Are Turks Armenians under the hood?

Benedict Anderson’s Imagined Communities: Reflections on the Origin and Spread of Nationalism is one of those books I haven’t read, but should. In contrast, I have read Azar Gat’s Nations, which is a book-length counterpoint to Imagined Communities. To take a stylized and extreme caricature, Imagined Communities posits nations to be recent social and historical constructions, while Nations sees them as primordial, and at least originally founded on on ties of kinships and blood.

The above doesn’t capture the subtlety of  Gat’s book, and I’m pretty sure it doesn’t capture that of Anderson’s either. But, those are the caricatures that people take away and project in public, especially Anderson’s (since Gat’s is not as famous).

When it comes to “imagined communities” I recently have been thinking how much that of modern Turks fits into the framework well. Though forms of pan-Turkic nationalism can be found as earlier as 9th-century Baghdad, the ideology truly emerges in force in the late 19th century, concomitantly with the development of a Turkish identity in Anatolia which is distinct from the Ottoman one.

The curious thing is that though Turkic and Turkish identity is fundamentally one of language and secondarily of religion (the vast majority of Turkic peoples are Muslim, and there are periods, such as the 17th century when the vast majority of Muslims lived in polities ruled by people of Turkic origin*), there are some attempts to engage in biologism. This despite the fact that the physical dissimilarity of Turks from Turkey and groups like the Kirghiz and Yakut is manifestly clear.

Several years ago this was made manifestly clear in the paper The Genetic Legacy of the Expansion of Turkic-Speaking Nomads across Eurasia. This paper clearly shows that Turkic peoples across Eurasia have been impacted by the local genetic substrate. In plainer language, the people of modern-day Turkey mostly resemble the people who lived in Turkey before the battle of Manzikert and the migration of Turkic nomads into the interior of the peninsula in the 11th century A.D. Of course, there is some genetic element which shows that there was a migration of an East Asian people into modern day Anatolia, but this component in the minority one.**

Sometimes the Turkish fascination with the biological comes out in strange ways, Turkish genealogy database fascinates, frightens Turks. Much of the discussion has to do with prejudice against Armenians and Jews. But the reality is that most Turks at some level do understand that they are descended from Greeks, Armenians, Georgians, etc.

To interrogate this further I decided to look at a data set of Greeks, Turks, Armenians, Georgians, and a few other groups, including Yakuts, who are the most northeastern of Turkic peoples. The SNP panel was >200,000, and I did some outlier pruning. Additionally, I didn’t have provenance on a lot of the Greeks, except some labeled as from Thessaly. I therefore just split those up with “1” being closest to the Thessaly sample and “3” the farthest.

First, let’s look at the PCA.

The Turks are shifted toward the Yakuts, but not too much. In contrast, there is much more of Yakut shift in Tajiks, and especially Turkmens. These are two groups from further east, closer to the heart of the zone Turkic expansion. Curiously, the Tajiks, who are the dominant non-Turkic Iranian speaking people of Central Asia, actually have more East Asian ancestry than the Turks of Turkey. This goes to show that ethnicity is somewhat fluid, and Turkic people have assimilated into the Tajik identity. That being said, please note that the Turkmen are notably more east-shifted than the Tajik.

Let’s see how this looks on pairwise Fst.

Fst is kind of difficult for fine distinctions when you have outgroups like Yakuts and Dai. So let’s look at Treemix with five migrations:

On this, you can see that the relationship of the Greece clusters on Treemix to Lithuanians matches PCA. Greece1 is the closest, Greece 3 the farthest.

The Turks are close to the Georgians and Armenians, but not the Kurds, or Tajiks. And, they receive gene flow from the Turkmen-Yakut region of the graph. So do the Tajiks…but the Tajiks also remove gene flow from the Lithuanians. The admixture plot makes it more clear what’s happening I think.

Yellow ~ modal in Southern Europe, green ~ modal Northern Europe, red ~ Central Asian, while blue and purple are northern and southern East Asian. In comparison to Turks of Anatolia Tajiks have a lot more Northern European affinity, probably because of the common steppe heritage. Not surprisingly, Turks have more Southern European like ancestry.

Curiously the East Asian ancestry in the Turkic people seems to be both Yakut and Dai like, so perhaps it was more cosmopolitan than we might think? The Yakuts after all are from the northern edge of the range, and may have absorbed a lot of indigenous Siberian ancestry.

Georgians have none of the Northern European sort of ancestry, but Armenians do, and Turks even more. One could posit that this is due to Slavic ancestry arriving with the Rumelian Turks who arrived in the 20th century, but just as likely is the possibility that Turks have a lot of ancestry from western Anatolia which was Greek, and Greeks have more of this than Armenians.

It’s hard to tell from these results whether Turks have more of an affinity with Greek or Armenians as their non-Turkic ancestors. So I ran a three population test.

Outgroup X1 X2 f3 error z
Turkey Armenians Yakut -0.00253688 6.70852e-05 -37.8158
Turkey Greece3 Yakut -0.00246931 6.72384e-05 -36.7247
Turkey Georgian Yakut -0.00256555 7.60158e-05 -33.7502
Turkey Armenians Dai -0.00246779 7.40038e-05 -33.3468
Turkey Greece3 Dai -0.0024101 7.34629e-05 -32.8071
Turkey Georgian Dai -0.00249174 8.11957e-05 -30.688
Turkey Greece2 Yakut -0.00222382 7.62368e-05 -29.1699
Turkey Greece2 Dai -0.00231001 8.39207e-05 -27.5261
Turkmen Turkey Dai -0.00288213 0.000108049 -26.6742
Turkmen Turkey Yakut -0.00254805 0.000102816 -24.7826
Turkey Greece1 Yakut -0.00225638 9.94722e-05 -22.6836
Turkey GreekCentral Dai -0.00235681 0.000104014 -22.6587
Turkey Greece3 Tajik -0.000622671 2.76666e-05 -22.5063
Turkey GreekCentral Yakut -0.00221985 0.000101654 -21.8373
Turkey Greece1 Dai -0.00243254 0.000112011 -21.717
Turkey Greece3 Turkmen -0.000640439 3.33529e-05 -19.2019
Turkey GreekThessaly Yakut -0.00208436 0.00011042 -18.8767
Turkey Dai GreekThessaly -0.00225435 0.00012241 -18.4163
Turkey Greece2 Turkmen -0.000584983 3.29819e-05 -17.7365
Turkey Armenians Turkmen -0.000520887 3.07253e-05 -16.953
Turkey Armenians Tajik -0.000421139 2.55274e-05 -16.4975
Tajik Turkey Dai -0.00140423 8.51697e-05 -16.4875
Tajik Turkey Yakut -0.00124601 7.60725e-05 -16.3793
Turkey Georgian Turkmen -0.000532496 3.80694e-05 -13.9875
Turkey Greece2 Tajik -0.000412419 3.04172e-05 -13.5587
Turkey Armenians Lithuanians -0.000459831 3.75838e-05 -12.2348
Turkey Greece1 Turkmen -0.000570715 4.7753e-05 -11.9514
Turkey Kurds Yakut -0.00146087 0.000124799 -11.7058
Turkey GreekThessaly Turkmen -0.000516877 4.46683e-05 -11.5714
Turkey Georgian Tajik -0.000328859 3.02443e-05 -10.8734
Turkey GreekCentral Turkmen -0.000504962 4.92555e-05 -10.2519

Armenians beat out Greece3 a bit better, but really it’s hard to say from this that this is definitive. It’s likely that my Turkish sample has both, and/or the original Turkic nomads had Iranian-like ancestry which was more like Armenian than Greek? Hard to say. Additionally, the face that Greece3 is better than the other options suggests to me that the source are Anatolian Greeks who were less impacted by migrations from the north than Greeks in Greece proper.

 

* The Mughals were Central Asian Turks, while the Safavids were mostly Azeri Turks.

** Since the Turks who arrived in Anatolia had long sojourned in Turn and Iran it is important not to assume that their contribution is limited only to the East Asian component of ancestry.

33 thoughts on “Are Turks Armenians under the hood?

  1. ‘Benedict Anderson’s Imagined Communities: Reflections on the Origin and Spread of Nationalism is one of those books I haven’t read, but should.’

    Political Tribes: Group Instinct and the Fate of Nations by Amy Chua is also excellent. I think you would like it. Her analysis of tribalism in America is fabulous. Besides it is also a great detoxifying agent after Pinker’s book.

  2. What’s interesting is that the regional analyses I’ve seen don’t show much of a pattern. With the exception of East Black Sea Turks, who look like unadmixed Georgians/Lazes/Armenians, apparently. Pontic Greeks aren’t so different either (somewhat East Med-shifted, but still very *Pontic*).

  3. Looks like there’s significant structure in the Turkmens too, some of them have way more “red” component than other Turkmens or Tajiks but no Yakut or “European”.

  4. Fascinating post. Thanks for the free research!

    Any paper suggestions on the genetic distance between syncretic, Christian, and Muslim communities in the Middle East? These populations also strike me as an interesting place to test the “imagined-ness” of community (if not nationality).

  5. Turks and Hungarians are extreme cases because these national identities were formed after there was an established statehood. So the foreign conquerors could base their rule on well established structures, sometimes even on a functioning administration.
    At the same time they were culturally different enough for keeping up their identity long enough, so that they slowly but surely imposed it on the local populace. Helpful was the cultural transition to a new religion. The new Hungarian identity was Catholic, the new Turkish Islamic. Contrary to what the majority of the conquered people was before. In Turkey the distinction between Turkish and Greek/Armenian was largely based on confession. There was no place for a Turkish Christian or a Greek Muslim. That made the transition much easier.

    For the formation of most nations existing today, thats rather exceptional. Obviously ethnic identities can be constructed and being “abiological”, but only as long as the “new nation” is still work in progress. As soon as its established, a biological identity, even if it is a new or mixed one, will be present again.

    But as for all human conditions and cultural products, they can degenerate and even work against their biological carriers, like in modern Europeans after the 1960s. Humans are just prone to cultural degeneration, thats the negative aspect of their behavioural flexibility and cultural creativity.

  6. The Turkmens are extremely heterogenous, even from century-old craniometrics, and the causes are well known. The more Westerly tribes raided Persian oases for brides, while Easterly tribes raided each other or Khiva oases.
    Yakut is a very imperfect proxy for the Turkic ancestors.
    Descendants of Balkan Muslims settled in today’s Turkey long before XX c., and a huge number of Muhajirs from the Caucasus brought layers of autochthonous North and South Caucasus people in XIX c.

  7. “There was no place for a Turkish Christian or a Greek Muslim.”

    There was one place: Pontus. Its Greek population was so heavy and the land so rugged, that its Muslim converts retained their language. On Greek independence and the various Ottoman depredations against their minorities, the Pontic Muslim Greek population stayed put. They are still there to this day.

  8. “One could posit that this is due to Slavic ancestry arriving with the Rumelian Turks who arrived in the 20th century, but just as likely is the possibility that Turks have a lot of ancestry from western Anatolia which was Greek, and Greeks have more of this than Armenians.”

    “Turkish administrators would scour their regions (but especially the Balkans) every five years for the strongest sons of the sultan’s Christian subjects. These boys (usually between the ages of 6 and 14) were then taken from their parents and given to Turkish families in the provinces to learn Turkish language and customs, and the rules of Islam. The recruits were indoctrinated into Islam, forced into circumcision and supervised 24 hours a day by eunuchs. They were subjected to severe discipline, being prohibited from growing a beard, taking up a skill other than soldiering, and marrying.” https://en.wikipedia.org/wiki/Janissaries

    A lot of them would have been Slavs. They may not have been permitted to marry, but a lot of them set up households and fathered children. The marriage stricture was removed in the 17th Century.

  9. The new Hungarian identity was Catholic, the new Turkish Islamic.

    huh? this makes no sense. hungarians were predominantly protestant before the catholic reformation slowly brought them back (well, tbh, habsburg sponsored efforts at conversion through coercion and persuasion). eastern hungary is still heavily protestant (under ottoman protection), and the current president, viktor orban, is from a reformed protestant background.

    In Turkey the distinction between Turkish and Greek/Armenian was largely based on confession. There was no place for a Turkish Christian or a Greek Muslim.

    it is true confession was the most important thing. the ottomans didn’t have an explicitly turkish ethnic cast insofar as balkan peoples in particularly were assimilated into the elites. but it’s just false to imply that there weren’t plenty of christians who spoke more turkish than greek and muslims who spoke more greek than turkish. during the great exchange of the 1920s lots of people migrated who spoke the ‘other language’ apparently.

    now it’s more well sorted out for obvious reasons (though as one commenter pointed out greek speaking muslims are still a thing in the pontic region, and people from that area who are turks are often extremely frank that they have little to no nomadic background).

  10. You seem to be ill-informed about the Byzantine Anatolia, which explains the funny title.

    1. Anatolia was mostly inhabited by Anatolian Greeks prior to the Turkish migration, those Greeks (genetically) were not identical to modern mainland Greeks. How do I know this? Well, we have Cappadocian and Pontic Greek samples. You can’t draw any sane conclusion using modern mainland Greeks as a proxy for Anatolian Greeks.

    2. Western and central Anatolia (where overwhelming majority of the Turkish population is concentrated) were inhabited by Anatolian Greeks (not by Armenians), which means modern Anatolian Turks can be modeled mainly as a mix of Anatolian Greeks and Oghuz Turks. This is also what historical sources tell us.

    “There is every reason to suppose that intermarriage took place rather extensively from the very beginning of the Turkish occupation of Anatolia and for several centuries thereafter. Anna Comnena speaks of the offspring of such unions as mixovarvaroi, and the twelfth-century Balsamon refers to their curious practises. When the Greek historian Nicephorus Gregoras passed through Bithynia en route to Nicaea in the middle of the fourteenth century, just one generation after the conquest of Nicaea, he observed that the population consisted of Greeks, mixovarvaroi (Graeco-Turks), and Turks. Thus intermarriage of Muslim and Christians at every level of society played a very important role in the integration and absorption of the Greek Christian element into Muslim society”

    The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh through the Fifteenth Century – Speros Jr. Vryonis

    3. Using Yakuts and Dai as a proxy for the Seljuk era Oghuz Turks is ridiculous, they can’t even be used as a proxy for early Turks, let alone the Seljuk era Oghuz Turks. The Central Asian admixture in Anatolian Turks is far from being Yakut-like. The link below contains the comparison of Anatolian Greeks, Anatolian Turks and Central Asians.

    https://oghuzturksdna.blogspot.com.tr/2018/02/the-genetic-structure-of-anatolia-from_21.html

    4. Why are you obsessed with explaining the Turkish ethnogenesis with “acculturated Armenians” despite the fact that western and central Anatolia were mostly populated by Anatolian Greeks prior to the Turkish migration. Is there any special reason? Anatolian Turks can be modeled as a mix of Anatolian Greeks and Central Asians (not Northeast Asians/Yakuts or Dai). The Central Asian admixture in Anatolian Turks seems to be Transoxianian/Khwarezmian-like.

    5. The genetic profile of Anatolian Armenians (who are different than eastern Armenians) resemble that of Assyrians.

    We are in 2018, not in 2010, any sane person with knowledge of history (and genetics) would laught at the title. Your article contradicts with history. The fact that you insist limiting the Central Asian ancestry to Northeast (Yakut) and East Asian (Dai) discredits the whole article.

    Least-squares method

    Using 1 population approximation:
    1 Turk_Anatolia @ 0
    2 Azerbaijani_Turkey @ 8,945804
    3 Azerbaijani @ 9,580063
    4 Azerbaijani_Iran @ 10,011403
    5 Azerbaijani_Karabakh @ 11,317268
    6 Kurd_Kurmanji_Turkey @ 11,763043
    7 Zaza @ 13,04868
    8 Turkmen_Iraq @ 13,565882
    9 Greek_Central_Anatolia @ 13,610301
    10 Greek_Cappadocia @ 14,479651
    11 Circassian_Kabardian @ 14,725936
    12 Balkar @ 15,59073
    13 Armenian_West @ 16,213126
    14 Circassian @ 16,725841
    15 Kurd_Sorani_Iraq @ 16,752008
    16 Kumyk @ 16,797623
    17 Lebanese @ 17,571103
    18 Iranian @ 17,894575
    19 Turkmen_Ashgabat @ 18,086303
    20 Ossetian @ 19,68327
    220 iterations.

  11. Elçi,

    you make some good points. but one difference from what i can tell btwn u & i is that i don’t just use other peoples’ calculators. so let’s not lecture about genetics….

  12. @Walter Sobchak

    “They were subjected to severe discipline, being prohibited from growing a beard, taking up a skill other than soldiering, and marrying. The marriage stricture was removed in the 17th Century.”

    Yeah, that’s also when the devshirme practice was abolished. Janissaries did not contribute to the genepool. Even if they were allowed to marry, their contribution would have been negligible.

    “In Turkey the distinction between Turkish and Greek/Armenian was largely based on confession. There was no place for a Turkish Christian or a Greek Muslim.”

    False. There are Muslim Armenians in Turkey even today, they are known as Hemshins. The presence of Muslim Slavs (Pomaks), Greeks (Romeika-speakers) and Armenians (Hemshins) destroys this myth, unfortunately for you.

    “One could posit that this is due to Slavic ancestry arriving with the Rumelian Turks who arrived in the 20th century.”

    Any Turk with Rumelian ancestry is well aware of his/her Rumelian ancestry, it has been only a few generations since their Rumelian ancestors migrated to Anatolia. The presence of North European-like admixture in Anatolian Turks has nothing to do with Janissaries or Rumelian Turks. Keep in mind that even the Anatolia_Bronze_Age sample has some NE-like admixture.

    Central Asians have more NE-like admixture than West Asians, which can explain why Turks have more NE-like admixture than Anatolian Greeks and Armenians.

  13. @Razib

    Elçi is correct when he says that non-Central Asian ancestors of modern Anatolian Turks are mostly Anatolian Greeks rather than Armenians and that modern Anatolian Greeks, who should be the best modern-day proxy for Byzantine Anatolia, are genetically not similar to modern Balkan Greeks (a.k.a. mainland Greeks). Also, modern Balkan Turks (a.k.a. Rumelian Turks) are genetically not similar to modern Anatolian Turks. I know all these facts quite well because I manage a Family Tree DNA project on the genetics of Anatolia, the Balkans and the Caucasus with hundreds of members. Here is our project link:

    https://www.familytreedna.com/groups/anatol-balkan-caucas/about

    And not just because of that. I am also an active member and moderator in some genetic discussion boards such as Anthrogenica where people of various origins share their genetic results. In particular, I am a moderator of the Turkic section of Anthrogenica, among other sections.

    Even more to the point, as part of our genetic project we have closely analyzed the genetics of Anatolian and Balkan Turks and Anatolian and Balkan Greeks, among other populations, not just by using their GEDmatch calculator results and PCA and dendrograms based on them, but also using PCA we have done ourselves directly based on autosomal raw data (using unrelated samples). Here are the PC1-PC2 and PC1-PC3 graphs of our West Eurasia PCA respectively:

    https://drive.google.com/file/d/0B9n4j3PQ81RcU0FkYXhLSl9UeVU/view?usp=sharing

    https://drive.google.com/file/d/10Dathyg9qaV0FVKKwdeRB2PYiZ01LADz/view?usp=sharing

    As you can see from the PCA, Anatolian Greeks (designated as Anatolian Hellenic here) and Anatolian Turks are genetically much closer to each other than either one of those populations to Balkan Greeks (designated as Balkan Hellenic here) or Balkan Turks. Also, Anatolian Turks are positioned closer to Anatolian Greeks than to Armenians. Indeed, Anatolian Greeks seem to be among the genetically closest populations to Anatolian Turks overall. Balkan Turks are positioned closer to Balkan Slavic populations such as Bulgarians than to Balkan Greeks, which makes perfect sense because all our Balkan Turkish samples are from regions mainly inhabited by Balkan Slavs, and particularly Bulgarians, rather than Balkan Greeks. PC3 captures the East Eurasian-related ancestry of Anatolian and Balkan Turks, so on PC3 those two populations are shifted towards east compared to their non-Turkic neighbors.

    Razib, since you are well-versed in doing genetic analyses yourself, if you want I can share with you the autosomal raw data and GEDmatch kit numbers of the Anatolian Greek and Balkan Turkish samples I use in my analyses and provide detailed ancestry info on them. You can contact me by email, you already know my email address. Also feel free to ask me questions.

    As for Elçi’s analyses, I should state that there does not seem to be a good modern proxy for the Turkic population that invaded Anatolia and brought the Central Asian input into the modern Turkish gene pool due to various factors too numerous to enumerate here. My best guess for a modern proxy for them would be Uyghurs, but I would not bet my money on them.

  14. @razib: I was writing about the formation of the Hungarian kingdom after conquest. Without the Catholic kingdom and identity the Hungarians as a people wouldnt have survived. In later times Catholicism was indeed nothing constitutive any more. The tribal people of the Hungarian plane were united under the Catholic banner after the catastrophic defeat of the pagan way.
    Otherwise they would have shared the fate of the Avars.

  15. You wrote a similar article back in 2008 and looks like you haven’t made any progress ever since. Let’s put aside your lack of basic knowledge about mediveal history of Anatolia, looks like you are not familiar to the genetic make up of any of the populations you mentioned in the article either, as both Elçi and Onur pointed out. Hence the conclusions you have drawn are extremely false and misleading, not to mention the Turkish samples you used in this article seem to be highly dubious and are not exactly representative for all Turks. So, I don’t really understand the purpose of this article since it doesn’t offer anything new and seems to be repeating the same old factual mistakes.

    Being able to use your own “tools” for such estimations when it comes to “lecture” about genetics doesn’t give you any advantage against others if you are lack of basic historical/demographical and genetic knowledge about the subject populations.

  16. @Onur

    How many Anatolian Greek samples do you have?

    I agree with the part that there is no perfect modern proxy for the Turkic population that migrated to Anatolia, but among the samples available, Uyghurs are even a worse proxy than Uzbeks for they have elevated Dai-like admixture. There was no significant decrease in CHG-related components (in contrast to Neolithic-related components) after the Turkic migration, which alone allows us to eliminate the Kyrgyz, Kazakhs and Karakalpaks.

    @Razib
    “Of course, there is some genetic element which shows that there was a migration of an East Asian people into modern day Anatolia, but this component in the minority one.”

    In the 700s, the Oghuz Turks migrated to the area between the Caspian and Aral seas from the Jeti-su area. The first wave of the Turkic migration into Anatolia was from the western portion of Central Asia, the following waves (especially the ones triggered by the Mongol invasion) were from Khwarezm and Khorasan.

    So, “a migration of an East Asian people into Anatolia” was not the case, the Oghuz Turks were Central Asians.

  17. @Elçi

    How many Anatolian Greek samples do you have?

    I have the autosomal raw data of three unrelated Anatolian Greek individuals, who are all fully of central Anatolian Greek origin, so they should be fairly representative of the average person in the Byzantine Anatolia. In addition, I have the GEDmatch kit numbers of many more central Anatolian Greeks and Greeks from elsewhere in Anatolia.

    I agree with the part that there is no perfect modern proxy for the Turkic population that migrated to Anatolia, but among the samples available, Uyghurs are even a worse proxy than Uzbeks for they have elevated Dai-like admixture.

    I think Uyghurs can be a good proxy for the Turkic immigrants to Anatolia in terms of their West Eurasian/East Eurasian ancestry ratio. But I agree with you that they are probably not a so good proxy in terms of the ratio of East Eurasian sub-components, the Turkic immigrants to Anatolia probably had a higher ratio of Northeast Eurasian/Southeast Eurasian.

    There was no significant decrease in CHG-related components (in contrast to Neolithic-related components) after the Turkic migration, which alone allows us to eliminate the Kyrgyz, Kazakhs and Karakalpaks.

    I did not say anything about Kazakhs, Kyrgyz or Karakalpaks. But among the three, Karakalpaks are probably the best proxy for the Turkic immigrants to Anatolia.

    In the 700s, the Oghuz Turks migrated to the area between the Caspian and Aral seas from the Jeti-su area. The first wave of the Turkic migration into Anatolia was from the western portion of Central Asia, the following waves (especially the ones triggered by the Mongol invasion) were from Khwarezm and Khorasan.

    So, “a migration of an East Asian people into Anatolia” was not the case, the Oghuz Turks were Central Asians.

    I think Razib already clarified what he meant by adding this footnote:

    “Since the Turks who arrived in Anatolia had long sojourned in Turan and Iran it is important not to assume that their contribution is limited only to the East Asian component of ancestry.”

  18. the sons of the wolves have their teeth out.

    i’ve been indulgement, but please chill with the patronization. honestly you guys know a little less about genetics than you think you do.

    1) east asians. i’ve looked at a lot of east asian samples. yakut-to-dai is pretty much the unadmixed cline. the east asian source of the turks has to be along with cline, unless there is an east asian population that is totally extinct and different. treemix should pick a migration arrow btwn them if it’s between them. it didn’t. it came straight from the yakut region (it can also take two edges if there was a large ‘chinese-like’ component, as the uygurs seem to have). also, the variation among east asian people north of mainland southeast asia is pretty low genetically once you remove admixture (central siberian, and west eurasian).

    2) the idea that the oghuz turks were ‘central asian’ as opposed to ‘east asian.’ this is why treemix is good. unless the ancestry of the oghuz turks is totally gone, something like treemix should be able to recapture the migration arrow. and it was central asian it’s weird how the turk samples are enriched for east asian vis-a-vis armenians and but not the ‘central asian’ like component (unless they mixed almost exclusively with greeks with no central asian to start with).

    3) i have a large # of “turk” samples. unfortunately i don’t have provenance for most. so i did some outlier analysis and removed ones that looked totally greek or armenian, and focused on those with clear east asian admixture. so sampling *coverage* shouldn’t be a huge issue. but sampling *labeling* is (and, perhaps lots of turks are pure greek and armenian, but the published literature does not show that).

    4) some of the same issues apply with the greeks. i don’t know who descends from people who migrated out of turkey. but with my N they are probably in there.

    i know plenty about the history of anatolia. my data is just limited. i’ll await the better samples from onur.

    i’ve think i’ve let enough insulting comments through so that’ll be that.

  19. the sons of the wolves have their teeth out.

    2) the idea that the oghuz turks were ‘central asian’ as opposed to ‘east asian.’ this is why treemix is good. unless the ancestry of the oghuz turks is totally gone, something like treemix should be able to recapture the migration arrow. and it was central asian it’s weird how the turk samples are enriched for east asian vis-a-vis armenians and but not the ‘central asian’ like component (unless they mixed almost exclusively with greeks with no central asian to start with).

    Any analysis without Anatolian Greeks is misleading. Anatolian Greeks have less CHG/Iran_Neolithic-related admixture (represented by the Central Asian component in your article) than Armenians.

    No need to be angry. I haven’t seen any insult in the comment section to be honest. You can’t expect people not to answer when you openly target an identity with baseless analysis and sensational titles.

  20. i have anatolian greeks. they aren’t labeled. i have 86 people whose 4 grandparents are greek. some of these have to be people whose families left turkey.

    if onur can get me the samples i can label them with the individuals with known provenance and we can finish this.

  21. Question: Many ethnicities resemble their neighbours (even if their neighbours belong to another language family). The difference between Turks and Armenians is bigger than the difference between various neighboring populations. Would you call another etnicity “X under the hood” as well? Or is this kind of approachment only limited to subjects corcerning Turks?

  22. Would you call another ethnicity “X under the hood” as well?

    yes.

    though in this case, the comparison is salient because of the historical conflicts/animosity.

    it seems likely that ancient jews are most like groups such as assyrian christians. modern day palestinians are surprisingly diverse and variegated.

    but you gave me a good idea. i should do a similar analysis of fertile crescent groups….

  23. “Yeah, that’s also when the devshirme practice was abolished. Janissaries did not contribute to the genepool. Even if they were allowed to marry, their contribution would have been negligible.”

    From what I’ve read, by the time the devshirme was abolished the marriage restrictions had long ceased to be strictly enforced and there was less need for Christian-born recruits anyway.

    In any case, during the empire’s classical age many Janissaries and/or Christian recruits were promoted to occupations where they could marry. During the period of Suleiman the Magnificent much of the Ottoman ruling class was Christian-born, including a majority of the Grand Viziers. Additionally, the women of the Sultan’s harem and those of the wealthiest upper class were usually non-Turkish concubines. Harem women who did not become mothers or companions of the Sultans (very few did) were often married off to high officials, as were the Sultans’ daughters.

    Putting aside military recruits, the genepool was also undoubtedly affected by the vast slave trade, not only of captured Western soldiers but also of millions of men and women captured by the Tatars in the Caucasus and Ukraine and sold in the empire. Islamic practices encouraged freeing slaves and many undoubtedly put down roots and started families.

  24. Putting aside military recruits, the genepool was also undoubtedly affected by the vast slave trade, not only of captured Western soldiers but also of millions of men and women captured by the Tatars in the Caucasus and Ukraine and sold in the empire. Islamic practices encouraged freeing slaves and many undoubtedly put down roots and started families.

    northern european ancestry is quite distinct. as are slavic R1a Y lineages. it’s actually less impact than i would have guessed judging by the mentions of slavery in the literature.

    this is a common phenomenon. slavery tends to have less genetic-demographic impact over the long term than we think. y?

    two reasons (related)

    1) slaves had lower total fertility rate in their generation. american slavery for example was *relatively* genteel, so the slave population was replacing itself or even growing. but carrib and brazilian slavery only persisted because of the slave trade. mortality rates very high. some ancient groups, such as anti-natalist areas of greece or sicilian latfundia, might have been like this

    2) most freedman and their descendants were in the lower social orders, though those who were high status or became so are salient (eg all of claudius’ numerous freedman). in pre-modern periods lower status people had lower fertility and contributed less and less to the subsequent generations. once the slave trade pipeline stopped for various reasons their ancestry would slowly be selected against through this process that impacts lower classes more generally

  25. From having seen Anatolian Turkish and Greek results, I have to agree with Onur and Elci.

    Part of the reason why Anatolian Turks have more “Northern European” seems to actually be due to the Central Asian (mix of steppe and East Asian) part of their ancestry that Anatolian Greeks lack.

    One con is that we have no *direct* pre-Turkic Western Anatolian representatives, like we do with Eastern Anatolia with Pontic and Cappadocian Greeks. The ones that left Western Anatolia in the 20th century were largely descendants of late Ottoman immigrants from the islands and the Balkan mainland, not pre-Turkic remnants, with maybe some potential exceptions in northwest Turkey. I actually haven’t seen too many of those but the few I have seen were quite “un-Anatolian”. Maybe Onur can weigh in on this specific point and correct me. I imagine the pre-Turkic ones would have been on a cline between Aegean islanders/Cypriots and Eastern Anatolian ones that ends at Armenia and moves on to the Caucasus (Georgians) to the north and Mesopotamia (Assyrians) to the south.

    It’s a shame you didn’t have a sample with just them.

  26. @Razib

    Please check your emails, I have now sent you the autosomal raw data of three unrelated fully central Anatolian Greeks and two unrelated fully Balkan Turks. There are also many relevant GEDmatch IDs in my email. I took consent for sharing data.

  27. In any case, during the empire’s classical age many Janissaries and/or Christian recruits were promoted to occupations where they could marry. During the period of Suleiman the Magnificent much of the Ottoman ruling class was Christian-born, including a majority of the Grand Viziers. Additionally, the women of the Sultan’s harem and those of the wealthiest upper class were usually non-Turkish concubines. Harem women who did not become mothers or companions of the Sultans (very few did) were often married off to high officials, as were the Sultans’ daughters.

    Let’s get some facts straight:

    – Janissaries were not allowed to marry.
    – The devshirme practice was abolished in the 17th century.
    – Contrary to popular belief, Janissaries and the kapıkulu (servants of the porte) in general were too few in number during the times of the devshirme practice. They were merely household troops, not the largest segment of the army. According to David Nicolle, the number of Janissaries in the fourteenth century was 1.000, and estimated to be 6.000 in 1475.

    Putting aside military recruits, the genepool was also undoubtedly affected by the vast slave trade, not only of captured Western soldiers but also of millions of men and women captured by the Tatars in the Caucasus and Ukraine and sold in the empire. Islamic practices encouraged freeing slaves and many undoubtedly put down roots and started families.

    Your fantasies have no limits. Most Turks were either farmers or nomads, they did not play any role in the slave trade. Let alone the slave trade, the vast Ottoman trade network was nearly monopolized by non-Turks (Jews, Greeks, Armenians). Turks were one of the least wealthy ethnicities in the empire.

  28. @RandomName

    One con is that we have no *direct* pre-Turkic Western Anatolian representatives, like we do with Eastern Anatolia with Pontic and
    Cappadocian Greeks. The ones that left Western Anatolia in the 20th century were largely descendants of late Ottoman immigrants from the islands and the Balkan mainland, not pre-Turkic remnants, with maybe some potential exceptions in northwest Turkey. I actually haven’t seen too many of those but the few I have seen were quite “un-Anatolian”. Maybe Onur can weigh in on this specific point and correct me. I imagine the pre-Turkic ones would have been on a cline between Aegean islanders/Cypriots and Eastern Anatolian ones that ends at Armenia and moves on to the Caucasus (Georgians) to the north and Mesopotamia (Assyrians) to the south.

    Parts of coastal western Anatolia received significant Greek migration from lands further west during the Ottoman times, especially beginning from the 17th century. But most of those immigrants came from the nearby Aegean islands and so were not genetically so different from native Greeks of coastal western Anatolia. Greeks of the Aegean islands that are geographically close to Anatolia are genetically quite western Anatolian-like and distant from mainland Greeks. Everyone can check it by looking at the Island Greek average results at the GEDmatch calculator MDLP K23b.

  29. i’m closing comments on this thread.

    1) i have more familiarity with proxy populations than some of you commenters seem to assume (sorry, it’s just true that e. asian diversity is pretty low, so unless the oghuz had a totally different parent set of pops….).

    2) cutting & pasting calculator results are informative, but really it’s not pushing the needle beyond earlier comments.

    3) i do understand i trolled with the title a little, so i’m not surprised ;-0) but i think the next post with onur’s data will be informative….

Comments are closed.