Tuesday, December 28, 2010

Valencia : Valencian Community (Spain)

From the Eneolithic (2000 BC) until the Roman conquest of Iberia, Valencia was populated by Iberians, more precisely by the Edetani who throughout this long period were influenced by Phoenician cultures and Greek settlers on the coast. It is quite probable that the modern extension of the Catalan language into the province emulates the former extension of the Edutani.

In Liria, Iberian inscriptions were found which have recently resurrected the old theories of basco-iberism : see this linguistic analysis by French-Galician linguist Hector Iglesias (L’inscription ibérique de San Miguel de Liria et le basco-ibérisme en général).

Catalan-speaking areas

  • Sample :

Full scale

  • Brief anthropological analysis :

- Type 1 : Dark complexion (dark hair, ... but there are light individuals), leptomorphic, long face (particularly on males), straight or convex nose rather parallel to the face, rather close-set eyes, high cheekbones, pointy chin, large jaw
~ (Dinaricized) Atlanto-Mediterranean

This is a very classical pan-Iberian type even though these individuals somehow show strong Catalan affinities when it comes to analyze secondary features such as a rather larger face than usual and more generally a mix of robustness and puffier features. Extreme robust types are illustrated by some males.

Lighter large-faced and dinaromorphic variants are found which are also rather Catalan-looking, nearly Pyrenean (high cheekbones are quite stunning).

Eventually, this Mediterranean-dominated sample also exhibits more classical Mediterranean individuals lacking peculiar ethnic traits but still very typical.

- Type 2 : Intermediate complexion (from dark to chestnut hair, from black to green/blue eyes, ...), more or less brachymorphic, reduced and "puffy" features, in some cases high-headed, rather puffy nose, strong jaw, close-set eyes
~ (Alpino-)Mediterranean

Such division is a bit artificial : those individuals also mostly fall within the Mediterranean spectrum as they don't strikingly exhibit traits associated with Alpinid types. Strong and robust brachymorphic Catalan types are absent from this sample which is very gracile on average. Nevertheless, a female series show more distant eyes, little noses, ... It doesn't really possess male counterparts.

  • Final morphotypes :


  1. Compare please with Catalonia or Murcia: these people are quite different (even if some individuals are clinal). In particular, I cannot spot a single one that looks "Basque-like" (a few could be somewhat ambiguous, very debatable, but that's all).

    I do suspect that they represent, largely, the original admixture of Cardium Pottery and other transmediterranean waves from Italy and the Balcans with a lesser apportion of aboriginal Iberian types. As you know Valencia and Alacant are the two provinces of Spain that have the most clear colonization episode in the Neolithic.

    Many individuals are very narrow/long-faced, often with marked East-Mediterranean features (A7, B4, B6, C2, C5), longfaced "Berid" variants (A8, C4, C7) or what I'd say are Balcanic types (D2, D5), Italid (D6) or just absolutely anomalous (D8, quasi-Mongoloid).

    Or am I seeing too much in too little? Your call.

    The most characteristic Valencian female type is surely described by A3 (nordo-med?) and D7 (with arguable Pyrenean affinities but very characteristic expression). Also A6. A more gracile (Dinarid?) type is represented by C8 and D2.

    The people who look less "ethnic" to my subjectivity are B3, B7, B8 and D4.

  2. You're right that this area was subject to more foreign influence than other Iberian areas. For instance, have a look at Dienekes' project.


    DOD189 is a full-blooded Valencian : he shows a stronger SW Asian admixture than the usual found in Spaniards (his Spanish sample). He also shows minor West African.


    Compare with DOD217, a Catalan : no signs of SW Asian (nor North African).


    Some comparisons :

    - Portuguese
    - French Basques
    - Me (DOD133)

  3. Not sure what to do with the Dodecad thingy. I make no sense of making world-wide runs when you are trying to describe people regionally. I mean Europeans are more properly compared with West Eurasians (including North Africans and West Siberians/Central Asians) but it makes no sense to be comparing all the time with Khoisan and Papuans for instance, when this is unavoidably bound to produce zero or near zero information.

    Also Dienekes has a fetish with the North/South European divide, which is not as real as it might appear in some studies. My reference is always Bauchet 2007, which IMO is the best research in European autosomal genetics to date, in spite of its limitations, and there you see that the N/S divide is an artifact that is eventually unmasked as minor NE and SE European components in Western populations, which have their own specific clusters.

    The resolution among Europeans of Dienekes' project is like k=2 in the Bauchet study, clearly insufficient and actually misleading.

    In Bauchet'07 you see that Valencians (tagged as "Spanish") have their own cluster, which is surely generic of Iberia (other stududies) but also maybe 20% or so of East Mediterranean component, which is not the same as West Asian component (because in Bauchet this is defined by Armenians, Greeks and Ashkenazim and there's no West Asian sample other than Armenians). We know that Armenians and Turks are very similar but other West Asians are not, specially not as you move to the South and East.

    Also a good deal of the Neolithic flow must have arrived not from West Asia nor even the Aegean but from the Adriatic area, where Cardium Pottery really began. So in order to know the true Neolithic element, Greece, Turkey or South Italy may be ok proxies but the real thing is in Albania, Montenegro, Dalmatia and Bosnia. Comparing with Palestinians, Arabs or Iranians instead is probably not going to tell much (and that is quite obvious in Behar 2010).

    Anyhow, in the Dodecad thing you appear very similar to Basques, maybe a bit more "Nordic" but that's it. And the Valencian appears also like Basques with minor West Asian admixture, while in Portuguese also some North African admixture appears (quite expected considering the presence of E1b1b1b in West Iberia).

  4. I don't really get why after getting that many samples from many areas in the Old World, he sticks to worldwide admixture runs : eventually he just detects two European components the distribution of which might be interesting but not that much when it comes to decypher the peopling of Europe. Still, he seems to be able to detect proper "Asian" imput (one peaking in the Caucasus, another one amongst Saudis) but that's a bit meaningless considering Neolithic migrants were not "pure" Caucasians or Saudis. I also presume his North African cluster is pretty convicing (as far as I know it peaks in one Galician guy which is consistent with what I've read so far on NW Iberia).

    Lately he has been refining his analysis and a Basque cluster now appears.

    Fine-scale admixture in Europe (Dagestan/Basque/Sardinian components)

    I've been in contact with "Poitevin" militants who contacted me asking me if I believed that organizing genetic sampling of Western France would be interesting (waiting for discounts) then communicating genomic data to amateur geneticians such as Dienekes or Polako. My firm opinion is that such studies should be investigated by official public organisms but since such studies are taboo in France ... I did not really know what to tell them. I fear that Dienekes would not be able to properly study such data, not that he's not knowledgeable (he is), but I fear that he doesn't really get "France".

  5. I am convinced that it is because he has a fetish on the North/South European dichotomy and does not want to see beyond (he's that way "cheating" himself and the rest).

    This is unfair for all three (or more?) West European distinct clusters and for the people who is providing their DNA in good faith. But on the other hand it's for free (right?) and he suggests you can do it yourself "easy" (though I could not understand the method yet - I miss all the technical issues such as convert this file to that and such).

    In that December post you mention, it does appear the distinct Basque cluster (which I'm sure you'd like to know how it flashes in your DNA, right?) but it still misses the (known to exist) Iberian and NW-Central European clusters, which are distinct from the Sardinian and NE European components with all likelihood. But that would maybe push a hole through Dienekes concepts of "Neolithic Europeanness".

  6. Just noticed that D. has just posted a K=46 structure analysis of West Eurasians and Pakistanis (as spreadsheet annex to this post). Notably French get their own distinct cluster, distinct from Basques, Iberians-North Italians, Sardinians and Central-NW Europeans.

    Other points of interest are several distinct Eastern European clusters (Lithuanian-Belorussian, Finnish, Russian-H, Daghestani, Adigey, Romanian), a rather unexpected Greek-Tuscan cluster, a distinct South Italian cluster (but including Ashkenazim), diverse clustering of Jews (distinct of what we have seen so far), distinction of Turkish and Armenians (against what Razib claimed recently), as well as of Georgians, identity of Kurds and Iranians, etc.

  7. I've seen that many attempts that I'm not sure that these results are that conclusive (for instance, he had a big problem in a previous run as he had included recently admixed people who had distorted the whole algorithmic process). Still, it's true that most French people fall in the first cluster alongside Brits, Hungarians (!!), Slovenians and Scandinavians. Six French individuals fall in the second cluster where one can find the Basques, Iberians, North Italians and curiously enough a consequent portion of Romanians. I was not included in this run (I'm the only French guy the genomic data he's got) : consequently his French are the ones used by most amateur geneticists (like Polako), a public sample in Lyons without data about origin. One can speculate that those 6 Frenchmen found in the 2nd cluster are from the South (Provence, Languedoc, ...).

    I must say that I'm highly dubious about the use of such sample not taking origin into consideration. Labelling as French 28 random individuals living in the second biggest French town is all but scientific. For instance, the very town of Lyons is situated on one major axis (the Rhône) and Provençal people strikingly differ people in modern "Arpitanian" lands (the border between "Oc" and "Oïl" around Valence and northern Ardèche being somewhat an ethnic border as far as I can tell with Ligurian placenames suddenly vanishing).


    When seeing such results, I'm seriously thinking of founding an association in France the aim of which would be to properly sample that country. I fear that it would be illegal though. If the price of such tests lower, I'll have a thought about it. CNRS in France won't move : they have to hide such studies with medical considerations.

    About other results : Armenians are found alongside Iranians and Kurds which is strange as I believed having read that Kurds differed from their East Anatolian neighbours.

    Structure in West Asian Indo-European groups

  8. Erratum : I've just realized I had not read the most interesting table with 4 dimensions. My mistake ! In the Basque cluster one can find 3 Spaniards, one French individual and more curiously one WASP from Utah.

    West Eurasians K=46 with 4 dimensions

    BTW best wishes for the year to come to all my readers !

  9. You are looking at the table k=15 and not at the wider one k=46 (link), which was my point of interest in fact.

    I took some time making a color map I might post at my blog. But I'm somewhat unsure of the validity of Dienekes' samples and methods.

    Thanks for the info on the French sample being in fact from Lyons. I did not know that. They still make their own cluster at k=46, with very few exceptions.

    Even at k=46 Central and NW Europeans cluster together (excepted some Scandinavians), what suggests this cluster must be very homogeneous (and likely Epipaleolithic by origin). But East Europeans become detached instead in several groups.

    "If the price of such tests lower"...

    It will in few years.

    I wish you best luck on that endeavor, which can only be of high informative value for European genetics in general. A possibile workaround constrictive legislation could be making the tests outside the French territory, in Geneva or London, for instance - though obviously this is a handicap, not so much with modern globalization and internet technology.

    "Armenians are found alongside Iranians and Kurds which is strange"...

    At k=15 all highland West Asians cluster together, from Lebanon and Turkey to Iran and the Caucasus. But at k=46, Lebanese only cluster with Syrians, Kurds with Iranians and all other populations are unique clusters (even Armenians and Turks appear distinct). With so many populations clustering only with themselves and several clusters empty or nearly so, the remaining trans-population clusters are very interesting:

    1. Central-NW Euro
    2. Iberian-N. Italian (maybe through Provence-Languedoc?)
    3. Tuscans-Greeks
    4. Lithuanian-Belorussian-Belgorod (not too surprising)
    5. South Italian-Ashkenazi
    6. Armenians-Assyrians-Iraqi Jews-Caucasus Jews
    7. Druze-Samaritan-Iranian Jews
    8. All populations from Balochistan
    9. Lesser component of Balochistan and Tripura Jews

    However they are not too coincident with other research, so it'd be nice to see the whole sequence, which may produce a more optimal clustering at lower k values (k=46 is a bit too much, right?)

  10. I just noticed your erratum, sorry.

    Happy new year, indeed. Urte berri on. :)


I've chosen to let people comment freely on my posts. Nevertheless, you'll lose your time taunting me and calling me a fascist (which I'm really not) : I pray you to read my introduction which will reassure that my intentions genuinely aim at achieving amateurish knowledge. I understand that you may not share my passion for the history of the peopling of the World, just don't let me know as clear conscience gained by bashing a humble documentary work is useless.