« Love thyself | Gene Expression Front Page | QUOTE OF THE WEEK »
June 13, 2003


I have at last got my hands on C. Capelli et al.: A Y Chromosome Census of the British Isles, Current Biology, vol. 13, 979-984, 27 May 2003.

Capelli et al. took DNA samples from men in 25 small towns around the British Isles, excluding men whose paternal grandfathers were born more than 20 miles away. For comparison they also took samples from Norway, Denmark, North Germany (Schleswig-Holstein), Friesland (Netherlands), and the Basque region of Spain. Using comparison of Y chromosome haplotypes, the Danish, North German and Frisian samples are all closely similar to each other, but the Norwegian sample is significantly different from these, and the Basque sample is widely different.

In a Principal Components analysis the Irish and Welsh samples (with one exception) cluster together with the Basque sample, supporting earlier findings. As the Basques speak a pre-Indo-European language, this suggests that the Irish and Welsh (so-called ‘Celts’) have a largely pre-Celtic genetic ancestry, possibly going back to the Palaeolithic. In Britain, the Orkneys, Shetlands, Western Isles, Isle of Man, and Cumbria (Penrith) show a clear Norwegian input, as expected. Elsewhere in mainland Britain there is no obvious Norwegian input, but varying degrees of German/Danish ancestry. Scottish mainland sites are intermediate between English sites and the ‘indigenous’ (Welsh/Irish) ones. However, all the English and Scottish sites show some ‘indigenous’ ancestry. The German/Danish component is strongest in eastern England and weakest in England south of the Thames.

Most of this is unsurprising, but there are two more controversial conclusions.
One is the claim that ‘the results seem to suggest that in England the Danes had a greater demographic impact than the Anglo-Saxons’. This is based on the finding that the German/Danish element is strongest in areas like Yorkshire that are known to have been settled by Danes. The conclusion seems to me a non-sequitur. The areas settled by Danes were the areas most exposed to invasion from Denmark and North Germany, and they got a double dose of German/Danish genes: first from the Angles, then the Danes. It would be very surprising if they did not have the strongest German/Danish element.

The other controversial conclusion is that the German/Danish element in southern England (south of the Thames) is limited, and that the male ancestry of this area ‘appears to be predominantly indigenous’. This may be true, but I would want to see it replicated with different samples and methods before taking it as firmly established. It should perhaps be noted that the samples with the smallest German/Danish element all come from areas (Wessex, Sussex, and Kent) reputedly settled by Saxons and Jutes, while the samples with larger German/Danish elements come from areas settled by Angles (East Anglia, Mercia and Northumbria). Conceivably there was already a genetic difference between these three ethnic groups before migration, though this does not seem particularly likely, as they all came from much the same area of Northern Europe.

As Capelli et al. recognise, their results seem to conflict with those of Weale et al., ‘Y Chromosome Evidence for Anglo-Saxon Mass Migration’, Molecular Biology and Evolution, 2002, vol. 19, pp.1008-21, which found a sharp distinction between central English and Welsh populations, but no significant difference between the English population and a Frisian sample. This discrepancy needs to be reconciled.

As I am a historian and not a geneticist it may help if I outline the historical evidence on the ethnic origins of the English. There is no dispute that British Celtic) elements were predominant in Cornwall and Cumbria, where Celtic languages survived long after the Anglo-Saxon invasions. There is also good evidence of British elements surviving in Kent and Wessex (see esp. Myres, ‘The English Settlements’, pp.147-73). But beyond that, there has been controversy since Victorian times. At one extreme, which I call the ‘Wipeout’ theory, it is believed that Celts were virtually exterminated or expelled by the invading Anglo-Saxons. At the other extreme, which I call the ‘Upper Crust’ theory, the Anglo-Saxons took over as a ruling elite but left the peasants largely untouched (rather like the later Norman Conquest). And of course there are intermediate positions.

The main lines of evidence are as follows:

Written sources: the main sources - Gildas, Bede, and the Anglo- Saxon Chronicle - make it clear that invaders from the Continent took political control of what is now England, and that in many places there was violent conflict between the invaders and native forces. But there are no reliable written sources on the numbers and proportions of different groups.

Language: the Old English or Anglo-Saxon language, in its various forms, is purely Germanic in its grammar and vocabulary, with no discernible Celtic element. If the Celts learned English, they learned it very thoroughly. The later Danish settlements strongly influenced the form of Old English spoken in eastern England, but did not replace it.

Place-names: the names of major towns and rivers often show some derivation from Celtic or Romano-British names, but the names of rural settlements are overwhelmingly Germanic (Anglo-Saxon or Danish), except in western England, where there is a ‘cline’ of increasing Celtic influence. However, there have been controversial claims that some Anglo-Saxon names have disguised Celtic origins.

Continental evidence: before the Anglo-Saxon settlement of England there were people known as Angles in northern Germany, and after it there weren’t. Around the same time, the Armorican peninsula was settled by Celtic Britons, to the extent that the area became known as Britain (Bretagne or Brittany). This certainly looks like a mass displacement of populations.

Religion: late Roman Christianity and Celtic religions disappeared from England and were replaced by Anglo-Saxon paganism until Christian missionaries from Ireland and Rome arrived.

Archaeology: there are few recognisable remains of any kind from the 5th century. After that, archaeological remains are mainly Germanic in style. It was formerly assumed by archaeologists that a change in style of this kind involved a migration of people, but the recent tendency has been to assume that styles change by ‘cultural diffusion’ or elite influence. Sometimes archaeologists seem to forget that ‘no conclusive proof that A’ is not the same
as ‘conclusive proof that not-A’.

Social structure and customs: the evidence from Anglo-Saxon poetry, laws, etc., is of a Germanic/Scandinavian society and customs. However, some sources do refer to ‘wealh’ (Welsh) inhabitants, who are presumed to be surviving Britons. The laws of Ine, king of Wessex in the late 7th century, make it clear that ‘wealh’ people could be either free or slaves (theow), and that they could belong to ‘wealh’ kinship groups, which implies survival of more than isolated individuals. Also, some charters and other documents refer to substantial numbers of slaves. (It complicates matters that the word ‘wealh’ itself, which originally meant ‘foreigner’ or ‘stranger’, may sometimes be used to mean ‘slave’, implying status rather than necessarily ethnic origin.)

The positive evidence, so far as it goes, seems to me consistent with something closer to the ‘Wipeout’ theory than the ‘Upper Crust’ theory, though with survival of ‘wealh’ populations in varying proportions. The advocates of the ‘Upper Crust’ theory rely heavily on an ‘argument from impossibility’: it is impossible, they say, that a relatively small number of Anglo-Saxon invaders can have wiped out a much larger Romano-British population. However, I think this is a misunderstanding of the invasion scenario. Roman-British society rapidly broke down when the Romans left. Even without invasion there would have been a population crash. The Romano-British were virtually defenceless apart from mercenaries who were themselves mainly Germans (Saxons), and quick to invite their relatives over to share the spoils. To destroy a defenceless population, it is not necessary to kill them individually. Just take a few captives in the first village you come to, skin some of them alive in the market-place, and let the rest of them go to spread the news. A wave of panicking refugees will spread out in all directions, and starvation and disease will do the rest. For analogy, suppose you heard that Martians with invincible weapons and sadistic habits had landed twenty miles away. You would run like buggery!

However, the feasibility of a scenario does not mean it is true. Further genetic evidence may finally resolve the controversy. If it is in fact proved that the ‘Celtic’ element was predominant in southern England, this would have interesting implications for cultural history and evolution, for it would show that a complete change of language and culture can be imposed by a dominant minority, in an illiterate pre-industrial society, and in a short period of time.


Posted by David B at 03:18 AM

Actually there isn't a great deal of conflict between the two papers. First of all, the Weale paper concentrated on Wales and Central England. Capelli and his colleagues covered a larger area of the island of Britain. Both papers show that in the island of Britain the highest levels of Celtic Y-chromosomes are found in Wales while the highest levels of Germanic Y-chromosomes are found in Central England. Capelli's paper found that there was not a great deal of difference between the Frisians, Danes and northern Germans, but that the Frisians were a little closer to the various British populations than the other two Germanic populations were. What Capelli seems to be saying is that the Anglo-Saxon invasion of Britain was quite small while the Danish invasion that led to the establishment of the Danelaw was much larger.

Posted by: Diarmid Logan at June 18, 2003 12:03 PM

Any read of Micheal Wood's book "In search of England" shows that Southern England suffered huge population expulsions in the early centuries after the Normann invasion.

Then there was the ethnic cleasing of "transportation" to Australia, borne most heavily by the southern counties. Then there was those that fled persecution to "Amercia", failing to recognise that religion and ethnicitity go often ahand in hand, from Southern England in the main. The issue that remains to be discuessed is which kind of English did the leaving and which kind of English did the staying.

Then there were the huge human expulsions caused by the "Enclosure Acts. Then there is the effect of motor transport and communitng which hit Southern England after the First World war long before other areas and continues unabated to this day. The substratrumn of the "affordable housing" issue is about retention of the orignal ethnics. Then we should talk about the human dispersals on the continent. The research is weak in its drawn pool.

The jury is out on the Celt thing re Southern England (excluding Cornwall) The reasrch has been snapped up and trumpeted. It is at least amusing to some to postulate that the Anglo-Saxons do not really exixt.

Having trawelled the web it is interesting to note what parties grabbed the story, The NY Post,and the Guardian are the ones that persistanetly come up. Neither can be said to have a majority Anglo-Saxon readership.

We appreciate your constructive analysis of the issue. It is the first we have discovered on a review of 1,300 site on the web on a Anglo-saxons search. We were not looking per se for an anaylsis but wrote because finding it we respected the moderate tone of the views experessed.

Posted by: The editor at July 2, 2003 08:44 AM

To "The editor":

It wasn't just the Guardian and the NY Post that picked up on the story. Check out the links below:




Posted by: Diarmid Logan at July 30, 2003 10:28 AM

Here are links to the two research papers concerning Germanic and Celtic DNA in Britain:



Posted by: Diarmid Logan at August 14, 2003 10:57 AM

Very good paper. I wish the scientists would do a much more widespread study, taking more DNA-samples from even more areas in Britain and also including samples from Sweden, the putative Urheimat of the Germanic version of the Nordic Race. Have any studies tried to make distinctions between the different subtypes of the Nordic Race?

Posted by: Kenneth Stephen Doig at September 21, 2003 08:57 PM

For information on Scandinavian genetics, check out the research papers below:



Posted by: Diarmid Logan at October 10, 2003 09:06 AM