Substack cometh, and lo it is good. (Pricing)

Function & phylogeny, where the twain shall diverge


Every few years I get asked about Nuristanis and Kalash. The reason is that these people are often white. By white, I mean that some Nuristanis and Kalash are fair-skinned, blonde-haired, and blue-eyed. Entering “Nuristani” into Google images returns some very white faces. And you have weird news stories about ‘white’ Taliban, because non-locals don’t realize that some Nuristanis look like Northern Europeans.

Since the vast majority of people who look like white Northern Europeans are white Northern Europeans, many people assume that the Nuristanis and Kalash must have some kinship to white Northern Europeans. More precisely, many have spread the legend that these people have some relationship to the soldiers of Alexander the Great (even though Macedonians are Southern European…details).

As it turns out, they do have some kinship to Europeans…but not inordinately more than any of the other peoples of the region. The TreeMix plot at the top of this post shows that Greeks are far closer to Iranian Jews than they are to Kalash. In fact, the Kalash clearly have a non-trivial proportion of “Ancestral South Indian” South Asian ancestry.

Because of their high genetic drift (they’re endogamous and kind of inbred) a lot of population genetic analyses are a bit more difficult with the Kalash samples that are out there. But their genetic affinities are clear:

Table S4 show highly significant evidence (p value < 10−10) in the Kalash when using Armenia and Chamar as surrogates. Eight other pairings of surrogates give p values < 10−5. In all cases, the surrogate pairs include one group from South Asia (Chamar, Kol) and the other from West Eurasia (Armenia, Adygei, Brahui, Hungarians, Palestinians, Tuscans), consistent with admixture from a West Eurasian source.

Chamar are a Dalit caste of Northern India if you don’t know.

So what’s going on with the Kalash and Nuristanis? Appreciable frequencies of alleles which are correlated with traits like blue-eyes are found amongst them. Though the frequencies are much lower than in Northern Europeans. Very white looking Nuristanis and Kalash may be highly salient to photographers and the Western media, but it turns out most Nuristanis and Kalash look West Asian, with a minority who are dark-skinned enough that their South Asian ancestry is also quite clear.

This disjunction between appearance and ancestry should not surprise us. There has been a lot of recent change in physical appearance across populations over the last 10,000 years. Europeans themselves have changed in appearance. Similarly, other populations have as well. Some of them look similar to Europeans due to happenstance or convergence.

Another case are the Ainu of Japan. Though as an unadmixed group they no longer exist, old photos show some of them exhibiting an appearance not typical of East Asians. This led early anthropologists to posit that the Ainu were a “lost white race.” And yet to my knowledge, no European ancestry is found in Hokkaido, or in the Tohoku region, where Ainu-like people lived down to the early medieval period.

The moral of the story: don’t judge the contents of the book by its cover.

33 thoughts on “Function & phylogeny, where the twain shall diverge

  1. The Kalash have y-dna H1* at 20.5%. Even taking into account the bottleneck and endogamy , that is pretty good evidence that H was at one time much more common in North-Western South Asia.

    They speak a Dardic language, and considering their isolation and drift from neighboring populations, Dardic languages likely preceded Vedic Aryans to South Asia?

  2. Dardic is a branch of Indo-Aryan which is commonly interpreted as meaning it derives from Vedic Aryans.
    Nuristani however is an independent branch of Indo-Iranian.
    The border between these language groups is approximately the Afghanistan-Pakistan border.

    I wonder how much shared drift/population substructure there is between these groups (also Wakhi and Burusho)

  3. Interesting thoughts.

    With that being said, the Greek connection was never seriously entertained by a majority of scholars.

    Rather, from the very beginning of “Western” scrutiny, direct “Aryan” descent has been hypothesized (although, I do recognize that you don’t have in mind serious scholars; rather, you are referring to the countless journalists and armchair anthropologists who love the Greek/Macedonian/Albanian angle).

    So, considering that the current aDNA data seems to show that the Kalasha have around the same amount of Steppe_EMBA as most contemporary Northern European populations, the old notion of direct descent from ancient Indo-Aryan-related Central Asians seems to have panned out.

    I mean, at the end of the day, Greeks being closer to Iranian Jews makes good sense; both Greeks and Jews have low/moderate Steppe_EMBA.

    For a European comparison to the Kalasha, in terms of steppe-related genetic ancestry, Lithuanians probably make more sense.

    Anyway, I’ve met Kalasha IRL (I’ve been to Pakistan), and in my experience, they tend to have this sort of appearance:

    http://c8.alamy.com/comp/CY7ECE/kalash-men-with-characteristic-feather-in-their-caps-at-the-joshi-CY7ECE.jpg

    http://c8.alamy.com/comp/CXNJ62/elder-kalash-men-listening-to-stories-and-songs-of-the-history-of-CXNJ62.jpg

    http://c8.alamy.com/comp/CXNHK0/elder-kalash-man-wearing-a-traditional-cap-and-carrying-a-pakistan-CXNHK0.jpg

    https://i.pinimg.com/originals/2f/5d/f7/2f5df703b333fc50c03513c1d5a5f00b.jpg

    Totally subjective, but I feel that most of the Kalasha look much more “European” (in the North Eurasian-shifted steppe sense, not the classical Southern European, or the North American-centric West European/Central European sense) than most Kurds and Iranians, even the Kalasha individuals that aren’t extremely pale and light-eyed. But again, subjective.

    Regardless, the fact that many Kalasha look vaguely West Asian/Caucasus is interesting, considering that their West Asian ancestors probably looked rather similar to Indians.

    I mean, Iran_N, based on their pigmentation genetics, and based on quick reading I’ve done concerning their physical anthropology, didn’t much resemble modern West Asians in terms of phenotype.

    Rather, probably imagine something in between the Brahui people of southern Pakistan and the Toda people of southern India.

    Switching gears though; Razib, is there any chance you’ve heard some news on that big Central Asian/South Asian aDNA paper?

  4. There was a TED talk about the history of white explorers finding “lost white races” everywhere in the nineteenth century.

    https://m.youtube.com/watch?v=gn4bvjMh4vc

    Their overactive imaginations went to work whenever they met e.g. Africans who didn’t look like the other Africans. In reality, it’s just that all Africans don’t look the same.

  5. “The Kalash have y-dna H1* at 20.5%. Even taking into account the bottleneck and endogamy , that is pretty good evidence that H was at one time much more common in North-Western South Asia.”

    That’s interesting. What surprises me is H is absent in Paniya tribes, yet it is common in other southern tribes. Unlike R1a1 or L, y haplogroup H1 has a significant presence in all ethnic groups of India.Outside south Asia, H1 has a little presence in central Asia and Arabian Peninsula.
    Recently I discovered it peaks in Bangladeshis.
    I agree with you, H1 had a robust presence in NW before neolithic.
    The Kalash people might have significant pre-neolithic ANI DNA.

  6. FWIW both the Macedonians and Spartan individuals were described as blond hair – blue eye. I seem to recall an association with the problematic Doric Invasion. Not to mention a number of Gauls (Galatians) passing through at various points in time. So there is a basis for the idea of blond Greeks.

    The Bactrian Greek State was around for a little over 100 years. So the formal Greek presence lasts a lot longer than one invasion by Alexander. So the idea that they would be related to Southern Europeans isn’t an implausible idea. Doesn’t mean its correct.

  7. To me the interesting question is more why these groups (along with other nearby groups to a lesser extent, like Pamiri Tajiks and the Burusho) all tend to have a relatively high proportion of the population with lighter complexion.

    It’s of course possible in small, isolated population groups simple genetic drift will cause initially uncommon phenotypes to become more common. However, why does it crop up so frequently here and not elsewhere in the world?

    Common ancestry is also an unlikely reason. Certainly all of them (even the Burusho) have some sort of Indo-Aryan component to their population (which ancient DNA suggests included some very fair individuals) layered upon the older neolithic populations. However, the same is true of the lowlanders in surrounding areas.

    This leaves the most likely reason in my mind to be selection, which happened essentially in parallel across many valleys the Hindu Kush/Pamir region. That selection for lighter skin/hair/eyes happened during historic times in this region would not be surprising, given we already have indications that Europeans have become more fair themselves over the last few thousand years. Perhaps the generally colder and more seasonal climate in the mountain valleys meant people spent more time indoors and heavily clothed, which relaxed previously balancing selection for UV protection and allowed fairness to be selected (whether due to sexual attractiveness, or due to other effects of the fair genes)?

  8. Selection that merely allows for fairness due to low UV and keeping out of the cold, just results in slightly fairer skin, but still black hair and brown eyes, as in northern eurasian and far northern american hunting or pastoral people like the Inuit.

    To get the full “white” effect (ultra pale skin, fair hair, and light eyes) seems to require the fierce harsh selection that comes from both living in high latitudes, while also abandoning hunting and herding, and living instead on grains, leading to vicious rickets. Doubtless these people are farmers as well as herders and dairymen, but I would think they would be uniformly white if selection against rickets was going on.

    Is this maybe a case of regression to the mean, that in small populations, large variations can pop up, where in large populations, variation is much narrower? Then photographers come along and take pictures of just the white people.

  9. Is this maybe a case of regression to the mean, that in small populations, large variations can pop up, where in large populations, variation is much narrower? T

    no. none of this is right. in small pops you have more homogeneity in genes and looks.

    as alluded to above this could be drift. but it’s suspicious that the drift happens in the same way to highland mountain people. so as u and karl suggest it has to be something to do with selection.

  10. The Kalash prove that the cover matters, because your argument would make more sense if talking about individuals. Looking at groups, those with a significant proportion of European-like individuals are, as a rule, genetically closer to Europeans.

    On an individual base similarity in appearance can be misleading indeed. A Northern European looking Greek is still highly unlikely/never to be more Northern European shifted genetically and on the whole than an actual Northern European, regardless of its appearance.

    Considering Ainus: I never read they were considered Caucasoid proper, rather they were considered non-Mongoloid, what’s right. Even more interesting, Amerindians were quite often considered not fully Mongoloid because of their physical appearance as well. Looking at what we know about their genetic make up by now, this seems to be spot on. The ANE component is clearly different from what constitutes modern Mongolian and Chinese people.

    About the “cover”: The physical phenotype is more than a cover, its part of the content, part of the ancestry and adaptation of a people. A lot of traits considered random and believed to be caused by drift currently will show up as being functional and adaptive in future studies. At least if those future studies will be done properly and honestly by the researchers, considering more variables and scenarios.

  11. Considering Ainus: I never read they were considered Caucasoid proper, rather they were considered non-Mongoloid, what’s right.

    read more. coon thought they are related to europeans: https://books.google.com/books?id=HuRcAyXWJxIC&pg=PA274&dq=ainu+coon&hl=en&sa=X&ved=0ahUKEwjCr9716PHYAhWr1IMKHUj9DN0Q6AEILzAB#v=onepage&q&f=false (though there were arguments, and some people seem to have guessed correctly that they’re basal east asians).

    About the “cover”: The physical phenotype is more than a cover, its part of the content, part of the ancestry and adaptation of a people. A lot of traits considered random and believed to be caused by drift currently will show up as being functional and adaptive in future studies. At least if those future studies will be done properly and honestly by the researchers, considering more variables and scenarios.

    very few people think these traits are caused by drift who work in population genetics. the non-adaptive explanation is that it’s pleiotropy. do you work in this field to tell us this?

  12. The Bactrian Greek State was around for a little over 100 years. So the formal Greek presence lasts a lot longer than one invasion by Alexander. So the idea that they would be related to Southern Europeans isn’t an implausible idea.

    This doesn’t make sense to me, because the number of Greeks in Bactria was extremely small. It’s not impossible, but is very implausible.

  13. Considering Ainus: I never read they were considered Caucasoid proper, rather they were considered non-Mongoloid, what’s right.

    Hilariously, there are still white supremacists who think that the Japanese are part “Aryans” because of the Ainu intermixture. In their minds, this explains why the Japanese are so “European-like.”

  14. @Kev
    “I agree with you, H1 had a robust presence in NW before neolithic.”

    Very likely it was a major Y lineage during the Indus Civilization. It’s still found in low frequencies in Pakistan and NW India. It increases in frequency as you go South.

    Subclade of Y-Dna H1 was found in 60% of Balkan Roma. The Roma are descended from low caste groups in NW India. They left around 1000 c.e.

  15. You could have at least said that the Kalash have their different looks due to their large share of Indo-European ancestry.
    And that at least until the end of the Bronze Age, the whole of Central Asia was inhabitted by Yamnaya/Corded Ware/European peoples, which gradually admixed with other peoples in its margins (Siberians in the North, Chinese in the East, Indians in the South East and Iranians in the South West).

    This is important to point out, because, for instance, the Wikipedia pages on the Kalash simply don’t mention anything about the Indo-European presence.

  16. Razib,

    “nothing diff than ppl u’ve seen talking about it on the interwebs. paper is largely written. ppl are holding it up.”

    Incredibly unfortunate; we’ve all been waiting long enough.

    All,

    Looking at Razib’s TreeMix, I find it rather interesting that it’s the Iranians, and not the Kalasha, who receive an Indian-related migration edge.

    Of course, Iranians do seem to be only 1%-2% ASI, while the Kalasha are more around 10% ASI, but I guess the Indian-related edge is necessary in the context of this tree, since Iranians have a much higher Iran_N affinity in comparison to Greeks and Iranian Jews, and Iran_N affinity always translates to South Asian affinity when using contemporary references (which is why aDNA is always crucial).

    Also, with regard to Y-DNA H, I like to think of it as an imprint from Iran_N-related populations, and thus not indicative of ASI.

    I mean, we have Neolithic Anatolians and Early European Farmers with a distinct subclade of this haplogroup.

    Of course, that could be a wrong inference. One never knows.

  17. You could have at least said that the Kalash have their different looks due to their large share of Indo-European ancestry.

    And that at least until the end of the Bronze Age, the whole of Central Asia was inhabitted by Yamnaya/Corded Ware/European peoples

    it doesn’t look sintasha/andronovo in south asians (well, most of the times it doesn’t). it mostly looks more yamna, who did not look like the sintasha/andronovo (though all the requisite alleles seem to have been around in w eurasia, just at low frequencies). the yamna source in the kalasha probably explains hugely inflated ANE drift signals ppl detect.

    also, i wouldn’t say european. you need substantial EEF. sintasha had some. but from what i hear BMC did not. though we’ll see when published. if there is no EEF, i’m not going to call european.

  18. Not related to the main point but the following statement is questionable.

    > or in the Tohoku region, where Ainu-like people lived down to the early medieval period.

    Much later, if you narrow the scope to a subregion of Tohoku. There is ample documentary evidence of Ainu villages along the coastline of the northern tip of the Tohoku region during the early modern period. The Ainu, probably numbering much less than a thousand, lived in Hirosaki and Morioka Domains. They are today known as the Tsugaru Ainu and Shimokita Ainu, respectively. The historical connection of these people to earlier peoples of the region is unclear. They might be latecomers from southern Hokkaido.

    If you are speaking about the Tohoku region in general, it is wrong, in my opinion. There is a persistent assumption that the Ainu(-like) peoples are the indigenous inhabitants of a large part of Japan who have lived “down to” some period. However, a growing body of archaeological evidence from Tohoku and Hokkaido shows that the front line between farmers and hunter-gatherers greatly moved back and forth as the region went through the cycles of global warming and cooling. During the warm early-to-middle Yayoi period, wet rice fields reached northwestern Tohoku (cf. Sunazawa site) but they subsequently disappeared. I believe Ainu-speaking peoples expanded southward during cooling phases, leaving Ainu-like toponyms in the Tohoku region, and retreated to the north during warm periods. The peculiar distributions of -pet and -nay toponyms (both meaning river) suggests at least two waves of southward expansion of Ainu-speaking peoples. As for the Emishi in Japanese history books, those horse breeders must be Japanese-speaking people from central Japan, as I said in this blog before.

  19. no. none of this is right. in small pops you have more homogeneity in genes and looks.

    Thank you for the correction.

  20. also, i wouldn’t say european. you need substantial EEF. sintasha had some. but from what i hear BMC did not. though we’ll see when published. if there is no EEF, i’m not going to call european.

    Their geographical genesis was in Europe. Also, would WHG, EHG and SHG be Europeans by your analysis? The current Europeans didn’t even existed at the Yamanaya time – both the EEF and Yamanya were two very different flavours of Europeans, just like Sardinians and Russians today.
    Also, Steppe_MLBA clusters perfectly with today Europeans (I know that South Asians and the Kalash have a Steppe_EMBA-like ancestry, put this is not my point, I was talking about the popupations of the whole Central Asia before the Iron Age).

  21. Also, would WHG, EHG and SHG be Europeans by your analysis

    if you are talking more than geography, before corded-ware i wouldn’t say anyone is european (some overwhelmingly EEF pops like sardinians perhaps? but numerically marginal).

    I was talking about the population of the whole Central Asia before the Iron Age).

    i doubt sintasha/andornovo are the whole of central asia. we don’t know if the yamna-like s asia is from yamna, or is part of a group in central asia later that had less EEF. i’m pretty sure like europe (where WHG-like seems to have been somewhere) there were reservoirs of different stuff that are not sampled for whatever reason.

    finally, the admixture with east eurasian/siberian groups seems to have happened a few times. before the eastern scythians (who may or may not derive from the early admixtures). local extinction though seems pretty common in much of northern eurasian so there’s not a monotonic increase in that ancestry.

  22. also: some of the highest signals of ANE drift are on the southern fringe of c asia in mountainous zones (in north cauc & kalash). i don’t think this is coincidence, and can’t be explained sintasha-andronovo gene flow. either very old yamna, or, there were higher ANE-like groups in the steppe zone who moved south too. i think latter as likely as former.

  23. CHG/Iran_N peoples were very ANE-like themselves.
    The best ADMIXTURE graph I find for this is from the Supplement of “Extensive farming in Estonia started through a sex-biased migration from the Steppe”, as the ANE components are very visible.

  24. CHG/Iran_N peoples were very ANE-like themselves.

    Iran_N?

    haven’t seen follow-up on the north cauc ppl. mait’s group doesn’t think they are mostly in siberia btw. he does not like name “ANE”

  25. Regarding the Ainu, I presume their “exotic” phenotype for the region was mostly due to not having the derived “East Asian” version of EDAR. A 2015 study of Japanese genetics confirmed they were highly differentiated from mainland Japanese regarding EDAR (and another gene called COL7A1). Certainly the more profuse beard/body hair, and more “Caucasian” hair texture would be attributable to a lack of derived EDAR.

    I do wonder how much of the earlier supposition about their being “caucasoid” just came down to the beards however. I have seen many historic pictures of Ainu men (and a few contemporary ones) which look “white” – but I’ve never seen an Ainu woman who didn’t look broadly East Eurasian.

  26. the japanese papers on the ainu/jomon are all over the place. i think they’re pretty clearly basal east eurasians who can’t be well placed in modern topologies

  27. Twinkie – It is my understanding that at least 2 “Greeek” cities were founded in Bactria. Ai Khanoum, in today’s Northern Afghanistan, would be a prime example.

    Of course, we don’t know how many “Greeks” lived there for sure. Nor, how what percentage were “blond w blue eyes”.

    My recent reading on the “Enochian” changes in the pre-Christian Judaism, has reinforced what I think a lot of people have forgotten. The Successor States may have been light on administration, but were very much set on both colonization of peoples and culture.

    The Bactrian Kingdom would not have been a handful of colonial officials collecting up their salaries, waiting for an opportunity to go home. A whole of host of commercial interests would have followed them into the area. The Jewish population was spreading out, presumably interested in expanding commercial possibilities, and I presume a whole bunch of other folks (some possibly blond 🙂 would have done likewise.

  28. “(also, EDAR been segregating in eastern eurasia for a while; non-na dene amerindians have it, so at least 20 K BP)”

    But have we found EDAR in any ancient samples from the Americas? I assume Kennewick Man did not have it because he had a Sundadont dentition.

  29. @Name: at least until the end of the Bronze Age, the whole of Central Asia was inhabitted by Yamnaya/Corded Ware/European peoples

    Not 100% wrong, but my picture would add:

    1) we’ve got samples from a fairly thin (if very wide) Northern strip of Central Asia (really Southern Russia). most of Central Asia by land area is well south of the zone that we have any Indo-European samples in. See here for a very quick visualisation on an equal area map: https://i.imgur.com/ybnXzIu.jpg. What happened north or south of this zone is subject to much more uncertainty.

    2) If we talk about the Bronze Age-Iron Age, samples like the Okunev samples or the Andronovo outliers already show some ancestry from Siberians from before what we would consider to be Iron Age.

  30. I think part of this whole deal is just some phenotypes really sticking out to Northern Europeans in places where they aren’t expecting to see them.

    The Kalash seem to have lower frequencies of the known pigmentation derived alleles than any European population does (except Sardinians for some) yet some people are often amazed at some of their phenotypes compared so say Southern Europeans since in the European mind, all “Europeans” are expected to have them but South Central Asians aren’t.

    Without getting into details, I’m also a bit skeptical of the Kalash being that steppe-derived in actuality though they can be modelled appropriately with a good amount as Sein wrote.

Comments are closed.