Myths of British ancestry revisited

Stephen Oppenheimer responds to readers' questions and comments on his October 2006 article on British ancestry
June 29, 2007
Discuss this article at First Drafts, Prospect's blog

Stephen Oppenheimer's article "Myths of British ancestry" in the October 2006 issue of Prospect attracted a huge online readership, and continues to generate comments and responses. Here, Oppenheimer replies to a number of them.

Q—Stephen Oppenheimer's fascinating thesis helps to answer one of the most vexing questions of dark-age British history: why is there so little trace of Celtic culture in England and in the English language? The fact that so little remains of Celtic influence in England in terms of place names—outside Cornwall and Cumbria—and in the language points to a long process of cultural conquest by the 4th and 3rd centuries BC Belgic invaders, who were Germanic, as implied by Julius Caesar's history of his British adventures. The cultural and linguistic origins of the English are thus pre-Roman. The Anglo-Saxon elite invasions of the 5th and 6th centuries AD reinforced, rather than created, a pre-existing difference between the proto-English and the culturally Celtic of the western fringes of the British Isles.

Mark Hudson

A—This letter draws attention to an aspect of the evidence that I understate in my book, namely the near-absence of Celtic influence in modern English place names. Whereas there are a couple of examples of near-complete "language shift" with absence of borrowing from a previous aboriginal vocabulary, indigenous place names are in general more resistant to extinction. This can be seen in America and Australia, which retain a considerable number of indigenous place names. These two examples are interesting, not only because massive replacement and genocide took place, but also because Australian and American English retain far more aboriginal vocabulary than native English retains Celtic. England itself retains pre-Indo-European place and river names, but few Celtic names, and the English language has literally only a handful of Celtic words.

The fact that England is such a "Celtic desert" is a problem for linguists who believe that Anglo-Saxon triumphed in what had been a totally Celtic-speaking region, even given the gory stories of massacre. This problem is because the Angles and Saxons apparently carried out a much better job of language extinction than in Australia and America, where genocide and massive replacement are so well documented. The "overkill" problem is acknowledged by English place name authority Richard Coates in a recent article "Invisible Britons: the view from linguistics," where he concludes either that the genocide was complete or that there were few Britons actually living in England to interact with the invaders: "I argue that there is no reason to believe large-scale survival of an indigenous population could so radically fail to leave linguistic traces."

Rather than pause to question scholarly assumptions that England had been 100 per cent Celtic-speaking until the 5th-century invasions, Coates prefers to use the linguistic evidence to challenge the genetic evidence: "These are the questions that need to be answered by those who propose a massive contribution of Britons to the "English" gene-pool."

I guess I would see it the other way around. While there is no reason to expect that language change, resulting from invasion, should necessarily be massively reflected in the genetic picture, there is every expectation that complete genocide predicted by linguists should be—if it really happened.

Stephen Oppenheimer

Q—Oppenheimer's article shows the futility of letting scientists loose on purely historical questions, which are better tackled by historians, archaeologists and linguists. There is no essential connection between where your ancestors came from in the Neolithic period and what language you speak or how you behave culturally. In any case, statistically all of us are descended from everyone: allowing 25 years per generation, in the 62 generations since 450AD, we have had 4.6 x 1016 direct ancestors, more people than have ever existed, and so we must be related to everyone on earth many times over.

Martin Nichols

A—From your first sentence it seems you must long for the good old days when historians, archaeologists and linguists could speculate on European invasions by Indo-Aryans, Kurgan horsemen and Celts, free of troublesome biological evidence. If you read my article and book, you will realise that your second sentence contains my starting point or null hypothesis: that connections between culture and genes are likely to be tenuous and that individual cases where this is claimed have to be tested appropriately.


Q—It is true that, "The genetic evidence shows that three quarters of our ancestors came to this corner of Europe as hunter-gatherers, between 15,000 and 7,500 years ago." This is the R lineage group and most European males have an R Y chromosome. But it is rather silly to say that, "Our ancestors were Basques, not Celts. The Celts were not wiped out by the Anglo-Saxons; in fact, neither had much impact on the genetic stock of these islands." Angles, Saxons, Celts and Basques are not lineage groups. They are ethnic groups that developed within the last 2,000 or 3,000 years. Like most Europeans, they probably belonged to the R lineage. Most Germans, Poles, French, Spaniards and Russians also belong to the R lineage group. None of this negates the established history of the British Isles.

Has Oppenheimer read the research of Weale et al—"Y Chromosome Evidence for Anglo-Saxon Mass Migration" (2002)—which shows that the male populations in two central English towns were genetically very similar, whereas those of two north Welsh towns differed significantly both from each other and from the English towns? Using novel population genetic models that incorporate both mass migration and continuous gene flow, they concluded that this was best explained by a substantial migration of Anglo-Saxon Y chromosomes into central England—but not into north Wales.

Douglas Forbes

A—I cannot claim responsibility for your second quotation, which is from the article's standfirst. As you must realise, authors of magazine articles rarely have control over these. I cannot disagree with your complaint, but hopefully you read the whole article.

On your second point, it is misleading for you to talk about frequencies of the R male lineage in different European countries as if this constituted a uniform genetic background, since there are actually two main R groups, which split tens of thousands of years ago outside Europe and had completely different modes of spread and present distributions in Europe. R1b expanded from the Basque Ice Age refuge and predominates in extreme western Europe, being found at only 20 per cent or less in Russia and the Baltic states. R1a1, on the other hand, predominates in eastern Europe, and to a lesser extent in Scandinavia. I deal with the spread of both major R lineages at length in chapters 3 and 4 of my book The Origins of the British.

I have indeed read the research of Weale et al. I discuss it and similar papers at length in chapter 11 of my book, where I register my disagreement with their method of reconstruction from relative gene group frequencies, presenting instead my own phylo-geographic re-analysis of their data, based on fine detail of individual founding lineages.


Q—Interestingly, Robert Graves, in his book The White Goddess, developed a theory about early settlement of these islands similar to Stephen Oppenheimer's. Graves's evidence is based on early literary sources, mythology, local tradition and the archaeology known at the time of writing. I gather that Graves is not popular among archaeologists. But if you are prepared to tease out strands of DNA from human body fluids, looking through The White Goddess should be no greater challenge.

Christine Peace

A—Thanks for this information. I have read several of Graves's books, but not The White Goddess. I shall rectify. Incidentally, another European scholar, linguist Theo Venneman, has a reconstruction of post-Ice Age recolonisation of the British Isles, which gives a relative of the Basque language primacy of place as the first entrant. I outline his theory in the new paperback edition of my book The Origins of the British.


Q—Regarding your statement that 75-95 per cent of paternal genes in Britain are of Iberian origin, is this genetic material distinct and specific only among Basque-type peoples, or does some of it share features with other, non-Basque Europeans? If the latter is true, why is it omitted from your findings?

Timothy Burton

A—I do discuss the questions you raise, but in chapters 3 and 4 of my book The Origins of the British, not in the more condensed Prospect article. Part of the answer to your query is in my answer to Douglas Forbes above, but allow me to expand a little more here.

As you suggest, the re-expansion of paternal group R1b and maternal group H from the Basque Ice Age refuge spread up the coasts of all the countries facing the Atlantic, after the ice melted. The British Isles retained higher rates than the other countries, for several reasons related specifically to early movements directly from the Basque country rather than from general diffusion from western Europe. First, as a result of lower sea levels, the British Isles, in particular Ireland, were connected and at the furthest edge of the extended Ice Age European continent, and thus received the bulk of early coastal migration. Then, as sea levels rose, first Ireland then Britain became islands, relatively insulated from further migration from elsewhere in Europe, thus preserving their high rates of R1b and similarity to the initial settlements.

The means by which I could separate the R1b types in the British Isles from those on the other side of the channel is by the use of "Founder Analysis." That is, looking at the detail of their gene types (so-called STR haplotypes). These revealed 21 founding clusters, which could only have arrived direct from the Basque country. Their descendant twigs are unique to the British Isles. Furthermore I was able to date the arrival of these individual clusters using their diversity.


Q—What about the genetic make-up of the Man Islanders? Did it suffer few modifications from its origins because of? geographical remoteness, or is it very different from the rest of the British Isles because of the impact of invasions (such as the Vikings) on a small population? Alexandre Cogan

A—The simple answer is that your first suggestion is closer to the truth than your second. The Isle of Man received more Norwegian gene-flow than anywhere else in the British Isles, except for Shetland and Orkney, which received the most. This does not, however, account for more than 20-25% of the male Isle of Man gene pool. Fig 11.4b in my book gives a very approximate genetic distance map, illustrating this in more detail.

Stephen Oppenheimer

Discuss this article at First Drafts, Prospect's blog

You can read more about Stephen Oppenheimer's work here, at the Bradshaw Foundation website. And you can buy his new book at our online bookshophere