be either Indo-European or not, is wrong.

    [i]The time of the differentiation of Proto-Indo-European that has been inferred from the reconstructed vocabulary is erroneous, because the shared words that inference is based on are associated with technological innovations which spread over Europe after the initial differentiation, and which carried their names with them as they spread.[/i]

    But linguists can distinguish between borrowed words and cognates. Cognates are words that were inherited from the proto-language. They typically conform to the regular sound changes that distinguish the language in question from the proto-language.

    [i]Languages don’t just change for no reason.[/i]

    I’m not so sure about that. I have the impression that most of the language change is performed by the youth – for the simple reason that they want to be different from the grown-ups. Listen to teenagers of your country, you’ll often have the impression that they speak a different language than yourself.

    Of course there is evidence in favour of the assertion that Indo-Europeans were not living in Iberia since the paleolithic. The proto-Indo-European vocabulary is suggestive of a Copper Age culture, living either in West Asia or in central, southeastern or eastern Europe, depending on the authors. There is accumulating genetic evidence pointing to the Proto-Indo-European Urheimat, and this doesn’t point to western Europe.

    [i]The historical data show without doubt that R1b is associated with the centum IE languages, and finding that the DNA of the people of a prehistoric culture, such as the Bell Beaker culture, is R1b allows us to infer that those people spoke languages that were ancestral to the historical centum IE languages.[/i]

    In doing so, you assume that the carriers of a haplogroup can only be associated with one language family, and this is wrong, as I have explained above.

    Also it’s noteworthy that you have to explain the presence of the ancient Raetic language somehow. It’s related with Etruscan and Lemnian, pointing to a southeastern European origin. The latest possible arrival seems to be the eastern group of the Bell Beaker culture, which was continued by the early Bronze Age. The subsequent Tumulus culture definitely must have been Indo-European, and in my opinion it didn’t originate in southeastern Europe anyway. But the relationship between Raetic and Etruscan/Lemnian isn’t so distant that they could have split much earlier than the Copper Age. And then there are the southeastern relations of the eastern Bell Beaker people I mentioned above. And the observation that R1b is very common in the formerly Raetic areas in Tyrol and the Ladin areas in northern Italy.

    I was saying that the Bashkirs have West Asian ancestors because that’s where their R1b has originated in my opinion. Of course they are now ethnically Turkic. But as always, there was also a substrate that had been absorbed, especially in the topographically elevated Urals. I think partly they are Turkicized proto-Italo-Celts.

    Globe13 doesn’t have enough K to produce something akin to the Gedrosia component. Remember, K12b was focused on Eurasia and had 12 components, Globe13 deals with the whole world and has just one component more. But if you look at the Lazaridis et al. admixture analysis closely, it becomes clearer what this K12b Caucasus/Gedrosia distinction in Europe is about. At K=20, the Brahui (the peak of Gedrosia) have some admixture from the Kalash component, while the Georgians (the peak of Caucasus) don’t. And if you look at northwestern Europe at K=20, at the Orcadians and the Scottish, you see a thin, but consistent layer of the Kalash component. While in northeastern Europe, in the Lithuanians, Belorussians and Ukrainians, the Kalash component is had only by a few individuals. Of course, both Caucasus and Gedrosia in Europe are mostly about West Asian admixture. It’s just the finer distinction of the Kalash association of Gedrosia that made ADMIXTURE decide about the distribution of the Caucasus and Gedrosia components the way it did.

    And there you have the link to the MDLP World-22 Indo-Iranian component. Yes, it is a Kalash centered component, but just look at the map, some of it is in Europe, too, and its distribution in Europe resembles the Gedrosia component.

    It’s comparatively strong in the areas where Gedrosia peaks (Brahui, Makrani, Balochi):
    Kalash 84,4%
    Burusho 8,7%
    Pashtun 8,5%
    Balochi 6,9%
    Brahui 6,6%
    Makrani 6,6%

    And look at the European distribution, it surely resembles Gedrosia:
    (in %)
    Norwegian 3,8
    Orcadian 3,7
    Welsh 3,4
    Swedish 1,6 – 2,9
    British 2,8
    Uygur 2,8
    French 2,1
    Bashkir 2
    Sorb 1,8
    Austrian 1,8
    German North 1,6
    Mordovian 0 – 1,5
    Portuguese 1,5
    Czech 1,4
    Provencal 1,4
    Mari 1,4
    Slovakian 1,3
    Chuvash 1,3
    Slovenian 1,2
    Spaniard 1,2
    Basque 1,1
    Italian 0,4 – 1,1
    Russian 0,3 – 1,1
    Hungarian 1
    Greek 0,2 – 0,9
    Polish 0,4 – 0,9
    Serbian 0,9
    Ukrainian 0,4 – 0,9
    German South 0,8
    Finnish 0,3 – 0,7
    Kosovar 0,7
    Swiss 0,7
    Bosnian 0,6
    Sicilian 0,6
    Bulgarian 0,5
    Estonian 0,5
    Karelian 0,5
    Latvian 0 – 0,5
    Montenegrin 0,5
    Corsican 0,4
    Croatian 0 – 0,4
    Lithuanian 0,3 – 0,4
    Belarusian 0,4
    Moldavian 0,4
    Macedonian 0,3
    Cypriot 0,2
    Romania 0,2
    Saami 0
    Sardinian 0

    [i]The Gedrosia component indicates some kind of gene flow between Western Europeans and peoples of Pakistan, but it doesn’t tell us in which direction that gene flow occurred. You have assumed that the gene flow went from Pakistan to Western Europe, but there’s no genetic evidence to support that assumption. [/i]

    I don’t think so. If the European Gedrosia had originated in Europe, it would be present in either the European hunter-gatherers, or the early farmers. But La Brana, Ajv52, Ajv70, Ötzi and Gok4 all have 0% Gedrosia:

    The same holds true for the World-22 Indo-Iranian component:
    La Brana 1 + 2 = 0%
    Gok 4 = 0%
    Ötzi = 0%
    Ajv 52, Ajv 70 = 0%
    MA1 = 8,99%

    Also in the Lazaridis et al. admixture analysis, at K=20, none of the ancient European samples has any sign of the Kalash component, the only one who has it is MA1.

    There is no way to deny it: That Kalash/Gedrosia influence arrived later. And it cannot be explained by Caucasus admixture either. The World-22 Indo-Iranian component in the Caucasus:
    (in %)
    Tabassaran 3,8
    Lezgin 2,9
    Azeri 2,5
    Avar 2,1
    Chechen 2,1
    Kabardinian 1,3
    Ossetian 1,2
    North Ossetian 0,9
    Georgia 0,8
    Balkarian 0,7
    Adygei 0,4
    Armenian 0,6
    Cirkassian 0,3
    Abkhasian 0,2

    So, the Caucasus peak of that component is just as strong as its Scandinavian peak.

    [i]Your scenarios are contradicted by estimates of the ages of R1b subclades.
    It was estimated here that R1b1a2a1a1c-Z381, a subclade of the Germanic and Northwestern European R1b1a2a1a1-S21, originated 7,000 years ago.[/i]

    But notice the „important note“ in that link. The age estimates perhaps have to be halved, which would yield an origin around 1500 BC, consistent with my theory.

    [i]It was estimated here that the Western European R1b1a2a1a-L11 originated 6,100 years ago. That’s during the Megalithic culture of the Neolithic, long before the Copper Age Bell Beaker culture.[/i]

    If that age estimate is correct (there are still uncertainties), then there is a problem with my theory. Perhaps a way to resolve it would be to turn it around: That R1b-M412* and the subsequent haplogroups evolved from the R1b that was already present at Bell Beaker times. And that the later newcomers from Siberia only brought some additional R1b-L23*. The Kalash/Gedrosia/Indo-Iranian admixture in Wales can hardly be explained with R1a.

    [i]It’s not accurate to say that the early European farmers were dark pigmented. The depigmentation allele in SLC24A5 was homozygous in Stuttgart. The depigmentation alleles in both SLC24A5 and SLC45A2 were homozygous in Ötzi.[/i]

    I didn’t mean the skin colour, but the hair and eyes.

    Regarding the distribution of mt-haplogroup K2b, you can study the various branches here:
    https://www.familytreedna.com/public/mtDNA_K/default.aspx?section=mtmap

  41. I’ve changed my mind and given up that theory of the South Siberian origin of Italo-Celtic. And I agree now that the eastern Bell Beaker group probably spoke a precursor of Celtic. This is what I wrote on Dienekes Blog:

    Genetiker has alerted me of two things: R1b-M269* is also present in 5.3% of Germans, so it had already reached central Europe, too. And, more importantly: Current age estimates of R1b-L11 place it at 6.1 kya. This would strongly contradict my theory that it is descended from south Siberians who arrived in central Europe during the middle Bronze Age. There is also another, a linguistic objection: The diversity of Italo-Celtic languages makes it hard to believe that they reached central Europe as late as 1600 BC. There were two quite different branches in Italy, a p-Italic (Oscan-Umbrian) and a q-Italic (Latino-Faliscan), similar to the disctinction of p- and q-branches in Celtic. Plus there were related languages that diverged even more, like Ligurian and Lusitanian. So that theory of mine isn’t really tenable.
    However, recently I noted that the difference between the Gedrosia admixture and the Caucasus admixture in Europe is mainly a slight Kalash relatedness of the Gedrosia component. And this explains its similarity with the MDLP World-22 Indo-Iranian component, which is mainly a Kalash component. Now, it occured to me that this influence seems correlated with R1b-L23. In the Caucasus the World-22 Indo-Iranian component is strongest in Tabassarans (3.8%), followed by Lezgins (2.9%). The Tabassarans have 37.2% R1b-L23*, and the Lezgins have 12.9%. The Indo-Iranian component is also relatively strong in Bashkirs (2%), and they have 32.2% L23*. In Europe, the Indo-Iranian component is strongest in Norwegians (3.8%), Orcadians (3.7%) and the Welsh (3.4%). They all have lots of L23, especially the Welsh, though mostly more derived. Notably the Indo-Iranian component is also present in the Basques (1.1%) which parallels the Gedrosia presence there. And BTW, this all parallels and explains the mysterious Dagestan component of the early Dodecad experiments. So, I would suggest that both R1b-L23* and some slight Kalash-like admixture entered southeastern Europe early, from West Asia / eastern Anatolia, presumably with the chalcolithic Baden culture. Presumably this didn’t carry yet a lot of general West Asian admixture. It has been suggested that the domestic pottery of the eastern Bell Beaker group is derived from Vucedol pottery, and this is derived from Baden. So it’s all connected: R1b, Gedrosia, Bell Beaker pottery and most probably also the Dinaric Bell Beaker skulls. And, most likely Celtic languages. Bell Beaker-like culture and people continued to dominate southern Germany in the early Bronze Age. In the Middle Bronze age they were pushed to the west, by the invasive Tumulus culture which brought more Germanic-like languages.

  42. genetiker says:

    The zoological definition of domestication requires human control over breeding, and such control is evidenced in the archaeological record by a decrease in the average size and an increase in the variability of the skeletal remains of animals. The first time those changes are seen in horse remains is 2500 BC, at the Bell Beaker Csepel-Haros site in Hungary. They appear later in Bronze Age sites in the Russian steppes, Spain, and Eastern Europe.

    I don’t face any problems regarding the Indo-Europeanization of any place, because I don’t tie the Indo-Europeans to any particular material culture after the Gravettian culture. After the Gravettian I tie them only to R1b and R1a.

    Ajvide 52 was V, so V was definitely a hunter-gatherer haplogroup. A Gravettian sample from Italy and a Magdalenian sample from Spain were found to be either R0 or HV, and those are ancestral to V, so it’s certainly possible that V is ultimately descended from Paleolithic Europeans. Sami are 41.6% V, and Sami have an especially close relationship to the hunter-gatherers of Paleolithic and Mesolithic Europe. Also, neither U nor V appear in any of the samples from the Neolithic Near East listed here. And as reported here, X, K, J, H, and T were found in samples from Mesolithic and Neolithic Greece, but no U or V was found there.

    If you were to assume that V was a farmer haplogroup, it would only make my argument that R1b was a hunter-gatherer haplogroup stronger. In that case, the discrepancy between Basque haplogroup frequencies and their autosomal composition would be 58 percentage points if R1b was a farmer haplogroup, and only 26 percentage points if it was a hunter-gatherer haplogroup.

    H is far more strongly associated with the farmers than with the hunter-gatherers. None of the Swedish hunter-gatherers were H. Three out of four of the Swedish farmers were H. Of the many Mesolithic European samples listed here, the vast majority are U, and only one is H. In this large study of ancient European mtDNA, all of the hunter-gatherers were U; none were H. H was found in the samples from the Neolithic Near East mentioned above, and it was also found in the samples from Mesolithic and Neolithic Greece mentioned above.

    The autosomal composition of Sardinians doesn’t make any sense if H was a hunter-gatherer haplogroup. If we assume that the European hunter-gatherer Y haplogroups were I, R1b, and R1a, and that the European hunter-gatherer mt haplogroups were U, V, and H, then Sardinian haplogroups are 41% farmer and 59% hunter-gatherer, which is a 73 percentage point discrepancy from their 71% Mediterranean and 16% North European autosomal composition. But if we assume that H was a farmer haplogroup instead of a hunter-gatherer haplogroup, then Sardinian haplogroups are 63% farmer and 37% hunter-gatherer, which is only a 29 percentage point discrepancy from their autosomal composition.

    I of course do not think that haplogroup frequencies and autosomal component percentages are directly proportional to each other, but there are correlations between them, and calculations like the ones above are therefore useful in testing proposed relationships between haplogroups and autosomal components.

    The 13% frequency of R1b-M343* in the Kurds of Kazakhstan was only 3 out of 23 people. It would have been a much more significant finding if it had been 130 out of 1,000, or 13 out of 100.

    The Iranian study genotyped P297, which defines R1b1a, but it didn’t genotype P25 or any of the equivalent SNPs that define R1b1, so there’s no way of knowing whether what it found was R1b-M343* or R1b1-P25*. The study that found R1b-M343* in West Germany, Slovakia, West Ukraine, and Romania did genotype M415, which is equivalent to P25, so we know that it really was R1b-M343*.

    Older clades of a haplogroup being replaced by younger clades of that same haplogroup is a very different phenomenon from clades of one haplogroup being replaced by clades of another haplogroup. The latter necessarily involves competition between two different groups of people. The former needn’t involve any competition at all; it can simply be the natural process of newer generations replacing older generations within the same group of people.

    Imagine drops of water falling and hitting a pond at a particular point, and ripples then radiating outward from that point. The oldest ripples will be found furthest from the common point of origin. This is analogous to ever newer subclades being generated in a haplogroup’s place of origin, and those subclades then spreading outward to other regions. The highest frequencies of the oldest subclades can end up being found far from their place of origin, because the newer subclades haven’t had time to spread outward and replace them.

    One can either accept or reject the proposition that a haplogroup originated in the place where the highest frequency of the paragroup for that haplogroup is now found.

    If one rejects this proposition, then one must acknowledge that there are no grounds for the prevailing view that R1b originated in West Asia as opposed to Europe.

    On the other hand, if one accepts this proposition, then the prevailing view of Indo-European origins, the Kurgan hypothesis, must be false.

    The above proposition implies that R1b1a2a1-L51 originated in either France or Ireland. R1b-L51 is estimated to have originated between 7,600 and 5,900 years ago. Since R1b-L51 men were the originators of the Germanic and Italo-Celtic languages, a language ancestral to those languages must have been spoken in France or Ireland between 7,600 and 5,900 years ago, which completely contradicts the Kurgan hypothesis.

    The above proposition implies that R1b1a2a1a-L11 originated in central England. R1b-L11 is estimated to have originated 6,100 years ago. Since R1b-L11 men were the originators of the Germanic and Italo-Celtic languages, a language ancestral to those languages must have been spoken in central England 6,100 years ago, which completely contradicts the Kurgan hypothesis.

    Whether one accepts or rejects the above proposition, one is forced to reject a majority viewpoint regarding European prehistory.

    There are no one-to-one relationships between languages and haplogroups today, but things were very different many thousands of years ago, when the world’s major language families were coming into being. The complexity we see today is the result of the innumerable population movements of the past few thousand years, and the subsequent intermixing of all of the many different populations. Prior to those events, populations were much more isolated from one another, and they were much more homogeneous. We can see this for example in the mtDNA of Paleolithic and Mesolithic Europeans, which was almost entirely U. Compare that to the complexity that we see in the mtDNA of Europeans today. It is likely, therefore, that most languages were originally spoken by people belonging to only a single Y haplogroup.

    The phenomenon that one sees today in the West of each new generation trying to be as different as possible from the preceding generation is an anomaly. That becomes obvious when one compares modern Westerners with one of the few remaining hunter-gatherer populations, such as the San bushmen. The hunter-gatherer societies of the Paleolithic and Mesolithic were far more conservative than what one sees today in the West.

    As I said above, change in human lifestyles has been accelerating over time. This accelerating change can be seen in human technology. In the Upper Paleolithic, a single type of technology would be used for many thousands of years, whereas today countless new technologies appear every year. Accelerating change can also be seen in music. Today many new musical styles appear every decade, whereas some styles of classical music existed for over 100 years. The same accelerating change occurred in human languages, and any calculation that fails to take that into account will vastly underestimate the age of a proto-language.

    The claim that proto-Indo-European differentiated during the Copper Age is based primarily on the fact that words related to the wheel in the different IE languages are cognates. First, it has to be noted that there are, not one, but three different reconstructed proto-IE words for the wheel, which is not what one would expect if proto-IE had still existed at the time the wheel was invented. Second, the idea that the existence of these cognates implies a Copper Age differentiation is simply false, because it’s possible that the cognates are descended from words that originally meant ‘circle’, ’round’, ‘turn’, ‘roll’, or ‘rotate’, and which were only later adapted for use as words for ‘wheel’.

    I’ll have something to say about the Gedrosia component in my upcoming posts.

  43. @ Genetiker
    I’ve always oscillated between associating the early Indo-Europeans with (part of) the North European component on the one hand or with (part of) the West Asian component on the other. This brandnew study makes me seriously doubt the theory of the West Asian origin of the Indo-Europeans:
    http://dienekes.blogspot.com/2014/05/ancient-dna-from-balkans-iron-age-thrace.html

    But if they are associated with North European admixture, they must be derived from eastern Europe and/or central Asia – not from central or western Europe. There are several lines of evidence in favour of this view. Perhaps the most important piece of evidence is the ANE admixture. You have claimed that it is nothing other than the West Asian admixture. But, as I’ve pointed out, the Scottish are among the most ANE admixed populations of Europe, they have more of it than the Lithuanians for instance. Yet you obviously cannot say that the Scottish are among the most West Asian admixed populations of Europe. Also, this ANE admixture wasn’t present in the western European hunter-gatherers. Some of it was present in Scandinavian hunter-gatherers, but not enough. Most importantly, it wasn’t present in German hunter-gatherers, because the Loschbour hunter-gatherer had zero ANE admixture, and he was from Luxemburg, which is immediately near Germany. Then there is the mt-DNA evidence, provided by the Brandt, Haak et al. 2013 study: In addition to the hunter-gatherer and early farmer mt-DNA, there is a late Neolithic / early Bronze Age component composed of haplogroups I, U2, T1 and R, which links it with eastern hunter-gatherers and the Caucasus area. The Yamnaya culture had 20% of these, and both the Corded Ware and the eastern Bell Beaker people had 13.8% of them (according to my own counting). As a patriarchalic society, the early Indo-Europeans had probably much more influence on the paternal side, the y-DNA. Both R1a and R1b probably came from the east too, and mark the trace of the early Indo-European and ANE admixed North European influence. There is zero evidence for a west European origin of R1b, and its first appearance in central Europe is documented in Bell Beaker individuals, which are obviously neither Paleolithic nor Mesolithic. I also think that the MDLP World-22 Proto-Indo-Iranian component traces some of this eastern European or central Asian early Indo-European influence, and I still think there is a relationship with R1b-L23. This „Proto-Indo-Iranian“ component (it isn’t exclusively linked to Indo-Iranians) is distinct from the West Asian component. It peaks in the Kalash in the secluded Hindukush region of northwestern Pakistan. The Kalash have been noted for their often European-like looks. I think the component has been more common in central Asia in prehistoric times and its current concentration in the Kalash is due to a reduction to a refuge area. The Indo-Iranian component isn’t present in western European hunter-gatherers and neither in early European farmers. It must have arrived later and from the east. It follows the route of the early R1b-Indo-Europeans, and its signal is also present in the Basques, which causes them to have some Gedrosia-admixture even though they have no West Asian admixture. But the Basques are also somewhat ANE admixed, which fits well with their R1b and „Indo-Iranian“ admixture. As I said, in the Caucasus, the Tabassarans and (to a lesser extent) the Lezgins are a hotspot for both R1b-L23 and the Proto-Indo-Iranian component. It’s possible that they have received this from early Indo-European neighbours to the north of them. After all, they are in Dagestan, in the northeastern Caucasus.

    Indeed, controlled breeding is crucial for domestication, otherwise we are just dealing with taming, which is not the same. But note that according to the link I provided, metrical analysis of horse metacarpals showed that Botai horses resemble Bronze Age domestic horses rather than Paleolithic wild horses from the same region.  This is evidence for controlled breeding a 1000 years earlier than at the Csepel-Haros site.

    You used the doubts of some archeologists with regards to Indoeuropean invasions of western Europe etc. as an argument against the Kurgan theory. If you’re now saying that archeological evidence doesn’t matter, you invalidate your own argument. Moreover I would say it isn’t advisable to ignore the archeological evidence. For the question of Indo-European origins we need everything, linguistics, genetics, archeology and physical anthropology. Granted, the Gravettian was very widespread, and if the Gravettian people were the Proto-Indoeuropeans, then we don’t need a lot of subsequent migrations, except into South and West Asia. But the idea that Europe was already packed with R1a and R1b people during the Upper Paleolithic is seriously undermined by the ancient DNA evidence. Also there is evidence that R1b experienced a rather recent episode of rapid expansion:
    http://dienekes.blogspot.com/2014/02/recent-radiation-of-r-m269-males-in.html

    I still think your theory of old paragroup replacement by younger clades isn’t logical. Old haplogroups don’t magically transform into younger clades. Each mutation giving rise to a younger clade occurs only once. I think the loss of one single male should be negligible. It surely occured often that males died without leaving any offspring. So then we’ve got one male having the new mutation and all the other ones who don’t have it. Drift and founder effects may explain how the offspring of a younger clade become frequent. But there is no a priori reason why drift and founder effects should favour any particular haplogroup or subclade. Nor is it likely that an old paragroup will (almost or completely) disappear from the population because of drift, if it used to be very frequent.

    The origin of a haplogroup isn’t necessarily identical with the place where the highest frequency of the paragroup for the haplogroup is found – there may have been serious upheavals, invasions, emigrations and population changes – but it’s more likely closer than farther away from it.
    The same applies to R1b1a2a1-L51. I don’t believe that it originated in France or Ireland, but close to it, definitely in Europe. And in southeastern France, where its paragroup is particularly freuquent, there were the Ligurians, an old linguistic „paragoup“ of Italo-Celtic. And Gaelic, the native language of Ireland, is a q-Celtic branch, which is older than p-Celtic.

    As for R1b1a2a1a-L11, I’ve said before that the distribution of the paragroup reminds me of the eastern Bell Beaker culture. It’s older though, but maybe it just became common at the time of the Bell Beaker culture.

    On a sidenote, I would be careful not to simply equate the Germanic branch with R1b. Linguistically, Germanic is inbetween Italo-Celtic and Balto-Slavic, and R1a is not at all uncommon in the old Germanic area. It’s probably a fusion of both influences.

    Various aspects of a language may change over time, and not all of them do so proportionally to changes in lifestyle. But the Bayesian approach by Gray and Atkinson works with word replacement only, and it seems plausible that this is sensitive to changes in technology and lifestyle. Their model doesn’t work with a fixed pace, however; instead, they exploited multiple known calibration points (e.g., the breakup of Romance languages in Late Antiquity) which resulted in different rates of word retention in the different branches. It seems plausible that the pace of word replacement was slower in the Mesolithic than in Neolithic societies. But first the tree has to reach back into the Mesolithic, in order for Mesolithic rates of word replacement to become applicable. However it doesn’t seem to reach back into the Mesolithic.

    But anyway, I put much more weight on the evidence from linguistic paleontology, And there is much more than just the case of the words for wheel, see here, page 40-43:

    http://books.google.ch/books?id=bSxHgej4tKMC&printsec=frontcover&hl=de#v=onepage&q&f=false

  44. genetiker says:

    David Reich’s “ANE” neologism is idiotic.

    Y hg N and O Mongoloids were “Ancient North Eurasians”, but they didn’t have any close genetic relationship to the Caucasoid and Y hg R* Mal’ta 1.

    Y hg I Caucasoids were “Ancient North Eurasians”, and although the fact that the North European component is modal for both them and Mal’ta 1 does show that there was some kind of gene flow between them, the fact that Mal’ta 1 had a large amount of the Gedrosia component while Y hg I people didn’t have any of it shows that they were also genetically distinct.

    For my recent comments on the West Asian, Caucasus, Gedrosia, and Indo-Iranian components, see this post. I had already addressed some of the subjects that you later raised in that post.

    It was always a mistake to look at the West Asian component, because it lumps together the Caucasus and Gedrosia components, which flowed into Europe from different sources. The Gedrosia component was spread across Europe primarily, and perhaps entirely, by R1b people. The Caucasus component was brought to Europe primarily by J2 people, but there does also seem to be an association between R1a and the Caucasus component.

    Estonians had even more “ANE admixture” than the Scottish, which as I pointed out above should tell everybody that the “ANE” calculations are garbage.

    The North European component was found throughout what was the steppe-tundra region during the Upper Paleolithic, from La Braña 1 in Iberia to Mal’ta 1 in Siberia. So we know that there was gene flow throughout this region. Genetic distance between individuals in this region would have been proportional to geographical distance between individuals, just as is the case today, and there therefore would have been a genetic continuum that spanned the region. La Braña 1 and Loschbour were at one end of this genetic continuum, and Mal’ta 1 was at the other end, so of course ADMIXTUREGRAPH would not show them as being mixtures of each other. Motala 12 in Sweden occupied a geographically intermediate position between La Braña 1 and Loschbour, and Mal’ta 1, and he therefore would have occupied a genetically intermediate position between them as well. So of course ADMIXTUREGRAPH would show him as being a mixture of Loschbour (“WHG”) and Mal’ta 1 (“ANE”). This trivial result has apparently led a lot of people to erroneously believe that the Scandinavian hunter-gatherers must have had some admixture from Y hg R people. My analyses of the Stora Förvar and Ajvide genomes showed that the Scandinavian hunter-gatherers cannot have had any admixture from P, Q, R*, R1*, or R1b people, because they didn’t have any of the Gedrosia component. The possibility that they had some admixture from R1a people may still be open, but even this is doubtful, because my analyses of the Stora Förvar and Ajvide Y chromosomes showed that none of them were P, Q, or R. And if R1a is associated with the Caucasus component, as it appears to be, then the autosomal evidence also precludes the possibility that the Scandinavian hunter-gatherers had any admixture from R1a people, because they didn’t have any of the Caucasus component.

    The Gedrosia component is a distinguishing characteristic of the genetic signature of P, Q, R*, R1*, and R1b people, and we should not expect to see it in individuals that belonged to other haplogroups, like I and C. When we get autosomal DNA from prehistoric R1b people, then we will see the Gedrosia component. The absence of the Gedrosia component in La Braña 1 does allow us to conclude that he didn’t have any admixture from R1b people, but it does not allow us to conclude that there weren’t any R1b people in Western Europe during the Mesolithic, because it’s possible that there were R1b people there, but that they didn’t mix into the Y hg C population.

    I think that the account of European genetic prehistory here is basically correct. It shows R1b occupying an Iberian refuge, I occupying a Balkan refuge, and R1a occupying a Ukrainian and Southern Russian refuge during the last glacial period, and each of these haplogroups then spreading from these refuges starting at the beginning of the Holocene.

    The mtDNA study you cited actually supports my position, not yours. The genetic distance maps from that study showed that the mtDNA haplogroups associated with the Corded Ware culture now have a distribution that is similar to that of R1a, and we know that R1a was associated with the Corded Ware culture. But those maps also showed that the mtDNA haplogroups associated with the Bell Beaker culture now have a distribution that is very different from that of the haplogroups associated with the Corded Ware culture. The distribution of the Bell Beaker haplogroups is similar to the distribution of R1b, not R1a. This shouldn’t come as a surprise, because we know that R1b was associated with the Bell Beaker culture. So the mtDNA evidence shows that there were two distinct populations: one associated with R1b and the Bell Beaker culture, and expanding from Western Europe, and another associated with R1a and the Corded Ware culture, and expanding from Eastern Europe. Since R1b men were the originators of the centum Indo-European languages, and R1a men were the originators of the satem Indo-European languages, this means that the centum Indo-Europeans are derived from Western Europe, while the satem Indo-Europeans are derived from Eastern Europe.

    The genetic distinctness of the R1b and R1a populations that is implied by the mtDNA evidence is also shown by the autosomal evidence. R1b is strongly associated with the Gedrosia component, while R1a doesn’t appear to be associated with the Gedrosia component at all. Belarusians and Lithuanians have among the highest frequencies of R1a, but they don’t have any of the Gedrosia component. On the other hand, R1b shows no association with the Caucasus component at all, while R1a does appear to be associated with the Caucasus component. Basques have one of the highest frequencies of R1b, and while they have 9.8% of the Gedrosia component, they don’t have any of the Caucasus component. J2 is completely absent in Lithuanians, and yet they have 8–10.1% of the Caucasus component.

    For the genetic distinctness of the R1b and R1a populations to have emerged, they must have been geographically separated for a long period of time. The simple fact that R1 split into R1b and R1a around 25,000 years ago itself suggests such a deep geographical separation. And the different autosomal components that characterize the R1b and R1a populations suggest where these populations resided during their separation. Western Europe, at the westernmost end of the Eurasian landmass, would have been the perfect place for the R1b population to preserve the original genetic makeup of R* people, which was characterized by a large amount of the Gedrosia component. The association between R1a and the Caucasus component strongly supports the idea, mentioned above, that the R1a population resided in Ukraine and Southern Russia, just north of the Caucasus Mountains, during the last glacial period.

    There is in fact a very large piece of evidence for a Western European origin of R1b. The frequency of R1b is highest along the Atlantic coast of Europe, and there is a decreasing frequency gradient from west to east, which is the complete opposite of what one would expect if it had originated in Eastern Europe.

    The Indo-Iranian component is like the Gedrosia component. La Braña 1 didn’t have any of it because La Braña 1 wasn’t R1b. When we get autosomal DNA from prehistoric R1b people, then we will see the Indo-Iranian component.

    There may be evidence for horse domestication in Central Asia at 3500 BC, but the first evidence for horse domestication in Europe is at the Bell Beaker Csepel-Haros site. Not at a Yamna site. Not at a Corded Ware site. At a Bell Beaker site. Such evidence is then found later in the Russian steppes, Spain, and Eastern Europe.

    As I noted above, the earliest representations of mounted warriors appear after 1200 BC, which is 2,000 years after the invasions from Eastern Europe are supposed to have occurred according to the Kurgan hypothesis. Horses were used to pull chariots before there were mounted warriors, but spoked wheels and chariots weren’t invented until 2000 BC, which is 1,000 years after the supposed invasions. And we know that the proto-Indo-Europeans didn’t have chariots with spoked wheels, because there was no word for ‘spoke’ in the proto-Indo-European vocabulary.

    We know that the Indo-Europeans did not originate from Central Asia, because the proto-Indo-European vocabulary contained words for ‘bee’ and ‘honey’, and honeybees were not native east of the Ural Mountains.

    You simply cannot claim that the presence of R1b and R1a in Upper Paleolithic Europe is undermined by the prehistoric DNA evidence, because there are huge gaps in the archeological record, and even bigger gaps in the prehistoric DNA record. According to this article on the origins of burial, there are no convincing burials from the Aurignacian, then there are a few burials from the Gravettian, and then there’s another absence of burials from the end of the Gravettian until 15,000 years ago. The article notes that almost all of the individuals from the Gravettian burials were atypical in that they had pathological features. The article says that

    “What is clear, however, is that burial was never the norm for ‘ordinary’ people. We have to assume that most people were disposed of – as perhaps they had been for hundreds of millennia – in ways that were reverent and ritualistic but which are now archaeologically invisible.”

    The Solutreans must have used such an archeologically invisible way of disposing of their dead, because no Solutrean burials have been found.

    Since burial was so uncommon for most of the Upper Paleolithic, the few known Gravettian burials may provide the only prehistoric DNA evidence we’ll ever get that R1b and R1a have been present in Europe since the Upper Paleolithic.

    Of course R1b experienced a recent episode of rapid expansion. Every haplogroup experienced an episode of rapid expansion when the people of that haplogroup transitioned from a hunting and gathering lifestyle to a farming lifestyle.

    In arguing against the replacement of very old clades of haplogroups, you’re again projecting recent dynamics back onto a time when those dynamics did not exist. It is indeed unlikely that newer child clades of a haplogroup would completely replace older parent clades of that haplogroup in times since the Neolithic, because population sizes have been so large since the Neolithic. Population sizes were far smaller during the Upper Paleolithic, when R1b and its very oldest subclades originated. The hunter-gatherers of the Upper Paleolithic and Mesolithic lived in bands of 200 or fewer individuals. For all we know, R1b could have been represented in Upper Paleolithic and Mesolithic Western Europe by only one such band of people. It would have been very easy for older clades of R1b to be completely replaced by newer clades in such a band of people.

    We now know for a fact that R* was present in Siberia 24,000 years ago. But no R* has been found in the present-day populations of Siberia. So what happened? The R* that was present in Siberia 24,000 years ago has since been completely replaced by newer subclades of R, and also by other haplogroups. Since we know that complete replacement of old clades has happened at least once before, it’s impossible to assert that R1b* and other early R1b subclades were never present west of West Germany.

    The older a haplogroup is, the more possible it is for the highest frequency of the paragroup for that haplogroup to be found far from the place where that haplogroup originated. R1b-M343 is over 20,000 years old, so R1b-M343* has had more than 20,000 years to wander from its place of origin and rise to high frequency in faraway places. R1b1a2a1a-L11 is only 6,100 years old, so R1b1a2a1a-L11* has had much less time to wander than R1b-M343*. We can therefore be much more certain in inferring the place of origin of R1b1a2a1a-L11 from the frequency of R1b1a2a1a-L11* than we can be in inferring the place of origin of R1b-M343 from the frequency of R1b-M343*.

    As I pointed out above, looking at R1a-M420* would lead one to conclude that R1a-M420 originated simultaneously in Western Europe and the Middle East, which is absurd. So we know that looking at the frequency of a very old paragroup can lead to incorrect conclusions about the place of origin of the corresponding haplogroup.

    There is no doubt that R1b people were the originators of the Germanic languages. We know that I people didn’t originally speak Indo-European languages, because the branching of I doesn’t parallel the branching of the Indo-European languages the way that the branching of R1 does. Another piece of evidence that I people were not the originators of the Indo-European languages is that R1a goes down into India, while I does not. The frequency of R1b is higher than the frequency of R1a in all of the areas where Germanic languages are spoken. No subclade of R1a is found in all of the places that Germanic people have migrated to. The Germanic languages are centum languages, like all of the other languages originated by R1b people, and unlike all of the languages originated by R1a people.

    I acknowledge that some of the words in the proto-Indo-European vocabulary must have had a Neolithic origin. I think that the Gravettian language can be thought of as pre-proto-Indo-European. I think that this language evolved into proto-Indo-European during the Neolithic. Very little evolution would have been required for pre-proto-Indo-European to become proto-Indo-European, because only a small fraction of the proto-Indo-European vocabulary refers to things of Neolithic origin. Most of the proto-Indo-European vocabulary refers to things that would have been familiar to the Gravettians. I think that proto-Indo-European then underwent differentiation within the different groups of R1 people during the Neolithic.

    The above implies that the differentiation of Indo-European did not occur in synchrony with the differentiation of the R1 haplogroup. We should not be surprised by this, because differentiation within R1 would have occurred any time there was a split in an R1 population, while differentiation within Indo-European was driven by changes in lifestyle, which accelerated over time. The further back in time we look, the greater this asynchrony was. R1 split into R1b and R1a 25,000 years ago, but very little change in the pre-proto-Indo-European language would have been occurring at that time. But as we move forward in time, we begin to see more synchrony between the two differentiations. The TMRCA for R1b-L23 is estimated to be 8,000 years. R1b-L23* people were the originators of the Anatolian languages, and the Anatolian languages have been estimated to have branched off from the other Indo-European languages around 8,000 years ago. The TMRCA for R1b-L11 is estimated to be 6,100 years, and the TMRCA for R1a-Z645 is estimated to be 6,000 years. R1b-L11 people were the originators of the Germanic and Italo-Celtic languages, and R1a-Z645 people were the originators of the Balto-Slavic and Indo-Iranian languages, and the non-Anatolian Indo-European languages have been estimated to have branched off from each other around 6,000 years ago.

  45. The PCA in Lazaridis et al. 2013 shows that most modern Europeans are placed inbetween La Brana/Loschbour, the Early Farmers, and Mal’ta 1/Afontova Gora 2. And the diagram illustrating the calculated „ANE admixture“shows that even populations in the far northwest of Europe, such as the Scottish, are considerably „ANE admixed“. Even if ANE admixture isn’t really admixture, as you maintain, but instead relates to a more eastern origin on the continuum between La Brana/Loschbour and Mal’ta, we’ll nonetheless arrive at the conclusion that modern Europeans, also modern western Europeans, are not purely the descendants of Early Farmers and Loschbour/La Brana-like hunter-gatherers, but instead also have ancestry from a group with more eastern roots.
    And since the Gedrosia component is so strongly present in Mal’ta 1 and Afontova Gora 2 while being at the same time completely absent in the western La Brana sample, it makes sense to assume that Gedrosia admixture is linked with increased ANE-like ancestry and thus arrived at the same time as the above mentioned eastern group.

    The question is: When did this eastern group arrive? I have admitted before (not on this site I think) that it’s of course possible that more ANE-like groups lived in western Europe already during the Mesolithic. I just find it unlikely that Mesolithic hunter-gatherers somewhere west of Loschbour and west of the Scandinavian hunter-gatherers were genetically more eastern than either of them. And they would have to have been oddly isolated in their hypothetical western pocket, given that the rather central European Loschbour hunter-gatherer was genetically so strongly western. Also an Iberian origin can be ruled out, otherwise we’d see some of this influence in La Brana. That hunter-gatherer groups on the Iberian peninsula didn’t mix with each other isn’t credible to me.

    Regarding the genetic distance maps of the Brandt, Haak et al. study, my thought was: The Bell Beaker culture originated on the Iberian peninsula, at least the ideological package and the basic equipment, that’s what archeology tells us. And likewise the mt-DNA evidence and the conclusions from the authors of the study suggest that there was a genetic influx from the Iberian peninsula associated with the Bell Beaker culture, discernible in the high frequency of haplogroup H and the introduction of novel variants of this haplogroup. And these western connections of the Bell Beaker culture would prevent modern eastern European and extra-European groups from displaying low genetic distance to the Bell Beaker people. Yet, as I said, the domestic pottery of the eastern Bell Beaker group shows more eastern relations. And the evidence from tooth morphology presented by Jocelyne Desideri here http://www.fondationlatsis.org/plpdf/Prix_Latsis/UNIGE_2008.pdf
    suggests that there was a strong local substrate in the eastern Bell Beaker group. The same is also evident from craniometry: Iberian Bell Beaker people were typically Atlanto-Mediterranean, while more central European groups were strikingly often Dinaroid. Also see this map here, showing the different local groups of the Bell Beaker culture: http://upload.wikimedia.org/wikipedia/commons/thumb/2/22/Beaker_culture_diffusion.svg/500px-Beaker_culture_diffusion.svg.png
    The Kromsdorf individuals belong to the yellow, easternmost groups. I would be wary of ascribing the spread of R1b to the earliest Bell Beaker groups from the Iberian peninsula. I’d much rather suspect that it was absorbed into the Bell Beaker culture somewhere in central Europe.

    Now, you’ve made a good point, alluding to the connection of R1a with the Caucasus component. This is also nicely reflected in the mt-DNA evidence from the Corded people, and in the recent study showing a particularly high variance of R1a near the Turkey/Iran border (Underhill et al. 2014). Also of interest here are the Rolloff experiments by Dienekes, which showed that the date of the admixture of the West Asian component into the Ukrainians was around 3530 BC, and into the Lithuanians around 3800 BC. So no, I don’t think the Caucasus association of R1a resulted from its spread from a glacial refugium. Instead, the Rolloff date is perfectly in line with the origin of the Kurgan burial custom in the Leylatepe culture south of the Caucasus, its implementation in the North Caucasian Maykop culture from 4000 BC on, and the spread of the North Pontic „Kurgan“/Yamnaya culture afterwards, in the late 4th millennium BC. An interesting conclusion that also follows, is that, in all likelihood, the Yamnaya culture was R1a, not R1b.

    The French are an R1b population, so it’s interesting to consider Dienekes’ Rolloff experiment regarding the Burusho-like admixture in the French: http://dienekes.blogspot.com/2012/09/rolloff-analysis-of-french-as-mixture.html
    The date is around 4940 BC, so much earlier than the Kurgan/Yamnaya times. Also earlier than Bell Beaker. Unfortunately this tells us nothing about the direction where this Burusho-like influence came from. An eastern origin would make sense, though.

    The decreasing frequency of R1b from west to east doesn’t necessarily prove its origin in the west. First of all, a group of R1b males coming from the east might have procreated more in western Europe than in the east. Secondly, later migrations from the east might have reduced the R1b frequency in central Europe.
    Also consider this study, http://dienekes.blogspot.com/2012/07/huge-study-on-y-chromosome-variation-in.html
    concluding that „the variance distribution of the rare R1b-M269* Y chromosomes, displaying decreasing values from Iran, Anatolia and the western Black Sea coastal region, is also suggestive of a westward diffusion from the Iranian plateau, although more complex scenarios can be still envisioned because of its non-star like structure“.

    Regarding the paucity of burials from the Upper Paleolithic (and the complete absence of y-DNA data from that era) I note that any UP European population that gave DNA to present day Europeans must have continued to exist through the Mesolithic. Hence, you cannot treat Mesolithic hunter-gatherers as if they were a distinct kind of human beings. They were at least partly, and probably mostly, descended from UP Europeans. And UP Europeans that didn’t survive through the Mesolithic are irrelevant with regards to modern y-DNA. Hence it’s the record of Mesolithic ancient DNA that matters. I acknowledge that it’s fairly limited and that it might be larger, and hopefully soon will be larger. Until that day I’ll remain sceptical with regards to R1b in Mesolithic Western Europeans.

    Alright, I acknowledge that drift and replacement of old paragroups was stronger in the small bands of the Paleolithic and Mesolithic. So there you have a mechanism that could explain how e.g. R1b-L23* disappeared from western Europe, if (if!) it once had been present there, since, according to the link you provided, R1b-L23 originated 8000 years ago.

    If you look at the distribution map of R1b-L23* in Myres et al. 2011 a very wide distribution is indeed apparent, in line with the relatively high age of that paragroup. The distribution looks a bit like a circle, encompassing Switzerland, Poland, the southern Urals, northern Kazakhstan, the Caucasus, Anatolia. More plausible than assuming an origin in western Europe would be an origin somewhere in this circle.

    And R1b-L11 originated 6100 years ago. So at least the latter (4100 BC) was definitely in a Neolithic context. And yet its paragroup is rare in western Europe. An explanation might be that R1b-L11* (and even older variants of R1b) weren’t very common in western Europe. Otherwise you’d have to invoke a severe bottleneck.

    The Centum-Satem distinction isn’t the one and only all-decisive trait. Germanic is linguistically intermediate between Italo-Celtic and Balto-Slavic, see here:
    http://dienekes.blogspot.com/2011/11/splits-or-waves-trees-or-webs.html

    This net was based on the data here:
    http://www.languagesandpeoples.com/Eng/SupplInfo/AnttilaNeighborNet.htm

    Thus, traits in common between Germanic and Balto-Slavic, under the exclusion of Italo-Celtic are:
    No. 10: -i suffix instead of -r (marking the present)
    No. 11: -m- suffix instead of -bh- (case marker)

    A trait in common between Germanic and Baltic, under the exclusion of Italo-Celtic is:
    No. 3: long o instead of a, oe.

    Further information can be found here: http://webspace.ship.edu/cgboer/indoeuropean.html
    For the genitive singular of nouns, we have Slavic (OCS) and Baltic (Lithuanian) using -ó, Baltic and Germanic (Gothic) using -eso, and Celtic (OIR) and Italic (Latin) using -í.
    For indirect and dative cases, we find a similar pattern: Slavic, Baltic, Germanic, and Tokharian use -m. Celtic and Italic using -bhos (dative singular).
    The Celtic-Italic link is fortified by such constructions as the comparison in -samo (vs -tero, -isto) and medium voice in -r (vs -oi, -moi).

    Also consider this tree by Ringe et al., based on lexical data:

    I think the Balto-Slavic like influence in Germanic was brought by R1a people, either with the Corded Ware culture, or with Unetice, or both.

    Personally, after writing this post, I’m doubting again that R1b was originally Indo-European. I mean, the R1a-branch had some relations with the Transcaucasian area. The earliest splinter group of Indo-European was located in Anatolia. And although the Caucasus component is virtually absent in Ireland, it has a descent presence in France and Spain, and a slight one in England and southern parts of Britain. And it’s precisely the Basques, with their 0% West Asian who don’t speak an Indo-European language. Maybe Indo-European originated in Eastern Anatolia, and R1b was originally Basque-related.

    As for your argument that subclades of R1b conforming with subgroups of Indo-European proves the Indo-European origin of R1b: That isn’t a convincing argument. R1b-M153 is a specific Basque and Gascon variant of R1b-S116. So ethnically specific subclades of haplogroups can arise even in ethnic groups which don’t speak the original language of the haplogroup, which applies in the case of the Basques and Gascons, if your theory is correct. (The Gascons are Romanized Aquitanians, who were Basque-speaking in pre-Roman times.) The explanation being: a specific subvariant of a haplogroup is more likely to spread within the borders of an ethnic groups than outside of them. Especially if the mutation arose within that ethnic group. And it doesn’t matter where the parent clade originally came from, or what language it spoke.

    I’ll have something more to say, later.

  46. Regarding the idea that there may have been a considerably more MA1/AG2-like population somewhere in Mesolithic Western Europe, there is also the problem that Mesolithic western Europeans were mostly descended from hunter-gatherers who had spent the LGM in the Franco-Cantabrian refugium. It’s dubious that groups who had expanded after the LGM from one and the same refugium could differ considerably in their autosomal make-up.

    The date for the admixture of the Burusho-like ancestry into the French inferred by Dienekes, 4940 BC, is hard to explain if that mixture is assumed to have taken place in France. At that date, the LBK settled down in France. It’s unlikely that it immediately mixed with the local hunter-gatherers, it didn’t do that in Germany either. And the local foragers were probably Loschbour-like. But the date was inferred with the French as a population, and this isn’t necessarily tied to the geographical area called France. Unlike Dienekes however, I don’t conclude that the mixture must have taken place in Western Asia. In fact, some archeologists have suggested that there has been an early influence of rather eastern European groups in middle Neolithic central Europe. In the Malice group of Poland (5000 – 4400 BC), which gave rise to Brzesc-Kujawski (4600 – 4100 BC) and Jordansmühl (4400 – 3900 BC); in the Tchitcharovce group, which gave rise to Csöshalom-Oborin in northern Hungary (from 4500 BC); in Gatersleben (Thuringia, from 4400 BC) and inTisza-Polgar (4400 – 4000 BC) and Bodrogkeresztur (4000 – 3700 BC) in Hungary. Related groups were also present in Croatia, Austria and Bavaria (Lasinja-Balaton 2, 4200 – 3700 BC). Now, as is often the case, this eastern influence is very controversial, since archeologists also find a lot of local continuity. But what I find noteworthy: Zsuzsanna Zoffmann (Anthropological sketch of the prehistoric population of the Carpathian Basin) could verify such an external, Cromagnoid influence in the Tisza-Polgar culture via Penrose-analysis of cranial material. In any case, the date of these influences would be well in line with the date inferred by Dienekes.

    Personally I surmise that R1b/Gedrosia entered central Europe at that time, and that it first expanded to western Europe with the central European Bell Beaker groups. The question remains, if that movement was associated with Indo-European languages or rather with Basque-related people. Now, compared with Indo-European speaking northwestern Europeans, the Basques have less Gedrosia, less MDLP Indo-Iranian, and considerably less MA1/AG2-like ancestry. This means inspite of their high incidence of R1b, they are less like the original R1b people than the Indo-European northwestern Europeans are. Moreover, the astounding expansion of R1b males speaks in favour of a certain cultural dominance associated with them. So it’s more likely that they belonged to a linguistic group that was succesful and expanding, like the Indo-Europeans.

    As for the question if the Proto-Indo-Europeans were associated with the West Asian or more with the North European component, I note that already Ötzi had 6% West Asian in the Globe13 analysis. This is only a bit less than the modern French have, and comparable to other western European populations. The recent study on the two Iron Age Bulgarians seems to show that the more Mediterranean, early Iron Age Bulgarian had a similarly moderate West Asian admixture. Moreover, in both Ötzi and this Bulgarian this West Asian admixture is combined with a strong Mediterranean component and a rather weak North European one. So, most of the West Asian admixture of most parts of Europe was present quite early, earlier than most of the North European admixture. So, unless we want to ascribe an Indo-European language to Ötzi, the Indo-Europeans must have arrived with the North European component, since this was the change that affected Europeans after Ötzi much more. Strong West Asian admixture is restricted to southeastern Europe and Italy. From there, some additional, slight West Asian admixture seems to have expanded to other Europeans at a very late date. The Romans were probably one of the most important agents for this.

    Finally I want to address your suggestion that the Neolithic words of Proto-Indoeuropean spread and mixed into a pre-existent Pre-Proto-Indoeuropean population which became Proto-Indoeuropean in consequence. These words must have been spoken by people, by a group of people which spread. The wheel-wagon vocabulary for instance is well represented in Germanic, Indo-Iranian and Tocharian, and partly present in the other languages too:

    It’s kind of unlikely that these words spread that far and wide by mere borrowing. And I’ve heard there is linguistic evidence that there wasn’t much borrowing between Indo-European languages after they had spread.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories
Follow

Get every new post delivered to your Inbox.

%d bloggers like this: