February 29, 2012

On and around with Ötzi's genome

As you're probably more than aware by now there's a new paper on the market (yeah, 32 bucks - but worry not that I already got my hands on it) on the most loved mummy of Europe: Ötzi the Iceman.

A. Keller et al., New insights into the Tyrolean Iceman's origin and phenotype as inferred by whole-genome sequencing. Nature 2012. Pay per view.

The most notable conclusion would seem to make Ötzi closest in all to Sardinians or more like   Corsicans, at least by Y-DNA. This one has been described now as G2a-L91, what is per ISOGG 2012 G2a2b (although the authors use the old nomenclature G2a4) and is most commonly found in Southern Corsica and the Corsican-speaking parts of Sardinia (Gallura).

The autosomal DNA has been compared with an all-Europe sample (the Behar 2010 one, I think based on the nomenclature used), to which a Sardinian sample was added. The result (right) does suggest a Sardinian (or Corsican) affinity of Ötzi.

Notice please that in the supplemental material the Ötzi dot achieves three different positions depending on the level of refinement: while all place Ötzi to the bottom left corner, he's exact position varies quite a bit - it's not like PC analysis (nor genetics overall) is rocket science, you know.

Also another caveat I have with this kind of analysis is that all it says is that Sardinians and Ötzi are very negative for both PC components, th Northern and the Eastern ones. The only association at the bottom left corner is a negative one: neither Nordic nor Greek, and this is not too informative.

Yes, the Y-DNA points to an association with Corsica (rather than Sardinia), what reinforces the suggestion posited by the autosomal DNA basic (but negative) analysis, still it would be nice if the authors would have bothered to do some 'Admixture' type of analysis as complement. 

At the moment all we have is a negative: Ötzi, who belonged to a Cardium Pottery derived cultural group (Bocca Quadrata or La Lagozza, can't recall right now) and bears a quite clear Neolithic marker such as Y-DNA G2a, shows up as strongly non-Balcanic, unlike most modern Italians (Europe S sample).

It looks odd indeed... but it might be explained if we assume that from that time on, secondary (post Neolithic) Bronze Age flows from the Aegean (and Central Europe) altered gradually the genetic composition of Italy. This is supported by archaeology as far as I know: even before Mycenaean Greeks, the Aegean was influencing Southern and Central Italy more and more. This trend was reinforced in the late Bronze Age (Mycenaean colonization in the South, Etruscan migration in the Center) and the Iron Age (classical Greek colonization of Magna Graecia).

Before the Romans Italy was all or most of the time a recipient of cultural influences (from the Balcans, from SW Europe and from Central Europe) and did not, as far as I can tell, export culture except as secondary trampoline (the Cardium Pottery Case notably). Excepting the Cardium Pottery case, it acted more as a buffer between West and East and dead end than what its central Mediterranean position would suggest. Even in the Heraklean myth, original Greek version, the route to the fabled Hesperides does not go through Italy but North Africa. Only later, as the Romans rose to prominence, was Hercules made to journey back through Italy, something not specified in the original version. 

I'm saying all this because it may explain why the Europe S (Italy) component tends so strongly towards the Balcans (and to lesser extent Northern Europe) but neither Ötzi nor Sardinians do, even if they look Neolithic-blooded to some extent.


  1. I really like your refreshing analyzis, quite distinct from what one can read elsewhere.

    Still, if I get it well, it surely means that North Italy certainly and subsequently received a strong Balkanic genetic imput (well those are known facts, you sum up it well) but should a SW European imput be discarded ?

    Nevertheless, such results somehow seem to confirm that the ethnogenesis of Italy is rather complex with many contradctory influences in the game. Still it suits well what we already knew (for instance that North Italians are strikingly distinct from their "French" Alpine neighbours - f.e. people from Lyons - who never show that much "Neolithic" admixture on all the runs I could read).

  2. I am glad they finally released the data for Ötzi. He is of particular interest to me, as I am also a G2a. The Lebanon/Syria area is a possible source for G2a, which fits with the origins of the Cardial. It is unfortunate that we don't have more/better SNP data for the G2a in Spain and France so that we can find out if they were also G2a4s.

    Most of the modern G2a population in Western and Southern Europe are G2a3s and appear to have moved in during the Roman period. This would indicate that if G2a4 was dominant in the Cardium Pottery culture that they were all replaced by later groups, except in isolated areas and islands.

  3. "it surely means that North Italy certainly and subsequently received a strong Balkanic genetic input"

    I think that the two main clusters in which "Europe S" is divided should be North Italians and Tuscans or North and South Italians, right?

    If so the presumable North Italian group looks (on the graph) more like having received also Central European influxes instead. But it's impossible to say because PCAs are quite useless except for a quick first impression.

    "... should a SW European imput be discarded ?"

    When? Ötzi is contemporary from the last possible (?) Western flow: Megalithism/Bell Beaker, which overlapped with the local Cardium and even the earliest Aegean influences.

    But I do not think we can jump to any conclusions with just a PCA. We need well-done Admixture-style analysis.

    The differences between Italy and France/Iberia may be argued to originate as soon as the LGM, when Italy continued with Epigravettian while the West evolved into Solutrean first and then Magdalenian. The Alpine ice shield no doubt acted as tight physical border.

    Then, of course, the West Asian and Balcanic (Neolithic? Bronze?) influences seem much stronger in Italy but Italy also appears to have got some Western input at least in R1b variants (Megalithic? Celtic?)

    Can't say with greater certainty.

  4. "It is unfortunate that we don't have more/better SNP data for the G2a in Spain and France so that we can find out if they were also G2a4s".

    In truth I do not know but I can only imagine that commercial DNA testing companies will soon be offering "the Ötzi test" and we will get to know something thanks to that. XD

    However the claim of the paper is that G2a2b (aka G2a4) is not found ever at more than 1% frequencies anywhere in mainland Europe.

    "This would indicate that if G2a4 was dominant in the Cardium Pottery culture"...

    All we can say right now is G2a, plus E1b1b1-V13, plus I2a.

    "... they were all replaced by later groups, except in isolated areas and islands".

    I would not go so far. We still lack anything near a complete aDNA picture that allows us to say that.

    Something apparent at the Ebro Mediterranean-Basque frontier area we can appreciate two clearly distinct populations side by side but almost unmixed still in the Chalcolithic:

    1. The majority group (13) was probably Cardium-derived and lactose intolerant (CC)

    2. The minority group (5) was surely native proto-Basque and lactose tolerant homozygous (TT)

    3. Only two individuals were apparently mixed blood (TC).

    There is too much linkage disequilibrium to pretend that the population was an old blend: they were beginning to mix when these people were buried.

    So there was probably some aboriginal backflow on the Neolithics but its difficult to discern the geographic, cultural and temporal patterns at our level of knowledge. This probably also happened in Central Europe, where LBK haplogroup patterns do not exist anymore and instead we see modern ones since the Bronze Age.

    Instead in places like Portugal or North Morocco, a much more stable situation can be described from the known data.

  5. Probably not very interesting as individual studies just tell one family history but amongst my contacts on 23andme is a pure-blooded Catalan guy whose Y-DNA is G2a3b2 (he's DOD217 on Dodecad).

    1. Hello Heraus. It's the 1st time that I hear s.thing about a Spanish man in G2a3b2. I would like to share some information with him. Is that possible? My e-mail is kalabi-yau@hotmail.com
      Thx a lot! Ben

  6. Ötzi is G2a2b, not G2a3b2. They only agree up to the G2a.

    It is interesting because it says that at least some of the other West Mediterranean G2a is something else than Ötzi's (what I already sort of expected).

  7. "still it would be nice if the authors would have bothered to do some 'Admixture' type of analysis as complement."
    Agreed, I am sure the authors had all the tools to do some type of model-based analysis, I wonder why they didn't? It would have been especially helpful because they only show two dimensions (PC1 vs PC2).
    Also, since I don't have access to the publication, do you know who all the populations included in Europe S are? I think it would also be helpful to include Anatolian, Northern Levant and Caucus mountain populations in the PCA as well...

  8. "since I don't have access to the publication"

    I sent you a copy (confirm arrival, cause I'm not sure if that's your email or not).

    "do you know who all the populations included in Europe S are?"

    In Behar 2010, Europe S were an Italian mix. I think these are the same populations (European part without Jews) but haven't double checked.

    There's another graph including a West Asian and a North African sample, but I found it less informative and hence did not include it. Razib did post it but all the structure became blurred in some sort of "Mediterranean wedge" shape (North Africans on one side, Europeans on the other and West Asians at the vortex).

  9. Thanks Again Maju.

    I think I understand what they did, within a PCA context of North Africa, Europe and the Arabian Peninsula (Qatari), the iceman firmly landed within Europe (albeit closer to Southern Euros), however, within just a PCA context of Europe only, the Iceman landed closest to the Sardinians. However in the 'projected' (133K SNPs) PCA it looks like the iceman was further away from the Sardinians, but when they intersected the data with the sardinians from HGDP, although there were only 28K SNPs remaining , and hence less refined, the Iceman appears to be more closer to them, nevertheless closest to the Sardinians in both cases.

    The European data came from POPRES & HGDP, by the way.

    The Qatari data came from here: http://www.sciencedirect.com/science/article/pii/S0002929710002661 (Not sure if the North African data came from there as well)

  10. On the R1b Question:

    A few months ago, various Bronze-Age-R1b-Theory zealots claimed that because the Iceman was not R1b, it suggested that R1b was not yet in Europe in 3300 BC, Iceman's era. As I recall, some even said that G2a could be Paleolithic in Europe, which if so means a dramatic post-Neolithic population replacement.

    From an article about Oetzi:
    "It [is] clear [Oetzi] was a pastoralist--a Neolithic herder".

    "The Iceman's stomach held cultivated wheat, possibly consumed as bread; game meat, and dried sloe plums."

    He was a pastoralist who had recently eaten bread. Were pastoralism and farming ubiquitous yet in 3300 BC in his area of Alpine Europe? If not, doesn't this suggest that he was associated with a Neolithic migrant group anyway, and not with the Paleolithic stock of Europe? So no one would expect him to be R1b, except those who proposed R1b came with farming, maybe (if anyone still does).

    I suppose my questions are, Maju,
    (1) When do you think Otzi's ancestors arrived in Europe?
    (2) Does the genetic analysis suggest he has any Paleolithic ancestry?

  11. "... within a PCA context of North Africa, Europe and the Arabian Peninsula (Qatari), the iceman firmly landed within Europe (albeit closer to Southern Euros)"...

    Indeed. But the main axes of difference are so unrelated to Ötzi or Italians that the resulting graph is not really informative, maybe if they would have tried with Europeans (less North Europeans maybe) and some West Asians only (no North Africans), the result could have been informative. But one dimension there is taken by the Euro-African duality, while the other is, in Europe, dominated by the Nordo-Italian duality (not really N-S because Iberia is right in the middle, like Central and West Europe, although the Balcans do seem to cluster with Italy here).

    In the Nordo-Italian axis, Ötzi is clearly super-Italian, nothing Nordic about him.

    "The European data came from POPRES & HGDP, by the way".

    Sure, OK.

  12. Bronze-Age-R1b-Theory zealots...

    I like that description!

    Were pastoralism and farming ubiquitous yet in 3300 BC in his area of Alpine Europe?

    Yes, for over 1,500 years.

    As to G2a, it could go both ways. However, since Ötzi has a highly derived group by today's standards, it seems pretty certain that G2a/G2a2 was present right from the beginning of the neolithic, and may have been common enough (pre-neolthic) in the Balkans to be carried with and move west with Cardium. Even today there is such a trail of G2a in the Balkans - particularly along the Drava river that leads into southern Tirol (but also along the Danube and its tributaries, following LBK's path, and also in ancient DNA). We need to dig deeper into the sub-groups to be able to say more.

    If the Cardium people were highly admixed Balkans (likely IMO), then they were already close to local South Europeans from the get-go, so, autosomally, they didn't require much admixture with (then) Northern Italians to have a signature similar to Corsicans/ Sardinians (who certainly also have some traces of later IE/Nordic and West European inflow).

  13. "Were pastoralism and farming ubiquitous yet in 3300 BC in his area of Alpine Europe?"

    Yes, indeed. It was already the Chalcolithic: the age of early metallurgy, sophisticated hierarchical societies with specialist work, first cities, long distance trade... and wars, of which is possible that Ötzi was a victim. For example Aztecs and Incas were in the advanced Chalcolithic upon European arrival, with the first indications of Bronze being developed in NW Mexico. Similarly much of Egyptian or all the Sumerian civilizations were not more than Chalcolithic in technological level.

    The Alps were an area only colonized relatively late, according to archaeological data, but I doubt it hosted true hunter-gatherers either.

    "(1) When do you think Otzi's ancestors arrived in Europe?"

    Naturally he could have ancestors from many different cultural layers but the bulk of them probably arrived with Neolithic (Cardium Pottery).

    "(2) Does the genetic analysis suggest he has any Paleolithic ancestry?"

    It can't be denied but it is not apparent.


    R1b anyhow is not that important in Italy, where "Neolithic" lineages (J2, E1b, I2a, G2a) are rather dominant, excepted some troncal remnants, it belongs mostly to a single lineage that is probably derived from South or East France, and it is concentrated in the North.

    The real problem with R1b is in the most isolated areas of Western Europe: Iberia, Britain and, of course, Vasconia and Ireland, where it is clearly hyper-dominant. This dominance extends itself up to Central Europe at lower frequencies but still >50%.

    Per the phylogeny, R1b in West Europe probably stemmed from Central Europe and then from (a) Franco-Cantabrian region (the main clade) and (b) the German Bight area (Nordic clade, maybe from Doggerland?). This is coherent with a Paleolithic time-frame and is NOT coherent with a Neolithic or post-Neolithic time-frame, for which the Franco-Cantabrian region should have not played any major role.

    So if there was a Neolithic genetic tide, there must have been a "neo-Paleolithic" counter-tide. This is something that we are beginning to realize now: Neolithic lineages and genetics are not modern.

  14. "it seems pretty certain that G2a/G2a2 was present right from the beginning of the neolithic, and may have been common enough (pre-neolthic) in the Balkans to be carried with and move west with Cardium. Even today there is such a trail of G2a in the Balkans - particularly along the Drava river that leads into southern Tirol (but also along the Danube and its tributaries, following LBK's path, and also in ancient DNA)".

    That's very interesting, Eurologist. Are you planning to write an article on this matter on your abandoned blog? Hope so because I have no idea of this and I'd love to learn.


    For the rest I'd say that everything in Neolithic aDNA suggests a displacement or such an intense admixture with "ghostly" hunter-gatherers that it's about the same. This applies to the Western Med and to Central Europe.

    The problem is when and how did "R1b aborigines" push back from (necessarily) the West. This is not obvious in the archaeological record at all and only Megalithism seems to be somewhat consistent with R1b distribution (but not in North Africa nor Italy).

    We need more aDNA, much more.

  15. Maju, Eurologist,
    Thank you for your informative replies.

    A further question: Assuming Otzi was representative of the inhabitants of his region and time: What happened to the Paleolithic-Europeans of south-central Europe?

    As I understand it, there three ice-age refugia, one in northern-Spain, one in the Balkans, one in the Ukraine. The SW-Europe refuge has been thought to be associated with 'R1b', the Balkans refuge with 'I', and the South-Ukraine refuge with 'R1a'? So, after LGM, 'I' expanded north, but then was partially or mostly replaced in its southern 'Urheimat' by G or whatever else. It eventually repopulated those regions (a 'Reconquista', to use a fitting analogy from Spanish history).

    Possible sequence of events:
    (1) Neolithic migrants (like Otzi's people?) push the Paleolithic-descent Europeans to marginalism or extinction in some of Europe (Italy, e.g.).
    (2) The Neolithics had their heyday for a few centuries or millennia in some/much of Europe.
    (3) The Neolithic cultures stagnated and declined for some undetermined reason(s).
    (4) The Neolithics were then themselves pushed to marginalism by resurgent 'neo-Paleolithics' as Maju puts it (R1b, I, R1a).
    (5) Eventually the Neolithics were assimilated into emerging Bronze Age European cultures. They represented only small minorities of the gene-pools in most emerging European cultures, except some isolated areas, like Sardinia.

    Is this plausible? Comments, please.

    In trying to understand this narrative, I can't help but be reminded very much of the fall of Rome and the Voelkerwanderung period. In that case, another stagnant but once-strong culture that spanned the Mediterranean Basin was toppled and replaced by resurgent and energetic tribes from outside that culture's zone of control. If I read Maju correctly, that is exactly what he proposes for why Italy is today only 7% 'G', with North-Italy only 2.5% 'G'.

  16. It is plausible in general lines, Hailtoyou. I am thinking in the lines of your model but I have many many many questions and doubts. And that's why I'm begging for more aDNA data, because ultimately only large enough aDNA samples can provide direct evidence (this however may have a problem of bias against cultures practicing cremation or abandoning corpses to the vultures or scattering their burials or affected by soil acidity).

    In regards to point 3, you probably want to take a look to this 2009 entry on Central-Northern European Neolithic demography.

    What we see in Germany and Poland is a demographic boom right after Neolithic arrival and a collapse to almost pre-Neolithic levels (in Germany, less so in Poland) in some 1000 years. This may be related with unsustainable farming practices (slash and burn) but also with climatic cycles, as the Danubian Neolithic boom took place at the very apogee of temperatures between the Ice Age and present day (when, as you know, is just going crazy).

    There is also evidence in Northern Europe of certain "reconquest" by huntergatherers with the complex Funnelbeaker cultural phenomenon. But it's complex, because the influences are not just from the North but also from the East (Dniepr-Don-derived Pitted Ware).

    The mtDNA is better known than Y-DNA and, while in Iberia continuity (or almost) may be argued in most localities, in Central Europe modern-like samples are not found till Bronze Age (but there may be a big blank in sampling in the key Chalcolithic period).

    Regarding the collapse of Rome, there is evidence and consensus of not being any meaningful demographic change, as far as we can tell. In fact Roman culture is still dominant in Italy, Iberia and most of ancient Gaul, as well as in the remote colony of Dacia/Romania, and is central to European identity in general. So I do not think it's comparable at all.

  17. hailtoyou,

    Your scenario(s) are a bit too specific, for me. I don't think we have sufficient data to corroborate all of that, today - although I thing much of it is likely mostly correct.

    I think the fall and retreat of LBK would most likely be rooted in animals and grains too highly specialized, and climate change and diseases affecting these, as well as farmers - as demonstrated millennia later.

  18. But, Eurologist, the boom of British Neolithic, 2000 years after the initial LBK but largely LBK-derived itself, shows that it was still able to successfully exploit new areas in some (not too clear) circumstances.

  19. 2,000 years later, and not at all the same style houses or arrangement of animal husbandry, living quarters, and hay & grain storage. It took to ~late Roman times when typical continental farming houses first appeared on the English East coast.

  20. "(3) The Neolithic cultures stagnated and declined for some undetermined reason(s)."

    Just speculating here but i don't think a decline is neccessary.

    If the farmers hit a crop limit then people on the other side of it could adopt just the domesticated animals and become pastoralists. This maintains the high protein diet (and larger size) they had as foragers while maybe allowing for a population increase as well?

    The Romans and Greeks both talked about the peripheral people being physically bigger and stronger and although this isn't neccessarily a big advantage against a disciplined army, in tribal warfare between two groups with similar weapons and level of organisation it might provide the edge.

    So farmers begetting herders who later conquer them?

  21. We do see a quite sharp decline in the number of LBK findings, but we do not see that "pastoralist replacement" until later (if at all). In Germany, the approx. BCE dates are:

    - 5500 beginnings of LBK: brutal demographic growth
    - 5200 plateau begins
    - 5000 plateau ends, collapse begins
    - 4600 at very low levels, Lengyel begins

    Note: Lengyel culture is a more advanced LBK version originary from the Middle Danube, the core LBK area, where it lasted for long.

    -4400 after the lowest levels, the duality Michelsberg/Epi-Rössen begins

    Note: Epi-Rössen is continuity with late local LBK = Rössen culture but it is "decadent" in quality. It dominates in the South. Michelsberg is also Danubian but rather innovative, might be related with Funnelbeaker and might be that kind of "pastoralist" thing (with many doubts???). It dominated in the North and expanded southwards somewhat (Westphalia notably). I have the impression that both cultural groups were immerse in some sort of lengthy "civil war", helping the arrival or Indoeuropeans that way. IMO the core Western IE group (Baalberge culture) might have been summoned as mercenaries in this struggle, getting a peculiar Danubian tribal district of the Elbe as reward.

    But these distinctions are very local to the West Danubian area (Germany, Low Countries, Switzerland, NW France) and can't be extrapolated to other regions.

    Also the Michelsberg "solution" only coalesced after the collapse of LBK was complete, so it was not a cause but a consequence.

    And never mind that the ones who really ripped the benefits would be the Indoeuropeans, who take over the area since the 4th and specially 3rd millennium BCE.

    One of the big issues with R1b is that it's so homogeneously dominant in all West Europe but, since Magdalenian (or epi-Magdalenian), there has not been any such homogeneous culture in all the area. Actually Magdalenian could only explain R1b1a2a1a1b (P312/S116) and we need older UP layers, like Gravettian or Aurignacian, to understand the link with the NW European clade R1b1a2a1a1a (M405/S21/U106) and the smaller "asterisk" occurrences of R1b1a2a1a (L51/M412/S167), the Western clade of R1b.

    But the problem is that we now stumble on Neolithic findings that are apparently inconsistent with this model, yet no other explanation is provided nor apparent in any way, with all that scatter of diverse cultures and lack of any single phenomenon that could justify this uniformity.

  22. Maju
    "but we do not see that "pastoralist replacement" until later (if at all)."

    "One of the big issues with R1b is that it's so homogeneously dominant in all West Europe but, since Magdalenian (or epi-Magdalenian), there has not been any such homogeneous culture in all the area."

    Yes. The thing i'm hooked on - because of coming to this from history rather than genetics or archealogy - is the Romans and Greeks remarking on the physical size of the northern tribes which i assume was related to diet(?) so i circle around that as the explanation but as you say it has to fit the other evidence.

    I suppose if there was a south to north or southeast to northwest cline in the *proportions* of pastoral vs arable diet using the same overall farming package i.e. more meat and/or milk/dairy and less grain in the diet as farmers moved north/northwest because of reduced grain yields then you could have the same broad culture groups but still get a gradual south to north cline in protein diet driving physical size driving a reverse expansion?

    That i guess would be testable if there was enough evidence of diet in enough sites along the cline - or i guess size of bones going north - south and SE - NW?

    Thinking aloud.

  23. * I don't think I've seen a post from you (Maju) before this one which is as receptive to really dramatic population genetic shift anywhere in the last five thousand years in Europe. Your acknowledgement that this does seem to have happened in Italy is noteworthy on that score.

    I certainly agree with you on the notion of Otzi seeming like a typical latose intolerant, G2a Cardium Pottery Neolithic individual (and by implication that Corsicans and Sardinians have been most stable since the Cardium Pottery Neolithic), and I likewise agree with you that the proto-Basque were part of a different wave of population with very distinct population genetics.

  24. @hailtoyou

    "A few months ago, various Bronze-Age-R1b-Theory zealots claimed that because the Iceman was not R1b, it suggested that R1b was not yet in Europe in 3300 BC, Iceman's era."

    The unfortunate fact is that we have almost no Paleolithic Southern European ancient Y-DNA, even though we have some mtDNA from that era. Ancient DNA does tell us that Y-DNA haplogroup R1b has yet to be discovered in ancient DNA from either the LBK Neolithic or that Cardium Pottery Neolithic, and while the number of early Neolithic ancient Y-DNA samples is not huge, it is still large enough to conclude that even if R1b was present at some percentage in either of those populations, that it was uncommon, and that it could have been rare or non-existent in the original Neolithic migrants (setting aside introgression with whatever Y-DNA haplogroups the locals may have had).

    This riddle then, is why haplogroups that were so common in the early Neolithic populations, like G2a, are so rare in most of Europe now, while specific subtypes of R1b are dominant in Western Europe and specific subtypes of R1a are dominant to the East (with some blending of the two in boundary areas). There are basically three ways to explain this (although one can make minor variations or blend them to some extent):

    (1) early Neolithic people and later waves of migrants were drowned in the demographic sea of Paleolithic European indigenous people who assimilated the Neolithic technology, outside places like Corsica and Sardinia and Swiss-Italian high country that may have been unfavorable to Paleolithic life styles for some reason allowing early Neolithic newcomers to become dominant there. Since R1b stays dominant, later waves of migrations from within the R1b zone are also R1b, but with serial founder effects. This is a Paleolithic continuity hyypothesis for R1b dominance.

    (2) The partilines of early Neolithic settlers swiftly become dominant, with Continental Europe looking genetically a lot like Sardinia and Corsica do today. Then, one or more waves of migrants in the Chalcolithic, the Bronze Age and the Iron Age, in turn, demically overwhelm the Neolithic settlers in turn, until early Neolithic genes are reduced to a tiny percentage of the overall populations. This is basically the migrationist hypothesis.

    (3) Early Neolithic Y-DNA haplogroups like G2a became less common relative to the Y-DNA haplogroups that are now dominant in Europe because Y-DNA haplogroups are, contrary to the unusual assumption, not selectively neutral at the individual level (i.e. apart from participation in a culture for which a haplogroup is a selectively neutral ancestry informative marker). For example, maybe people with G2a are more vulnerable to dying from influenza than people with R1b. In this scenario, over enough generations, regardless of the starting point distribution, selection will relentlessly reduce the frequency of G2a relative to R1b until G2a becomes very rare and R1b becomes very common. This is a gradual selection hypothesis.

    So, what can of evidence can discriminate between these three theories? The data from the last couple thousand years, which is comparatively rich and includes some pretty hard times seems to disfavor gradual selection.

    Cases like the great demographic turnover in Italy lead me to favor the migrationist view. So does the fact that mtDNA tends to be more stable than Y-DNA in populations, yet a decent volume an ancient mtDNA samples from Northern Europe (more or less) show a dramatic reduction in the frequency of mtDNA haplogroups like U5 on the order of 80% compared to current levels, so one would expect Paleolithic Y-DNA to be at least somewhat more reduced there, which disfavors a Paleolithic continuity in Northern Europe without real massaging of the idea.

  25. Thanks everyone -- Maju, Grey, Eurologist, Andrew -- for a great discussion in this thread.

    I believe we can divide your scenario-2 into '2a' and '2b'.

    (2a) Intracontinental Migration (Maju's "neo-Paleolithics")
    The peoples which overwhelmed and replaced the Neolithic cultures came from some non-colonized regions of Europe itself, the people of Paleolithic stock (a continental Reconquista). I am reminded again of the Fall of Rome scenario, in which Europeans outside the Roman sphere of influence took over a long-declining Roman culture that had lost the will to live on. This may well be what happened to the Neolithics, perhaps impelled by the reasons Maju mentions in a post marked March 1, 2012 12:32 PM above. (Although Maju points out the fall of the Neolithics may not be comparable to the fall of Rome, because there was not a near-total-replacement in many, if any, places. But it fits in the sense that in both scenarios we see a declining European culture overtaken by its European neighbors, and not distant migrants from Asia).

    (2b) Inter-continential Migration
    The people replacing the weakened Neolithic cultures came from outside Europe itself, from a population unconnected with the European-Paleolithics. This is a present-day version of Aryan Theory, it seems to me. A few papers in recent years have endorsed it, including several by Maju's favorite, Balaresque ;-) An equal number seem to be against it.

  26. A general comment:

    One problem, I think, is excessive focus on genetics to the exclusion of all else. Genetic studies very often raise more questions than answers. Speaking in terms of physical-anthropology alone, it's clear that a great number of today's Europeans look like the reconstructions of Cro-Magnon Man that have been made from Paleolithic skulls. The most reasonable explanation is that these Europeans are most likely of heavily-Paleolithic ancestral-stock. Is it not?

    Now to incorporate genetics in a subsidiary role: Cro-Magnon physical types predominate especially in areas of high R1b presence, it seems to me. This further strongly suggests that R1b is Paleolithic in Europe.

  27. As said above, this is a very interesting discussion - and surely one lacking a clear answer as of now. I'm just going to reply here to the penultimate post by Hailtoyou because there are a couple of details I disagree with:

    1. The fall of Rome scenario has really nothing to do: it's a political rather than demographic change, even where Vulgar Latin was replaced by native (Basque, Berber, Celtic) or immigrant (Arabic, Slavic, Germanic, Magyar) languages, and some elites were of foreign patrilineage (like Abderhaman III: 3/4 Basque, 1/4 Ummayad ) the population change was obviously so small and anecdotal that it is a counterpoint and not at all a valid comparison.

    To date tracking anything remotely indicating the Swedish/Polish/Romanian origin of the Visigoths in Spain, for example, or the Arabic origin of the Ummayads (or even the more plausible Berber origin of some of their soldiers) in the genetic pool has proven futile, indicating that there was not just nothing like demic replacement but not even noticeable immigration almost at all.


  28. ...

    2. Balaresque 2010, which is a very poor paper with a biased intent from the beginning (they were dead set to "demonstrate" the Neolithic hypothesis even if that meant cheating pathetically - I'm really angry with that team, seriously: how can you pretend to be a scientist and be such a hypocrite?!) and only useful for its raw data, claims a Neolithic (and not post-Neolithic) dispersal of R1b in Europe, what has been, I understand reasonably dispelled by the known fossil Y-DNA found in the Neolithic cultures' burials (as well as by the fact of R1b having a structure, much of which seems centered phylogenetically, not in Yugoslavia or Italy but in the Franco-Cantabrian region).

    Balaresque's paper has nothing to do with Bronze Age hypothesis anyhow, notably because it claims an expansion from Anatolia, not Central or Eastern Europe.

    The rapid expansion for R1b as a whole of for the Euro-Anatolian clade R1b1a2 (M269) is untenable without a prehistoric (archaeologically consistent) model of expansion from the Aegean area. Bronze tech did expand from there but not any specific culture we can identify); in fact the Aegean interaction in SE Iberia existed but was with Mycenaean Greeks (a low R1b, high E1b, J2 people), who did not even legate identity nor language (Iberian was the actual language spoken after that).

    The only Bronze Age expansion of some sort identifiable in the archaeological record is Celtic/Western IE (Urnfields, Hallstatt and La Tène successively) but alas Basque are not Celtic at all and the area was only marginally touched by Celtic influences, which seem mostly hostile, being a consensus among archaeologists of local continuity in the Basque area, incompatible with Y-DNA replacement so late in time (or actually at any time, for example Santimamiñe cave has a full successive record from Gravettian to the Iron Age, not being any exception but rather paradigmatic).

    The pre-Celticnes of Basque language and identity (as well as others like Iberians, Tartessians, maybe Picts) stands as an absolute barrier to allow the IE hypothesis of R1b expansion having any ground. In fact, if Urnfields, Hallstatt and La Tène would have starred any expansion, we should see the pattern still visible in highest phylogenetically basal diversity in the Rhine area. We do not, or at least the Rhine-plus area does not stand out as likely original of the major subclade R1b1a2a1a1b (P312/S116), which is strikingly SW European.

    The only "hope" for a post-Neolithic expansion time-frame of part of the R1b haplogroup and such, could be Megalithism. And still it should only affect some sublineages and not all the R1b1a2a1a1b (P312/S116). Also Megalithism can't explain North Italy, nor Central-East Iberia, nor some parts of Germany and Britain where Dolmenic Megalithism never really existed (and instead it did exist in North Africa, where R1b is low).

  29. I agree with Maju, here - y-DNA-wise.

    You can attempt working things backwards, and it ain't looking pretty. We have Norman, Anglo-Saxon, and Frisian-Dutch invasions into the Isles - yet nothing changes much. No genetic shift except perhaps a little more R1a,and some shifts in R1b and I sub-groups.

    Than we have Roman dominance without much at all except a trace, and then Celtic - but from a continental source almost indistinguishable, in y-DNA.

    See the pattern? All these "replacing migrations" do nothing but enhance y-DNA diversity, slightly - no replacement, at all.

    Perhaps IE dominance was only a few centuries before the Celtic cultural maximum, perhaps it was (likely) a millennium or more before it - but it does not change the fact that concerning y-DNA, there remain few differences to the continent regardless whether regions were Celtic (Switzerland) or not (Northern Germany, Poland, and Scandinavia).

    Perhaps arrival of agriculture, then. What we would expect, here, is a discernible North-South gradient. Both the Isles and Scandinavia received agriculture very, very late. But again, no: no N/S gradient.

    In the end, the only reasonable pre-agricultural groups left are the ones we find today - R1b, I, and R1a.

    No migration has seemed to change this.

  30. "This further strongly suggests that R1b is Paleolithic in Europe."

    I'm not seriously putting this idea forward but i do wonder about neolithic or even pre-neolithic people from the med taking a hopping coastal route to Southern Portugal which then becomes a base for trading, fishing and mining settlements along the Atlantic coast.


    If their crop package failed in those settlements because of the ocean climate they might have turned to fishing and cattle instead.

    And if high levels of LP came about as a consequence of relying on cattle and it first reached a high percentage among an R1b subclade population then it might have spread all along the coast from those founders?

    Either way i think there was some reverse expansion - maybe caused by diet - but whether this reverse expansion was conducted by paleos or a mixture of particularly adventurous Sherden who got lost in Ireland and paleos i wouldn't guess.

    I do find that megalithic map very intriguing though.

  31. I have been pondering a Megalithic expansion "option B" for R1b in Europe since at least 2008 (only to be dismissed by Neolithic-model fans once and again).

    It has several problems however:

    1.- it does not fit well enough, with high R1b non-Megalithic areas like Central and East Iberia, North Italy, East England, Central Germany, Austria... accompanied by low R1b Megalithic areas like North Africa or South Italy.

    2.- (Dolmenic) Megalithism is often understood as a super-cultural phenomenon, with many variants and local cultural substrates and never argued to be a single culture with a single expansion pattern of any sort. Challenging this would require at least some very serious prehistoric rethinking (and supporting archaeological evidence).

    3.- South Portugal (or SW Iberia in general) does not even seem so far the most dense nor diverse R1b area of Europe. Although it's true that Iberian R1b1a2a1a1b-P312/S116 is abundant and still needs a lot of study, even in the best case only this "South Clade" could be argued to have any relation with Megalithism, and North of France and East of ireland it is the "North Clade" R1b1a2a1a1a-M405/S21/U106 which makes up almost 50% of all R1b. This Northern clade clearly, even if a relative, has a different origin, maybe in Doggerland, Germany, Netherlands...

    4.- We know that E1b-V13 (and other Neolithic markers) now found in Portugal were already around at the founding of the Domenic Megalithism phenomenon in that area, c. 4800 BCE. It is most unlikely that these markers did not play some role in the proposed Atlantic Megalithic journey, establishing some founder effects or gene pools similar to that of Portugal all around the Megalithic zone.

    So I won't say that Megalithism did not play a role but it's not the only mechanism for the Western European R1b spread. It may have been central in the configuration of gene pools in NW Iberia (which is indeed similar to that of Southern Portugal) or maybe Ireland (assuming a very marked founder effect from, say, Brittany) or... you tell me. But it is not a single origin massive scatter.

    There may well have been other such secondary "distributors" like Celtic-plus migrations or less popular expansive phenomena like Artenac culture in Chalcolithic West France, Chassey-La Lagozza in late Neolithic SE France and North Italy, etc. But I do not think at this moment that any single phenomenon can be pin-pointed as sole culprit.

  32. What i was imagining was a westward expansion wave hopping along the coasts and islands of the med with SW Iberia / N. Africa as the high tide point with much smaller numbers than settled Southern Italy and less reinforcements from the eastern med over time - so wide-trading but demographically isolated from their root.

    Because of this i was thinking the second wave of expansion was less a wave and more a ribbon or thread along the Atlantic coast following fishing grounds and exploring around the coasts and the planting of small settlements near good harbours for their boats to run to in a storm - or temporary fishing camps that eventually turn into permanent settlements like in Newfoundland. Once settled and exploring inland plabnting other settlements near gold and silver mines in Ireland and Wales, flint mine in Jutland, later tin mines in SW England etc.

    So in short a thin population of trader-fisher-miners stretched out like a ribbon along the Atlantic coast possibly trading trinkets with the local foragers for food and protection.

    That might cover point 2.

    If their crop package worked fine in SW Iberia but began to fail further along the Atlantic Coast (around Galicia onwards?) because of the wet climate -


    - but for the same reasons their cows produced more milk - they may have switched to a heavier use of cattle and fishing for their diet.

    Then if high LP began to be selected for in one of the areas where they'd planted coastal settlements (Franco-Cantabria?) and that was associated with the south clade of R1b1a2a1a1 it could then spread to the megalithism areas like Southern Ireland and Southwest Britain.


    This could fit point 3. Although the base culture stems from SW Iberia the haplogroup element is a secondary effect from a secondary settlement.

    The north clade is created the same way but later and further north - the Jutland flint mine? - and then spreads separately which might fit point 1.

    In short.

    European R1b is pre-celtic and a result of the introduction of cattle into an ideal habitat by a trading-mining-fishing superculture originally based in SW Iberia that planted trading-fishing-mining settlements along the Atlantic coast alongside the native foragers.

    The south clade is the marker for an early expansion (from Franco-Cantabria?) to the places associated with megalithism: Franco-Cantabria, Brittany, Southern Ireland, SW Britain, Jutland?

    The north clade is a marker for a later event
    - coming out of Denmark or that vicinity(?) and spreading more by diffusion than hopping between coastal settlements - funnel beaker?
    - later event from a secondary expansion of cattle to regions outside the megalithism culture: Doggerland, Germany, Holland?

    Maybe the beakers were for milk :)


    As to point 4 if the R1b was associated with LP and LP caused a population expansion from a secondary settlement like Franco-Cantabria the E1b may have been swamped?


    "But it is not a single origin massive scatter."

    Yes. I'm thinking the bits on the Atlantic side of this:


    grew into the northern red areas of this:


    with a later expansion into the green areas of this:


    which may have been the north clade *or* the spread into new areas may have sparked the north clade.


    i'm humbling bits of information here so no doubt there are holes in this.

  33. "European R1b is pre-celtic"

    Meant to say the south clade of R1b1a2a1a1 is pre-celtic.

  34. "... so no doubt there are holes in this".

    I see too many holes. Things like "why cows and not sheep, which is at least as important if not more?"; or "hy that obsession with milk?, when it seems that the fixation of the lactose tolerance allele was a fluke, not adaptive"; or "sure, the Neolithic package might have stumbled on some sort of climatic problem at the Atlantic but it clearly overcame it since very early"; or "trade? Trade only became important since c. 3000 BCE, some 2000 years after Neolithic!"; or "I really fail to capture any relation between the Chalcolithic cultural maps shown and the R1b reality".

    Whatever the case I lack models to explain that the Neolithic samples in Languedoc and Catalonia (n=4) had R1b=0% and now the population has R1b>60%, just as nearly everyone around. Something is amiss but I don't see what.

  35. Maju wrote:
    "Whatever the case I lack models to explain that the Neolithic samples in Languedoc and Catalonia (n=4) had R1b=0% and now the population has R1b>60%, just as nearly everyone around. Something is amiss but I don't see what."

    Looking back at this thread on the Languedoc findings, I find that you, Maju, write that the people were probably "of immigrant origin."

    Commenter argiedude: "The 22 y-dna samples are clearly just 2 unrelated lineages, one a G2a, the other an I2a. There's no doubt wahtsoever about that, maybe in the case of I2a by a very long shot they are actually 2 unrelated lineages, but it's extremely doubtful. If R1b was 60% in the region back then, as it is today, then the chance of not finding R1b in a population of just 2 samples is 1 in 6, unexpected, but perfectly possible."

    Real-world probabilities are never as they are in the textbook. Considering also differences in burial practices, (as you often point out), or other factors I may be overlooking, the chances of not finding a major y-line in a given place in a few findings may be very high.

  36. "Things like "why cows and not sheep, which is at least as important if not more?"; or "hy that obsession with milk?, when it seems that the fixation of the lactose tolerance allele was a fluke, not adaptive"

    The primary reason is the
    - heaviest distribution of European R1b
    - heaviest distribution of European LP
    - heaviest distribution of Atlantic coast rainfall
    all closely overlap.

    This is correlation not causation but i find it intriguing because

    2) Cows can produce very large quantities of milk (calories) in the ideal conditions for no extra work and those ideal conditions (of very heavy rainfall and lush pasture) happen to overlap with the heaviest density of (south clade) R1b and LP.

    3) Dairying that extra milk would take extra work but drinking it wouldn't.

    4) Populations expand when they find a way of producing more calories on the same land and/or for the same (or in this case less) work.

    Fluke or not (and on reflection it makes more sense as a fluke) if

    that high rainfall strip along the Atlantic coast could produce more calories by focusing on milk than could be produced using the other available alternatives

    and if

    some population along the Atlantic coast developed LP so they could drink the extra thereby taking advantage of more calories for less work

    and if

    that group was R1b (south clade) and part of the megalithic cultural range with contact by sea to the other parts then they could have expanded to those parts of the cultural range which had the correct climate e.g Southern Ireland and SW Britain but not back to the places which didn't e.g. Southern Iberia.

    In short even if LP fixing was a fluke and the fluke happened among an R1b population purely by chance then that population could then subsequently have expanded to those places where that kind of farming was more efficient than the alternatives available at the time taking their R1b and LP with them.

  37. @Hailtoyou:

    Most importantly, the negative results for R1b were repeated in Catalonia (again effective n=2, adding up to n=4), what makes the odds of mere chance quite lower.

    It may be a coincidence but, so far, the data suggests that Cardium Pottery people carried no R1b worth mentioning. Their mtDNA pools were also not modern, with low amounts of H and high amounts of "neolithic lineages" like J1 or X2.

    Meanwhile contemporary Basque sites like Paternabidea show modern mtDNA pools, while the lactose tolerance duality found in the Upper Ebro military burials from the Chalcolithic, strongly suggests two populations: a "Basque" one with modern (and arguably native) genetics and a "Mediterranean" one with 'exotic' genetics that now make up only a smaller (but still important) fraction of the DNA pool in the larger region.

    But I still do not know how to explain the "Atlantic backflow", which seems more and more obvious as we speak.

  38. @Grey: You have a potential point in principle but I must distrust single element explanations, specially until we find an archaeologically plausible explanation for such expansion.

    Anyhow, milk is not so important in most diets and has never been. If you look at socio-economic history of the Middle Ages, the staple food were cereals without doubt and milk was mostly used to make cheese (and possibly other related dairies like fresh cheese, curdled milk and certainly butter). The main use of milk-producing animals was typically other than milk: bovine cattle was used mostly for traction in the fields and transport, sheep's main goal was to produce wool. Only goats, largely restricted to poorer areas, often mountainous ones, were herded primarily for their milk (goats are excellent milk producers and very cheap to maintain) but they do not seem to hold any correlation with lactose tolerance (they were specially abundant, it seems, in Italy, yet Italians are largely intolerant). This is surely because the milk was used to produce cheese and not drank raw.

    It does seem true however that Dolmenic Megalithism may have some correlation with pastoralism, specially in the Pyrenees and other mountain areas, but I fail to see a plausible mechanism for any "aboriginal" takeover of the Mediterranean countries or Central Europe.

    I need archaeological support, everything else seems mere speculations.

  39. Anyhow, milk is not so important in most diets and has never been. If you look at socio-economic history of the Middle Ages, the staple food were cereals without doubt...

    Or milk and cereal - as in Müsli.

    In cereal, raw, in coffee and cocoa etc., I have consumed about a liter of low-fat milk or more every day for more than half a century, and I think it's difficult to get the same caloric, protein, vitamin D, calcium, and other benefits as easily from other sources.

    Apart from the general nutritional value, lots of people suffer from Calcium and vitamin D and folic acid deficiency - which are correctable via milk consumption.

  40. The closest think is porridge (which may have water instead of milk) and was apparently important before bread took over in most of Europe. Naturally milk can be used for cooking but there are so many variants and often water and eggs are also involved that it's impossible to decide how important it was.

    "I have consumed about a liter of low-fat milk or more every day for more than half a century"...

    Not me. I am lactose tolerant as almost everybody here but adult milk consumption is much lower, typically restricted to some desserts and with coffee. I wouldn't know how to combine a whole liter of milk for daily meals and I am sure I do not need more calories anyhow.

    If you want more calories, I'd suggest you to drink the milk whole or, even more radical, eat butter or cheese instead.

    "... it's difficult to get the same caloric, protein, vitamin D, calcium, and other benefits as easily from other sources".

    Milk is an excellent source of calcium but not at all of vitamin D, which is only found in meaningful amounts in fish (and the sun). Natural (non-enriched) milk won't save you from rickets or vit. D deficiency in general.

    Milk is good for calcium, protein and fat (raw calories) essentially. You can get most of that from cereals alone if combined with legumes or occasional meat/fish/eggs.

  41. "but I fail to see a plausible mechanism for any "aboriginal" takeover of the Mediterranean countries or Central Europe."

    Yes i jumbled a bunch of separate ideas together. The main idea is the possible spread of the south clade of European R1b to those places where
    - heaviest rainfall / best pasturage
    - that clade of R1b
    - megalithism
    overlaps i.e Franco-Cantabria, Brittany, SW England, Southern Ireland, and maybe to a lesser extent NE Scotland and Jutland.

    The key point being the same number of the same type of cows on the same sized field produce much more milk in the ideal terrain/climate so the first group that adapted to those conditions - either through LP or simply switching their priorities to cattle and fishing (like the Danish Ertobolle culture) might have expanded but *only* to those areas where the ideal conditions applied.

    "Or milk and cereal - as in Müsli"

    I was thinking porridge too.

    "I wouldn't know how to combine a whole liter of milk for daily meals"

    Maybe unusual but in my late teens and 20s i used to drink four pints a day when i was doing hard physical work. I'd have about a pint in cereal and tea over the course of the day plus i'd drink one pint on the way to work, one at lunch and one on the way home.

    I love milk.

  42. I don't think that drinking so much milk is normal, not even in Nordic societies. Still I asked a friend what he thought about milk consumption here in the past and he said: goat milk and acorn bread, that's what people ate (long ago). He "knew" only from vague references anyhow but there goes a second opinion.

    I know that acorn bread is correct because it's mentioned by Herodotus or some other classical geographer but never before heard that goat milk was so important - so I still have my doubts. It'd be goat milk and not cow nor sheep produce in any case.

    Megalithism does indeed have a marked overlap with R1b in Europe but, as said before, there are also many "holes" in that pattern: both Megalithic areas low in R1b (North Africa, Southern Italy) and very specially many non-Megalithic areas high in R1b (Central and East Iberia, North Italy, North France Central Germany, East England, etc.)

    ... many doubts, you know.

  43. "many doubts, you know"

    Sure. It is just interesting (to me) speculation.

    The second and separate idea - again mostly spurred by map correlations, in this case of the overlap in the distribution of funnel beaker culture and the north clade of R1b



    - is wondering if the same thing might have happened again generations later i.e. the intoduction of cattle to the megalith culture part of Denmark not leading to a big expansion because the climate - judging purely by the rainfall map


    - isn't as ideal for cattle as Ireland but after generations of improving the cattle breeds maybe it led to an expansion by a group who had the north clade R1b into the funnel beaker area - and also accounts for later expansions from that same terriotory like the Anglo-Saxons to eastern England.

    Like i say this is all just idle speculation based on looking at a lot of maps so not to be taken too seriously.

  44. The map of the haplogroups is very good: even if one can probably question exactness of some details or the opportunity of some labels, it really gives the correct impression overall and the possible errors are minor.

    On the other hand the map of "Old Europe" is mostly junk. Vinca (or Vincha) is mostly a Serbian culture with connection via Macedonia to Thessaly (Dimini culture). While it had some influence in parts of Valachia after the 4th millennium IE-Kurgan raids, and earlier in parts of East Hungary, most of what is shown as "Vincha" in that map is actually something else.

    Also no LBK in Dalmatia-Bosnia, which were all the time before IEs in the Cardium area, as well as Montenegro and coastal Albania.

    TRB is many cultures in fact and only the Denmark-Scania bloc can be considered homogeneous, and arguably the origin. Its inclusion of Michelsberg culture is not acceptable: Michelsberg is essentially LBK, even if "reformist", not displaying funnelbeakers at any moment AFAIK. In any case consider Funnelbeaker a complex phenomenon, not any simple vectorial expansion.

    Do we know where is there the greatest basal diversity of the North Clade (R1b-U106)? It'd be nice to know which is/are the most plausible origins: Austria, Netherlands, Denmark, Doggerland, Bavaria, Low Saxony?

    One issue I have is that we lack of any ancient DNA data for West Germany between Early Neolithic and Urnfields. The same can be said for Low Countries and North France, so that is really a big blank of data. Modern-like populations could have been formed at any time between the Mid-Late Neolithic and the Late Bronze Age.

    A candidate is naturally Bell Beaker, whose demic influence from the Rhine Eastwards may have been quite important (unlike in the West).

    As for cattle, I'm not even sure that they were so important for milk then, but rather for traction (role number 1) and meat (#2). Sheep and goats were at least as important re. milk.

  45. yep - it's an interesting idea but more data needed.

    "As for cattle, I'm not even sure that they were so important for milk then"

    Yes, but like i say the critical point is simply that it's free extra calories simply from location. The same cow produces more milk in Ireland than it would in Greece simply because of all the rain. I think you're right i have been overstating the scale of the bonus but say for the sake of argument the farmers in the heaviest rainfall parts could get 1/2 to 2/3 pint extra milk per person per day. The question remains what to do with that bonus?

    Dairying the extra milk would take extra work which would reduce the overall cost benefit so the ideal would be some way of using the extra milk that didn't require extra work and which was suitable for both LP and non-LP.

    So now i'm wondering more about porridge especially in conjunction with eurologist's point about muesli. Muesli is eaten cold. Porridge is usually eaten hot. I suppose heating it kills the lactose? Not sure.

    So maybe
    1) extra milk as an accidental benefit of the Atlantic coast climate
    2) dairying the extra milk would require extra work reducing the cost benefit of the free extra milk3) So invention of bowl of breakfast cereal in the morning with a pint or so of milk mixed with oats or something
    - c400 calories if eat it heated
    - c700 calories if LP and eat it cold

    (or 200 and 350 if 1/2 pint per person)

    That sequence makes more sense actually as if a population starts with only c10% LP then the diet of the whole population won't shift to accomodate just that 10% whereas an addition to everyone's diet - a bowl of milk and oats in the morning - with just a minor difference - eat it hot or cold - is more plausible.

  46. Per Wikipedia muesli was invented by a Swiss physician in 1900.

  47. Yeah, people used to cook serial, because grinding into flower is a more modern development, and flower or bread were not available to the general population until recently.
    Rolled oats today are a slightly processed food (usually steamed and/or toasted) - which allows us to eat them without further cooking.

  48. Flour (not "flower"!) is not a modern development at all. Among true Mesolithic (transitional-to-Neolithic) findings in the Levant are hand mills (examples from the Sahara), indicating that they ground their cereals into flour even before cultivating them. Hand mills are also a common finding among many Neolithic cultures and some are even Paleolithic (Neanderthal-made Mousterian mill).

    Flour was not just made of cereals but, as mentioned before, also from acorns and then used to make "acorn bread".

    Another forgotten Atlantic traditional food is walnut puree (other nuts are also usable and surely delicious).

  49. "Per Wikipedia muesli was invented by a Swiss physician in 1900."

    Yes, that's what gave me the idea.


    "It was inspired by a similar "strange dish" that he and his wife had been served on a hike in the Swiss Alps."

    A bowl of milk and oats i.e. uncooked porridge, with added seeds, nuts and fruit when available. If the majority of people in the Alps were non-LP they could cook it first - like porridge - but if an individual was LP then they could have it uncooked like a modern breakfast cereal for an extra 200-ish calories a day.

    (And of course the Swiss Alps are the other part of Europe with the heaviest rainfall.)


    Also from the wiki

    "In Switzerland and southern Germany, it is also eaten as a light evening dish; Birchermüesli complet is muesli with butterbrot and milk coffee."

    which is an indication of a meal dominated by milk in various forms.

    A few other correlations

    The Two highest milk producing breeds come from Holstein - Frisia and Switzerland



    and http://www.squidoo.com/holstein

    "The common origin of all the "Black and Whites" appears to be a population of dairy-type cows found along the North Sea coast (from Friesland to Jutland, passing by the Holstein region) which had been more or less crossed with Shorthorns."

    Only circumstantial but interesting.

    Maybe one day there'll be testing of ancient cattle dna.

  50. Lactose intolerant people can't ingest cooked milk (without problems) either: the lactose must be broken and that means fermentation, as in yogurt or cheese.

    Lactose intolerant people would surely eat porridge with water, or as classical Greeks did often for breakfast, with wine. I'm unsure if their porridge was cooked or not, my reference was in Spanish and in Spanish "gachas" just means, it seems, some antique soup of cereal that is not eaten anymore. It's so rare today that for most of my life I could not figure what the concept of "gachas" (porridge) could be at all. Only recently I realized they were similar to cereal breakfasts.

  51. Maju,

    In most of central and northern Europe widespread preparation and eating of bread is indeed quite recent. People did not grind cereal to flour - just into smaller pieces, with the germ part and bran still attached. This then was cooked either in water or milk (the milk version was and partially still is a staple food for young children - eat your porridge!/ iß deinen Brei! still reverberates throughout the region in many languages...). Flat bread also was made from this, mostly without a baking oven.

  52. That's not what I have read. I read (Dhont 1967) that a capitulary (legal text) issued in Frankfurt in 794 (which AFAIK is part of "central and northern Europe") describes the prices of bread depending on what kind of wheat was used: a dinar paid for 12 wheat breads (weighting 2 pounds each), 15 rye ones, 20 barley ones or 25 oat ones. Another text mentions emmer wheat and makes it equivalent to barley in bread price.

    He mentions that the massive amounts of bread sold suggest that the job of baker already existed, profession which is mentioned in the Capitulare de Villis, the capitulary of Pîtres.

    However it's true that oat porridge was mentioned already in the Lex Salica, the oldest Frankish legal text preserved (c. 510).

    He's the same one who mentions that only goats appear to have been primarily/ used for milk production back then, with the main role for cattle being work.

    A problem here is that we don't know exactly what parts of the Empire these Carolingian laws were conceived for but, on one side, the center of the Carolingian Empire was around the Mid-Rhine (Central-West Germany, North France, Belgium) and, on the other, the mentions are too repeated and generic to be a southern peculiarity only.

    Bread was a common staple food in the Carolingian Empire c. 800 BCE at its greatest extension under Charlemagne.

  53. There is no question that mechanical mills and bakers in started showing up in medieval times in central and northern Europe - that does not mean bread was accessible to the masses (why don't they eat cake, then, instead?). The thing is that only certain farmers in some regions were free, and otherwise lords decided what grain was milled to fine flour and what happened to it (probably the court and the army, first) - there wasn't free and ubiquitous access to mills and to baking ovens and at times not even to the necessary firewood. There were few cities with independent bakers and few people being able to afford buying it until rather late medieval times. At any rate, I was mainly thinking about this region before large-scale dominance by kingdoms/ dukes etc. - i.e., before medieval times. Yes, in early medieval times there were water mills, and since late ones wind mills (in this region). But before all of that, hand stone grinding or pounding was not for flower - it was for very coarsely ground/ pounded groats, for all kinds of porridge, and at most, for flat bread made from that on heated stones/ plates (not in ovens). And for the poor subsistence farmers out in the country, it remained like that until early to late modern times, depending on region. Panem et circenses was way south of this - in a better organized society, 1,000 - 1,500 years earlier.

  54. "that does not mean bread was accessible to the masses (why don't they eat cake, then, instead?)"

    Uh? How does the false Maria Antoinette quote (she never said that) fit here?

    The denier or denarius was the equivalent to an English penny, what means that approximately you needed 240 to make a silver pound. With that "small change" you could buy breads by dozens, as mentioned above. So it does not seem that bread was unavailable for the poor. In fact the opposite seems right: that bread was normally the staple food.

    It may not be right for all regions but it was indeed in most of the Caroligian Empire. Nobody says "porridge" as equivalent of basic food, we say "bread". Maybe in East Asia they say "rice" but here we say "bread" - for a reason.

    "The thing is that only certain farmers in some regions were free"...

    That's true. In Germany most peasants were slaves in fact until the Black Death and the Drang nach Östen pushed for better conditions. The situation was better in France but "colonate" (what we now call "serfdom" - but then "servus" still meant slave and "colonus" were just a special quasi-slave status) was also common.

    But that does not mean they did not have access to bread and instead had access to milk, surely a much more desired and hence costly product.

    "But before all of that, hand stone grinding or pounding was not for flower"...

    That's not correct: bread is known since at least Roman times. Querns and animal traction mills were common, I understand. Whatever the case it does not seem to me that milk (considering how costly were most animals, so much desired for meat, traction or wool) would be cheaper than bread. There may be localized exceptions but not the rule in any case.

    If they had their porridge with water or beer, I don't know, but milk does not look so accessible unless you had some goats.

    "Panem et circenses was way south of this - in a better organized society, 1,000 - 1,500 years earlier".

    Well, the Roman Empire reached into the lands of modern Scotland, Netherlands, Germany and Hungary, and included all Belgium, Switzerland, England, Wales and most of Austria. Trier was a Western Roman capital for a time and the capitality of Rome ended with Constantine (Ravenna, near Venice, was chosen instead as capital of the West). Let's not exaggerate nor fall into false notions.

    Certainly York, London, Tier, Cologne or Mainz, among other Central-North European cities, had their 'panem and circenses' thing. And Roman influence extended through much further into the North of the continent (at least to Denmark). While the Goths may certainly have been quite detached from Roman customs initially, the Franks were for everything that matters border Romans instead (and that may have been the main reason of their long term success, as they were not perceived as "foreigner" or at least not so much).

    I think that you are largely just imagining things. You have not documented your claims in any case, so I have reasons to doubt them. Whatever the case, the popularity (or not) of porridge should not be automatically linked with availability or massive consumption of milk. That depends only of milk availability, because porridge can be eaten with water, mead or whatever else and milk can be accompanied with other foods (or drank alone).

  55. Maju
    "Lactose intolerant people can't ingest cooked milk (without problems) either: the lactose must be broken and that means fermentation, as in yogurt or cheese."

    Ah, that makes a difference.

    "But that does not mean they did not have access to bread and instead had access to milk, surely a much more desired and hence costly product."

    For my part i'm mostly thinking the milk thing, if true, is very pre-medieval when arable farming in northern europe was in its infancy, starting in the most suitable regions for milk production - like the Atlantic coast and Switzerland - and probably lasting the longest (as a ubiquitous food source) in the same regions while in other regions the diet moved on as arable farming caught up with pastoralism.

    (Although as you mention in an earlier post meat must have been the biggest component in those same regions because originally they wouldn't all have been LP.)

  56. "because originally they wouldn't all have been LP"...

    On the contrary, my best hunch is that LP was widespread, probably fixated, among pre-neolithic peoples but for no adaptive reason at all: it was just a fluke - like Rh- or R1b were more or less fixated, so was the LP allele. Otherwise it's very hard to explain its prominence, specially as Neolithic farmers (and cattle herders, mind you) appear to have been all lactose intolerant in fact.


Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).