Britain: a marvelous mongrel mix of migrants

March 20, 2015 • 10:35 am

JAC: I asked Matthew—actually, I twisted his arm—to tell us a bit about the new paper in Nature dissecting the genetic composition of inhabitants of the UK, which showed lovely genetic clusters reflecting history and ancestry. Of course, as the “race critics” would assure us, those clusters are only social constructs!

by Matthew Cobb

One of the key issues in British politics for some time has been the level of immigration. Over the last decade, hundreds of thousands of people have come to the UK from Europe, which some people find problematic for cultural and economic reasons. (Even more Brits have gone off to live in Spain and elsewhere in Europe, but that fact never gets mentioned, any more than the fact that those immigrants to the UK appear to be driving the economic recovery…)

A new study in Nature this week [reference below] shows that, in genetic terms at least, all this is nothing new. The land mass that is Great Britain (England, Scotland and Wales – the big island) has been subject to repeated waves of immigration over the last few thousand years. That much we knew from the history books. What we didn’t know is what happened when those immigrants – or invaders – hoved up on our shores. Did they settle, breed with the locals and contribute to today’s British genomes? Or did they simply sit in their castles (or big huts) and exploit the poor natives?

The study is part of a long-term investigation of the genetic structure of the UK population (the United Kingdom is a political entity that comprises Great Britain and Northern Ireland), called People of the British Isles (PoBI), based in Oxford and set up by Sir Walter Bodmer.

As Steve Jones has pointed out, the invention of the bicycle did a great deal to shuffle the genes in the UK, as people started moving around the country. To get over this problem, the PoBI project studies only people whose grandparents were all born within 80 km of each other. They were therefore effectively able to study the DNA of those grandparents, most of whom were born in the late 19th century. They recruited 2039 people who met these stringent criteria and then studied the variability in their DNA. They also looked at the DNA from 6,209 people from continental Europe to find matches.

The big picture results are quite astonishing. Firstly, they found groupings of people that basically fit with the key geographical areas of the UK, suggesting that prior to the end of the 19th century, many people did not move around very much. This is true even down to very small regions – look at Devon and Cornwall (bottom left hand corner) – there is a very clear division between the two areas, even though they are right next to each other. Similarly, people from north Wales and from the England/Wales border seem quite distinct. (Note that this is a very fine-grain analysis of small differences – overall, everyone was pretty similar, as you would expect.)

It’s important to note that these different genetic groups were identified without knowing where people came from. When they plotted the data onto the map, it fell out in this dramatic way. Isn’t that cool? The researchers must have been so excited!

So where do these genes come from? They compared the 17 groups they identified in the UK with the data from continental Europe, and you can see that that – despite what some in the UK might like to think, Britain is indeed part of Europe:

European ancestry profiles for the 17 UK clusters.

The height of the bars shows the proportion of each of the 17 UK groups that can be traced to that particular European group. So, for example, you can see folk from North Wales have a very high proportion of their genes from the light blue French group FRA14, but none at all from the dark blue FRA17 type. [JAC: note that the inhabitants of the northern islands, like the Orkneys, have a substantial genetic component from Scandinavia.]

All this can be explained in terms of the successive waves of migration/invasion over the last 10,000 years, as the authors show in this handy cut-out-and-keep summary of British history:

Major events in the peopling of the British Isles.

The data suggest that there was no ‘Celtic’ population in Britain before the Romans arrived – the various groups were already differentiated according to different origins before we were told to stop painting ourselves blue and to start wearing togas.

As the bottom right hand map shows, Britain experienced two waves of Viking invasions, the first from Norwegian Vikings, the second from Danish Vikings, who also ran a protection racket for quite some time, forcing the locals to pay up or be ravaged (this was called the Danegeld). Strikingly, there is no evidence of this long period of Danish occupation in the DNA sampled in this study – although a lot of the people sampled had Danish DNA, it is not localised to any region (you can see this by the relatively even spread of DAN14 across all 17 types, in the second figure).

In other words, the Danish Vikings kept themselves to themselves and did not intermingle with the locals. When they eventually left, they did not leave any DNA behind them.

If you want to know more about this study, I heartily recommend listening to Adam Rutherford’s excellent Inside Science programme* about it (this is available anywhere in the world), in which he interviewed the man who did the statistics, Peter Donnelly [JAC: Peter Donnelly used to be in my department at Chicago, as well as in Statistics here]. There is also this nice summary from Nature news.

* I shamelessly stole the title of this post from Adam’s celebration of the UK’s mongrel nature.


Leslie, S et al (2015) The fine-scale genetic structure of the British population. Nature 519: 309–314. Here (£££).


57 thoughts on “Britain: a marvelous mongrel mix of migrants

  1. In other words, the Danish Vikings kept themselves to themselves and did not intermingle with the locals. When they eventually left, they did not leave any DNA behind them.

    I think it was Jared Diamond in Guns, Germs and Steel who made the point that whenever one people ran into another people, there may have been trade or there may have been war — but there was always fucking. This data seems to contradict that.

    An alternative explanation though might be that the Danish Vikings raped and canoodled as usual, but that they were viewed so negatively by the population that most pregnancies and/or resultant offspring were swiftly terminated and/or dispatched. I’m not sure how to test that at this late date, though.

    1. Matthew’s comment about the Danes…just eyeballing the height of the bars for various groups, it looks to me like the Danes contributed more than any specific set of Norwegians or Swedes (though maybe not all those groups collectively). Their bars are comparable or larger than those of Ger 3, Fra 12, SFS 31, and maybe Bel 11. In fact of the 14 European groups listed, I’d estimate just by eyeballing ‘area under the columns’ that they’re the 4th or 5th most common. Am I reading the chart wrong?

      1. That’s what I thought too. Also, that part of southern Sweden that the genetic marker comes from was ruled by Denmark at the time of the Viking invasions.

      2. The point is rather that the genes shared with the Danish group are fairly uniformly distributed across the British Isles rather than localised to the Danelaw. That implies they’re mostly due to older common ancestry rather than admixture from the Vikings.

        1. The great Old English epic, Beowulf, is set in southern Sweden and Denmark (an odd thing, surely, for an epic supposed to be echt English), and it has been proposed that the ancestors of Raedwald, the king who is buried at Sutton Hoo, the Wuffingas, were originally Swedish; and also that English – the language – is in fact more closely related to Scandinavian languages than to West Germanic languages.

        2. Either that, or there was some truth to that old Dave Allen sketch (punchline “You play your games, and I’ll play mine”). Oh, those Danes!

          (just spent hours looking for the sketch on YT, no luck)

    2. With all respect, Mr. Cobb has this part wrong. His otherwise fine posting:

      “In other words, the Danish Vikings kept themselves to themselves and did not intermingle with the locals. When they eventually left, they did not leave any DNA behind them.”

      The conclusion the authors of the study draw is rather that the Danes (as well as the Romans before them and the Normans after them) had little genetic impact because they were small in number compared to the local, subject, population. They bred, but there just weren’t enough of them to make a dent over time. The cultural impact of small groups of conquerors can be much greater then their genetic impact.

      I’m surprised that Cobb missed the other interesting finding – that most of the French ancestry dates from a previously unknown migration and that the Normans had little impact on the genetic level.

      1. You’re right that ‘did not leave any DNA behind them’ is far too strong and therefore wrong. But I think the conclusion of the authors’ is stronger than you make out. They write:

        “we see no clear genetic evidence of the Danish Viking occupation and control of a large part of England, either in separate UK clusters in that region, or in estimated ancestry profiles, suggesting a relatively limited input of DNA from the Danish Vikings and subsequent mixing with nearby regions”

        If there was extensive interbreeding between the Danes and the locals, you would have seen that trace in those regions that they controlled. The key point (discussed above) is therefore that the contribution the Danish DNA makes does not distinguish any of the regions – it’s fairly constant across populations. This suggests an earlier (or later) admixture, and not a trace of that period of domination.

        A final point that needs reiterating is that, overall, these populations are highly homogenous. That is, they do not differ by much at all – however, they carry markers that enable us to trace their history.

        As to the Normans (who were simply Vikings, as their name indicates, who had been living a while in north-western France), this may be a product of the sampling – we know they left traces in the upper echelons of society (e.g. the aristocracy), because they basically took it over. Those traces may not have been picked up in what is I think assumed to be a more peasant-based sample, given the requirement that all 4 grandparents had to have been born within 80 km of each other.

  2. Bloody Vikings coming over here, taking our jobs, teaching us how to make bread and boats and work metal, no respect for our traditions. What have the Romans ever done for us?

      1. Mitchell & Webb sketches are so moreish (that’s my first time using that word in text), I’m going to have to spend the next hour watching clips.

  3. “We are each a multitude.”

    It would be very interesting seeing these sorts of analyses for other countries in Europe where there is also (wrongly) held to be a relatively homogeneous population.

  4. For me, the question becomes: Do those tight little isles need more humans? Will another few million humans make life better for the inhabitants of the British Isles? Is compression of space-per-indivisual and ever-increasing diminution of resource allotment a good thing for the citizens of the UK?
    Just askin’.

    1. Read Garret Hardin – “Lifeboat Earth” (just google his name and there is a foundation which has all his most interesting papers as PDFs including “Tragedy of the Commons”. You must have noted that seldom does anyone write (or talk) about excess population. Hardin took a hard look two generations ago, and academics and many do-gooders were aghast. He was right, of course. And here we are still failing to recognize it, and the politicians and religions are in complete denial.

  5. No wonder those commercial genetic tests don’t tell you all that much you don’t already know (since they don’t go into the detail of this one). I figure it would say “British Isles” as my origin since that’s probably where most of the genetics come from and would they really distinguish between all those Germans and Frenchmen that snuck (that’s right, I said “snuck” & not “sneaked”)into my gene pool?

  6. Of course, as the “race critics” would assure us, those clusters are only social constructs!

    The first step in arguing with a position is to understand that position, not a strawman version. Nobody says that clusters of genotypes are social constructs. “Race” is the social construct, and biological races (subspecies), in those species in which they exist, are not merely geographically structured variation. You can have the latter without having the former, and that’s the “race critic” position.

    The obvious question here would be “If there are races, what are they?”. And nobody has been able to agree on that, which is suggestive.

    1. Please stop saying I have a strawman position. Races are genetically differentiated populations, usually geographically isolated–as these once were. I never said that there were definitive numbers of races, or that they could be demarcated with certainty, but genetically different populations of humans exist, and you can place an individual within one group or another. Were “races” social constructs in animals?

      Anyway, I do not want to consider this discussion. Your question at the end is like saying “if there’s a difference between a child and an adult, which one does a 17-year-old belong to.

      1. At the risk of opening up a can of worms, wouldn’t the discussion be less fraught if it were about ‘populations’ instead of ‘races’? Oh never mind, that train has long since left the station.

          1. You are moderated until you learn to comment civilly, without going after the host, and until you stop acting like a sourpuss. Somehow you have not learned how to comment without being aggressive and abrasive. Do not bother to argue with this, nor email the proprietor. If you do, you’ll be gone for good.

            1. Just for the record, I am fully in agreement with you. I however, wound up in anthropology and then was hired by a sociology department. Neither my discipline (if you can call it that) nor the sociologists believe in races Nor does the United States Government, and worst of all neither has the American Association for the Advancement of Science which annually sent me a renewal form with “race” to be checked. My letters to the president of the AAAS for the embarrassingly unscientific “races” they offered always went unanswered.

        1. I think that instead of “populations,” which is a lumping term, “groups of individuals” may be better.

  7. All those arguments in the new world about what country are your “people” from and how about you. Kind of a joke really.

    My mother was a McFarland and I’m a Schenck. Probably more Italian than anything?

  8. I was wondering if Hadrian’s wall, which is shown in one of the smaller maps, still might have an historical effect in gene distribution today. The 1st map implies a discontinuity in genes near that region, but the lower maps suggest it was pretty much erased from later invasions from Europe. So I am not sure.

    1. I don’t think that wall was around long enough to be a significant enough barrier to cause genetic differences in populations on either side. Also, it probably wasn’t all that effective in keeping people out (the thing was long and hard to man) so it may have been a deterrent but not a real big one.

    2. The wall as a barrier was short lived. Even within its active lifetime it wasnt necessarily the border (for a period it was replaced by the Antonine Wall).
      It is also unclear how much of a barrier it was. The best guess seems to be it was more a customs barrier and wouldnt have prevented the people from either side mingling, so long as they didnt turn up enmass carrying sharp pointy things.
      Then once the Romans did a runner it became invisible. The Kingdom of Northumbria ran from York up to Edinburgh for several hundred years after that.

      1. I would have expected more North African ( for example ) genes showing through that traditional Italian. Apart from local “liaisons” ( which of course must have happened – soldiers being much the same throughout eternity) the majority of Roman legionaries were Roman citizens ( until 212 anyway when all free born men became citizens) and probably would have returned to Italy following military service. The same would not have applied to the non-citizen auxiliaries who would have probably settled in Britannia following their retirement . These guys were recruited from all over the Empire and certainly there is evidence of soldiers from the Middle East serving in Britain

        Another point ( which might be pertinent to the Viking situation) is that there could have been remarkably few Romans actually in Britain. A ruling elite, three legions ( post about 100 AD) and the auxiliaries. A few 10’s of thousands maybe? The situation was remarkably similar to the British in India where there was a tiny British presence – both bureaucratic and military,

        1. If you first learned Latin using the British version of the Cambridge Latin course, your introduction to the “Roman Britain” unit has two soldiers talking to each other in Britain complaining that they want to get away from “this island” because the “sun never shines.” 🙂

      2. Rather than a barrier, it’s probably more important that the Wall was a chain of garrison towns. That could be expected to leave a genetic mark.

  9. Going on viking is all business and no fun!? [/disappointed viking prospect]

    Of course, as the “race critics” would assure us, those clusters are only social constructs!

    Is that what they would say? I thought they would say that the variation between demes is smaller than the variation within demes.

    Something like: “this is a very fine-grain analysis of small differences – overall, everyone was pretty similar, as you would expect.”

    But I probably do not understand what race is supposed to mean.

    1. And first after posting, I noticed that Jerry had made a very clear definition of “race” above.

      So race can be something that only a genetic test can see, worst case?

      1. I think the definition of race is a physiological difference, or set of differences, (e.g. skin, facial structure, hair, stature, body proportions) which humans generally recognize. (We organized ourselves that way for a very long time; and some places still do — ask any Ainu.)

        These are all phenotypes of the genes these groups carry.

        I think this is what Jerry is saying; but I may well be wrong about that.

  10. The Danish Vikings didn’t leave. It’s a matter of historical record that they settled Eastern England and were eventually absorbed into the unified English kingdom established by the descendants of Alfred the Great. That’s why northern and eastern England are littered with place-names of Norse origin, e.g. all the towns and villages ending in “-by”, “-thorpe” or “-thwaite” .

    One obvious reason why there appears to be no distinct “Viking” genetic signature in eastern England is that they were scarcely different from the Anglo-Saxon population who were there when they arrived – people who were themselves descended from earlier migrants originating in northern Germany and Denmark.

    And I think it’s worth noting that despite the “mongrel nation of immigrants” spin that the post tries to put on this study, the data show that there’s been no numerically significant genetic input into the population of the British Isles for well over 1000 years, basically until the large-scale arrival of Afro-Caribbean and Asian people beginning in the 1950s.

    1. I am curious about “-thwaite”. It doesn’t ring a bell, and I thought it was old english.

      Do you know what it means?

      [Roughly, “by” is village, and “torp” is house or place.]

      1. It is a “woodland clearing”. According to the Jorvik Centre:

        “-thwaite: originally thveit, woodland clearing. Example Slaithwaite (Huddersfield)”

        It is Viking ….

    2. That has already been commented upon in the Nature news article when someone said: “Danish Vikings, who occupied Britain between the 700s and 1100s ad, by contrast, left little signature in most Britons’ genomes” How can they tell ? ….A lot of the “Danish” vikings would in fact have been descendants of the Jutes and Angles(who lived in what is now Jutland and Schleswig)meaning they would have had the same genetic make-up as the peoples that migrated to Britain 300 years earlier. Previous studies have been virtually unable to distinguish between the Jutish tribes(cimbri, teutons, harudes,vendel and jutes) and the Angles, with which they also shared the north germanic language. (as opposed to the west germanic spoken by the Saxons.) Recent genetic surveys in Denmark have also shown that modern “jutes” ie people from jutland, are genetically much closer to the britsh than people from the danish isles , Zealand in particular, that also being the area of the original tribe of the Danes.”

      ( I take no credit for that. )

  11. “Strikingly, there is no evidence of this long period of Danish occupation in the DNA sampled in this study…”

    I think that this can best be explained by understanding the true meaning of dane-geld, whereby a norse is rendered unable to breed.

  12. The no Celtic invasion data backs up what a lot of archaeologists have been saying for a while that while there was similar cultural aspects between the “celtic” nations it was probably spread via trade rather than by invasion or people themselves.

  13. I do hope the Big-3 of DNA/ancestry companies in the US take note of this study. I have complained of their British ancestral biases to at least two of these companies. The way I see it there has long been a pro-British bias in United States genealogical or family history circles, and that bias has been continued in determining what your DNA has to say about ancestry. That has created a British bias which discounts northwestern European origins. I happen to be a quarter Belgian and a quarter northwest German (Saxon/Friesian). However, AncestryDNA has me listed as 45% British and even 1% Irish when I have no ancestry in either area for the average 9 generations back I have traced my ancestry in Belgium, France, and Germany. Strange that in coming to America my immigrant ancestors from continental Europe created an individual who looks to be about half British.

Leave a Reply