Friday, June 12, 2015

Beakers from Bavaria and Bohemia (Allentoft et al, 2015)

The metadata in this study has been really difficult to navigate, so this is kind of an incomplete post and hopefully I'll be able to re-attack once some of the data (that we care about) crystallizes.

Kněževes - grave 8 (modified from Turek, 2012)
To start, one of the more interesting things about the Beakers in this study is that Kněževes - grave 8 was actually a woman.  It's difficult to say that it is her husband's mementos because she herself is oriented like a man.  I'm hoping I have the right graphic, but you can see below the sexing was different from the genetic result.  I'm guessing there was a question here and that's why this individual and the infants were chosen by the Czechs. (Don't quote me on this)

This doesn't necessarily mean she had a man identity or complex.  Certain items were associated with power in ancient societies, such as the queen who is depicted with the false beard, crook, flail and nemes (all things associated with male pharaonic power).  Clearly, certain regalia was buried with Beaker men who were importantly viewed in their society.  Jan Turek has written on this subject.

Bell Beaker individuals in this study lack carbon dates, this being because their pottery were historically satisfactory for placing them in a chronology.  But, the cemeteries at Augsburg, Osterhofen and Landau in Germany and Knezeves and Brandysek in the Czech Republic were subject to one of the earliest and largest isotope studies.  Price et al, 1997 (Table 7) found that Augsburg Beakers were around 15% first generation immigrants, Osterhofen around 38% and Landau around 44%. [also 2004]

These same cemeteries were also lumped into a dental study by Desideri and Besse (2010) which found population movement between groups.  Here's some of the sampled individuals from Table 1 of Allentoft et al.

RISE559 Augsburg Bell Beaker F0174, grave 4 tooth F F Adult
RISE560 Augsburg Bell Beaker F0187, grave 3 tooth M M Adult
RISE562 Landau an der Isar Bell Beaker F0228, obj. 136/92 = grave 9 tooth FF Adult
RISE563 Osterhofen-Altenmarkt Bell Beaker F0234, obj. 8 tooth MM Adult
RISE564 Osterhofen-Altenmarkt Bell Beaker F0241, obj. 25 tooth MM Adult
RISE566 Knezeves Bell Beaker F0521, A01168, gr. 14 tooth MM Infant
RISE567 Knezeves Bell Beaker F0523, A0766, gr. 8 tooth MF Adult
RISE568 Brandysek Bell Beaker F0525, A01623, gr. 16 tooth nd F Infant
RISE569 Brandysek Bell Beaker F0527, A01643, gr. 35? tooth nd F Infant

The last individual from Osterhofen F0234 is the male that is R1b U152 (not DF27) according to Richard Rocca   [See also Rocca et al, 2012].

If true, then that means that two of the Central Beakers where subclades can be discerned are U152/S28, not entirely surprising.  It's presence among Bashkirs is noteworthy however, especially given the stacks of ancient R1b coming out of the North Caspian.

*Ok, Correct map now, got a little careless.  Thanks to Krefter.

Bronze Age population dynamics, selection, and the formation of Eurasian genetic structure, Allencroft et al.  Nature 522, 167–172 (11 June 2015) doi:10.1038/nature14507 [Link]
Additional data [Link]



  1. Do you know the dates for these BB samples? I looked in the Supplementary information of the Allentoft paper, but couldn't see dates for the BB samples.

    1. There are no carbon dates for these Beakers. BB was historically put into chronology like Egyptian pottery, based on typology alone due to the uniqueness of various schemes. It's a shortcoming.

    2. But these are generally thought to be in the ballpark of 2600 to 2200 B.C.

    3. OK. Thanks, much appreciated.

  2. @bellbeakerblogger,

    Not a big problem. But you posted a map of R1b-U106/S21. The Beaker man you speak of had R1b-U152/S28. Eupedia has a map of that too.

    Also about the argument of where western R1b-L11 comes from. Evidence is stacking up for a East European origin, not Western anything.

    A Corded Ware man from Germany had R1b1a2a-L23(xZ2103, L11). He lived before Bell Beaker arrived in Germany, and when the very first Bell Beakers are found in Iberia. Considering the close autosomal relation between Corded Ware and Yamnaya(who was mostly R1b-L23) it's no surprise some R1b-L23 was carried by Corded Ware into Germany. Today there's R1b-L23* and R1b-Z2103 in Central Europe, so one could argue that's the only steppe-derived R1b in Europe.

    Keep in mind though. Abstracts have confirmed people in El Portalon Spain even just under 4,000 years ago were still Sardinian-like. Y DNA from Sardinian-like North Italians(inclu. Otzi) from 5,000-4,000YBP are all typical Neolithic I2a and G2a. Bronze age Hungarians had some Yamnaya-like ancestry, but were much more Neolithic-descended than Bell Beaker in Central Europe, and all their Y DNA has come out G2a, I2a, and one J2a.

    Late Neolithic/Bronze Denmark/Scandinavia was similar to Bell Beaker/Unetice and have come out all R1a and R1b1a2, except for a few I1s. R1b1a2 has even been found in Bronze age Siberia with people who had alot of Yamnaya-type ancestry, but not in Neolithic Europe.

    I just don't see how R1b1a2a1a-L11 was in the original Bell Beakers or spread out Western Europe, unless a rare R1b-branch from Corded Ware went all the way to Portugal and back up.

    1. Thanks for the Snafu alert. I was rushing trying to get out the door.

      I'll fix this and hopefully get to the ultimate question hhere in the next few hours

    2. I did not read Allentoft's paper, but according to, all the Unetice yDNA reported by Allentoft is I2... Can it be a mistake in transcription?

    3. Correction: "all the Unetice yDNA reported so far".

      I had always read that Unetice was ancestral to Urnfield, Hallstat and La Tene (and the root to italoceltic languages). But if no R1a/ R1b appears in Unetice and I2 does, then: 1) Unetice did not speak indoeuropean (because they are not R1a/ R1b) and then the italoceltic languages came from another culture, or 2) Unetice people were I2 and spoke indoeuropean, and then R1a/ R1b people from the steppes spoke another language (not IE). Messy

    4. I'll try and comment here quickly...

      It's possible that they are all I2 if there is a selection bias for which graves are archetypal Unetice. Conversely, there would have been a bias for and against individuals in the previous periods.

    5. Unetice draws heavily from BBC in religion annd burial customs, in many cases in Beaker cemeteries. In some cases are mixed.

      By and large Unetice is Beaker 2.0 in a new age with maybe a new aristcratic caste. If there is a marked changed in the lineages of some, it may have come out of the Baltic since the apex of yhe Amber road was Bohemia.

      The old Beaker network also went down the Danube into the Balkans. So if there was some turmoil, it probably came out of one of its own legs (beaker)

    6. "I had always read that Unetice was ancestral to Urnfield, Hallstat and La Tene"...

      There's greater complexity in Central European Bronze Age than just Unetice. Sometimes people speak of Unetice as broadly representing the great diversity of cultures in this period but that's a way too loose generalization. Unetice senso stricto is a Czech, Central-East German and Silesian culture only. This map represents quite well the complexity of Early Bronze Age Central Europe:

      After this complex EBA period, the Tumulus culture seems to unify many of those groups in the MBA but its distribution rather points to a Straubing or otherwise Nordalpine origin and not directly to Unetice. From Tumulus emerges Urnfields, which can be considered loosely Italo-Celtic, then Hallstatt (Illyro-Celtic?) and finally La Tène (strictly Celtic).

      I wouldn't hype Unetice too much: it is indeed the reference EBA culture in Central Europe but it is not the only one nor "ancestral" to later developments in any direct way.


      As for what BBB says on Unetice being related to BB culture, I wouldn't bet too much on it. Genetically they seem quite unrelated. I'd rather see Unetice as direct descendant of Corded Ware with some modifications induced by the BB superstrate.

  3. So... the famous "Celtic" R1b-U152(=S28) subclade here. Nice, because it means that that it is clearly pre-Celtic, and it means also that R1b-S116 (its precursor) is necessarily much older and could not have spread in this late time-frame.

    You may also need to correct this: "It's presence among Bashkirs is noteworthy however"...

    1. I believe this comes from North Bashkirostan fromMyers et al 2010
      I'll double check later today

    2. I can assure you, based precisely on Myers 2010, that S28=U152 has no relation whatsoever with Bashkirs. Bashkirs have the Volga-Central Asian R1b subclade which is almost totally unrelated to West European R1b subclades.

      U152 is derived from S116, which is the SW European sublineage, with greatest apparent basal diversity in Southern France. It's therefore related to Irish, Basque, Iberian, etc. R1b sublineages.

  4. From wiki on Bell Beaker: "A review of radiocarbon dates for Bell Beaker across Europe found that some of the earliest were found in Portugal, where the range from Zambujal and Cerro de la Virgen (Spain) ran between 2900 BC and 2500 BC".

    The immediate R1b haplogroups immediately upstream from R1b-U152 cluster in Iberia, which fits nicely with what we know about early BB.

    Another thing I would note is that with R1b-U152 now known to be in Bronze Age BB, and with the Z2103 (Balkans) /L51 (Alps/Western Europe) split upstream from this, it fits nicely with R1b expanding into Western Europe well before the Bronze Age.

    I've also looked at the Admixture data in Allentoft. Yamnaya has some admixture from North Eastern Europe that does not appear in the Beaker Samples. So we can pretty much rule out these Bronze Age Yamnaya samples in the Allentoft paper as being the direct ancestors of the Beaker Samples.

  5. And we can also rule out CWC as the direct ancestor of BB, for the same reason.

  6. Just realized something. See Figure 2b in the Allentoft paper. The Montenegro sample (1000BC) looks very BB like. None of the minor admixture components from North Eastern Europe that you see in the CWC and Yamnaya samples. The rest of the components are more or less the same as BB.

    Fits with the timing of the Z2103 (Balkans) /L51 (Alps/Western Europe) split. And fits with the idea that during the Neolithic, BB ancestors from the uplands of the Balkans followed a mountain root along the foothills of the Dinaric Alps/Alps/Pyrenees into Iberia and Western Europe.

    This is what I've been thinking for the last year.

  7. I looked through the Allentoft paper for more information on the two Montenegro samples. Does anyone know the dates and context for these two samples?

    1. I have not had a chance to look at these. Were there any haplogroups assigned?

  8. Hi BB blogger
    The MBA Armenia shows a sensible Atlantic shift. It is R1b Z2106+ I first suspect that it is from Yamna, but know reading Your blogs comments I have doubts.
    Could it be that some of R1b are seafarers? Related to Troy?

    1. Well I can't comment on this SNP specifically since I really don't know, however I think that Armenia itself is central to PIE mythology and the general in the foothills of the Zargos is a better fit for a PIE urheimat.
      All of the people consumed with Yamna need to read about its origins as well as the Maikop. In short, I'd expect to see some sort of R1b in the Armenian foothills around the ceramic neolithic, probably R1a of some sort as well.
      It's from the Upper Euphrates and NW Iran in general that the pastoral societies launched to the grasslands of Asia, African and Anatolia. Yamnaya is just one of ray of a supernova.
      I think also that there is a gross misreading of the earliest ceramic users in the region on up to Karelia, but that's another topic

    2. @Aryam: Z2106 is the R1b lineage found in Yamna but most probably migrated from West Asia to the Volga at some point (overall R1b seems to originate in West Asia). Armenians (and to lesser extent Turks) also have some mainline European R1b (cf. Balaresque 2009, raw data) which must be attributed to later back-migrations from Europe, probably those that precisely forged Armenians as non-Hurrians (and non-Anatolian-IEs in case someone is in doubt).

      @BBB: What do you mean by Armenia being "central to PIE mythology"? I can't think of anything like that. PIE mythology is not even cthonic, as seems more natural for mountain peoples, but celestial, as seems natural for people living in flatlands and seeing only the empty sky above their heads. It's also a conqueror mythology that celebrates winning rather than sharing.

      Anyways historical Armenia was once Urartu, a clearly Hurrian country. No IEs to see there before Armenization, which surely only happened in the Iron Age, apparently as Phrygian colonization, per ancient accounts.

      "It's from the Upper Euphrates and NW Iran in general that the pastoral societies launched to the grasslands of Asia, African and Anatolia. Yamnaya is just one of ray of a supernova."

      Hmmm. If you mean that Neolithic spread from the Fertile Crescent in all directions, then I agree. If you mean that Indoeuropeans specifically spread from there, then I have to disagree, even if there is surely a Zagros are element in the constitution of PIE peoples (otherwise they were dominated by Eastern European aboriginal and some Siberian aboriginal blood). Yamnaya is anyhow the left-behind branch after the first expansion of IEs: Maykop/Kura-Araxes, Afanesevo and the various European branches were already detached by then.

    3. //Zargos is a better fit for a PIE urheimat.//
      Of course it is.

    4. The Hervella paper from last week has shown that rather than there being a large hiatus between the initial farmer expansion and the movement of PIE or Steppic people, that in fact there may have been waves of immigrants from and into the Near East, in some cases being distinct from each other, moving in all directions, including Europe. IOW, there were several supernovas.

      I think it is often a mistake to assume that every non-IE language is 'always' the native and former language of peaceful people is grass skirts who were victims. This has colored the discussion on the Indus Valley Civilization where academia has already 'determined' that IVC was Dravidian speaking, without any evidence.

      It's also highly improbably that Sumerian was native to Iraq, or most it, since it has zero internal diversity. It also appears that it could possibly have a PIE substrate (hypothesized by Gordon Whittaker and others)
      So, Uratu and Hattic and all of that is unconvincing. And to the last point, I think PIE religion is an amalgam of two traditions, one that is probably native to the steppes and Asia, and the other to the Northern Middle East

    5. I find you suggestions quite logical! and on Sumerian-PIE relation see this short article there are some good evidence-
      and yes can you give the link on Harvella paper please?:).

    6. @BBB: I can't imagine how the Hervella paper demonstrates anything of the kind you say. It just sequences a bunch of ancient Romanians. She hypothesizes something about the Balcan peninsula being an active corridor for cultural flows but that's not something that corresponds to Genetics but to Archaeology to determine. And AFAIK, there is no such archaeological evidence at all. I still have to read that study in full, so I'll leave it at that: that Hervella demonstrates nothing at all, she just throwns a conjecture or hypothesis (that can be easily demonstrated false).

      The IE problem is first of all an ethno-linguistic problem, a problem that demands a single more or less homogeneous origin (and very probably an origin c. 6 Ka ago, per glottochronologists' estimates). There are not thousands of different overlapping cultural flows all of which can be happily claimed to be Indoeuropean, nope. There must be one single ethno-cultural expansion process, with one single root and with a shared cultural baggage, at least in the early stages that can be discerned archaeologically. That process can only be the Kurgan macro-cultural expansion, pretty much as Gimbutas described it several decades ago (only some minor refining applies): there's no other such process anywhere in the archaeological record.

      The fact that you appeal to phrases like "peaceful people is grass skirts who were victims", means that you are not judging the matter on objective grounds but emotional-ideological ones. Not everybody who was victim of Indoeuropean expansion was peaceful, probably most were not, and we can find evidence of their violence (West Danubian massacres, opium as crop, burial with weapons) or their hierarchization (Iberian sites). The archaeological terms of Gimbutas theory are valid but her idealization of "Old Europe" is definitely not. That does not mean that there is not something to it and that sedentary agricultural populations probably had stronger community values than errant pastoralists, especially when they reach the "horde" evolutionary stage and become aristocratic conquerors with it.

      Let's get out of Eurasia and move to Bantu Africa for emotional detachment. We can see similar developments among Bantus: some were organized in communitary form (Congo, Ndongo, etc.), others structured themselves in machista warrior bands that were very successful in some cases, as they were best equipped to win at the battlefield; Zulus are paradigmatic, but there were many others with similar "Spartan" organization, often implicated in the slave trade like the Jaggas of Angola (at the extreme you get the modern LRA and Boko Haram, just that our technological modern era is no match for them, hopefully). We can't idealize these warrior hordes: they were what they were, and they left a lasting legacy, regardless that we may dislike it. It's the dynamics of the Metal Ages, in which hierarchization, particularly on the military side, was rampant, and not just among nomads.

      "highly improbably that Sumerian was native to Iraq"

      It is pretty well determined that proto-Sumerians descended from the Zagros. Their language is probably most closely related to Hurro-Urartean and NE Caucasian for all I can discern.

      Imagining PIE substrate everywhere does not help. Pre-PIE must have been related to other languages but other than some relationship with Basque (Roslyn Frank's "paleoeuropean") I can't see nothing substantial. It's still possible that "paleoeuropean" was actually something West Asian but considering the large amount of Paleo-European genetics in both Neolithic Europeans and Kurgan ones, I'd rather lean for a true aboriginal European connection. Debatable in any case.

    7. I don't think it's true that Sumerian has no internal diversity. According to "A Descriptive Grammar of Sumerian", there are northern and southern dialects. Also, the well-known Emesal variety doesn't seem to have been solely literary, since it shows up in Akkadian loanwords.

    8. @Maju @capra

      No internal diversity may be an oversell, but Gordon Whittaker recently made the case that it does not exhibit the kind of characteristics one would expect in a native language and that there does appear to be a substrate in terms of rivers, ancient cities and land features.

      'Grass skirts' applies to the resistance of the idea that IE languages are rarely replaced. While this is my own words, the nativeness of Hattic, Sumerian and antiquity of Semetic in the Peninsula and West Asia is really baseless.

    9. I would have to consider the merits and demerits of that Whittaker guy's proposal but, considering (1) that Sumerians were already in decline in 4th millenium BCE (when IEs begin their expansion from the steppes, which is coincident with the Semitic expansion from the semi-desert), (2) that they seem to have originated in a more northernly region, namely the Zagros Neolithic area around Jarmo, (3) that there is no indication of IE languages in the region before the Anatolian branch (Hittites, Luvites) and (4) that there is a large array of West Asian linguistic families (mostly extinct today but attested direct or indirectly) that are NOT IE nor anything of the like... (conclusion) I will have a very hard time believing that claim, so it better be extremely well exposed and backed.

    10. Let's make a comprehensive list of non-IE language families or isolates in West Asia since antiquity:

      1. Elamite (Elamo-Dravidian?)
      2. Sumerian, Hurro-Urartean and NE Caucasian (probably related IMO)
      3. Hattic and NW Caucasian (said to be related)
      4. Kartvelian (only documented historically, unsure if intrusive)
      5. Semitic (Afroasiatic)
      6. Eteocypriot
      7. Eteocretan (must have originated in Anatolia according to many)
      8. Tirsenian (Etruscan, Lemnian, etc. - almost certainly from Anatolia as well)
      9. Possibly also very early on a Nilotic branch leading (partly) to Vasconic, which I'd associate with Natufian-PPNA, excepted the Hairifian group, which would be proto-Semitic instead.
      10. Gutian and other trans-Zagros "barbarian" languages of unknown affiliation

      A few of these groups may be related among them but even after a cropping, there is still huge diversity in terms of linguistic families: too large to allow much room for other families like pre-PIE.

      Another different objection. Let's assume that effectively the Iran-Turkey apparent origin of R1a indicates the precursor of PIE in West Asia (it would not be PIE yet in any way but an ancestor): why to imagine it in Low Mesopotamia (Sumeria) rather than around Kurdistan, for example in the Jarmo group (which Nirjhar documented as partial ancestor of Yangelskaya, which might be in turn partial ancestor of Samara/Khvalynsk, i.e. the first Kurgans).

      And another objection: if Hurro-Urarteans and Hattic are not "native", then how come we don't see any exogenous (European or Central Asian or Siberian) lineages associated to their regions but rather we see West Asian elements in the North Caucasus instead? Where did those peoples came from and why we cannot detect their would-be exogenous genetic input?

  9. Aram,

    They could have been seafarers, but some people also could have walked. Pastoralists, even as recently as a hundred years ago, walk hundreds of miles in a single season to find pasture for their herds.

    Even in the Neolithic, they could have walked along the highland corridor that extends from Anatolia to Iberia.

  10. BB blogger
    Thanks for highlighting that Hervella paper. I think it's results match those from Unterlander's abstract from Greece: in that the major change (post Palaeolithic) in SEE - at least from an mtDNA perspective- occurred during the EN-MN transition rather than LN -EBA as in Central europe.

    Given the limitations of mtDNA- only studies, I think an upcoming paper making full shotgun sequencing from Varna, Bulgaria will clarify many things, and give long overdue insights from the Balkans.

  11. (Nothing to do with the main issue here) In the R-S28 map you publish, there's a nice pink area that is quite coincident with the Hittite Empire location... I wonder if Hittites were U152, just for the fun of it

    1. Actually the spot you mention is West of the main Hittite area and right on Ankara (ancient Ancyra, medieval Angora), which incidentally is a Celtic foundation (Galatians). So in this particular case I would say that the "Celtic" reputation of the lineage may be justified.

  12. It's pretty clear PIE had at least some of it's earliest influence from the Iranian plateau (perhaps Maikop Culture). Copper work, has it's origins in Iran. and the plethora of early R1b/R1a support an Iranian PIE origin.

    1. Unless those lineages you mention have no correlation whatsoever with Indoeuropean genesis. Only some derived branches of those lineages can be considered to have any relation with IE expansion, the rest are older and have different origins (and associated ancient languages, such as Dravidian, Vasconic or many others). People change culture, language and identity, they can't do that with their DNA.