« Fragile chromosomes, cancer and paradigm shift | Gene Expression Front Page | Test your own son? »
October 11, 2003

Wales, England, Cornwall, genetics....

Genetics make Welsh distinct says a story in on a Welsh website[1]. Oh really???

Here are some quotes from the article:

  1. The results show that the Welsh are related to the Basques of northern Spain and southern France and to native Americans. All are descended from the Kets people of western Siberia.
  2. This shows that in the Dark Ages, when the Anglo-Saxons turned up, there was the most horrible massacre on the English side. They killed everybody and replaced them.
  3. In the male line, at least, the Welsh and the Basques are survivors or relics of a period before huge numbers of farmers filled Europe from the Middle East.
  4. The Cornish are in effect Anglo-Saxons who for a time used a language that was hanging around
  5. The genes of Scottish males also betrayed considerable inter-mixing with outsiders

Some of the points are clarified in the article, but I don't think they're clear enough. On point 1, it is the descent of the Y chromosome only, the unbroken patrilineal line. This does not address the mtDNA, which is the unbroken matrilineal line, or the autosomal DNA which generally encapsulates functionality in the genome and is the vast majority of the DNA that we carry. There are often sharp differences in the phylogeny of each genetic marker for various reason[2]. Spencer Wells talks about the Ket lineage in his book Journey of Man (Wells carries the marker himself).

On point 2, see this post and this post (hint: Jones engages in a bit of hyperbole).

On point 3, see my post on demic diffusion.

On point 4, look at this physical map of the UK. Note that the boundary between Cornwall and England is far smoother than that of Wales and England (not to mention artificial boundaries thrown up in the Dark Ages). Physical relief often correlates with high genetic diversity locally. Rather than one movement of Anglo-Saxons into Cornwall-it was probably the slow generational intermigrations that blurred the boundaries between Cornwall and Wales. To say that the "Anglo-Saxons picked up the local language" is like saying that Finns are Swedes who "picked up the local language." In the latter case, it seems likely that genetic exchange with their neighbors has worked to blend the genetic profile of Finns and Swedes together-though there are still differences (the cline of the TAT lineage, etc.). Also, the Cornish were conquered by the English earlier than the Welsh, and had an avenue of escape to Brittany as a safety valve, etc. The reasons to dissent and rebut Jones' point are myriad....

On point 5, duh. Heard of one nation, four peoples? The Angles were one of the peoples that made Scotland what it is, and this ignores the Viking contribution, so the evidence of Germanic genes should not be surprising.

fn1. The story seems to be based on unpublished research and reflects some of the points in a chapter of Steve Jones' Y: The Descent of Man. I couldn't find much more at The Center for Genetic Anthropology at University College-London.

fn2. For example, Y chromosomes are for more likely to show European ancestry in African Americans than mtDNA, for obvious reasons. The Y lineage of mestizos are often similar to Spaniards while the mtDNA lineage is similar to the indigenous peoples of the New World. Additionally, much of the autosomal genome is subject to selective pressures which distort phylogentic relationships but is far more informative of adaptive history.

Posted by razib at 02:56 PM

I think I remember seeing a website address at the end of an article in The American Conservative; it may have been this one. I was thinking about that when I looked at the weblog of another blogger whose article was in the second most recent issue of that magazine. From a past blog entry of his, I followed the link to Steve Sailer's blog, and I saw a mention of Mr. Khan's past article in The American Conservative. I came to this blog, and it looks rather interesting and impressive. I started my blog about a year ago, and it has been an good experience overall.

You site has a nice design and layout. I don't agree with many of the views expressed here, but you are covering important topics and issues. Some of the recent entries here appear intresting and though-provoking. I should look through them some time.

Posted by: Aakash at October 11, 2003 05:12 PM

thanks for the input aakash-i did much of the re-design :)

Posted by: razib at October 11, 2003 06:04 PM

The Basques and American Indians were far-fetched enough, but when the author dragged in the Kets I started to think of Monty Python. The Kets are a very obscure and almost extinct Siberian people, lumped with the Yukagir and the Nivkh as "paleo-Siberians" mostly on the wastebasket principle of wanting to have one place to put them all. Their languages do not seem related in a language-family way. The Ket population is ~500 and dwindling.

Posted by: Zizka at October 11, 2003 07:03 PM

well, there is a germ of truth in what is being said, sometime 20,000 or so years ago when humans pushed north out of central asia, they split in two directions-west into europe and east into northeast asia. ergo, you get the same lineage all the way in the new world and in western europe.

Posted by: razib at October 11, 2003 07:07 PM

Yeah but the Kets are a tiny, very specific minority. They are one of the few remaining survivors of the (certainly once more numerous, but the area has always been thinly populated) pre-Turko-Mongol-Tungus inhabitants of the area and COULD be related to the ur-Native-Americans, but for them to be related to the Basques and the Welsh would basically require that a fully-populated Eurasia be overwhelmed by invaders from the S, except for a few scattered peoples in the far corners. It's sort of a brotherhood of the not-quite-totally-defeated.

In other words this would seem not to be the people surging but the people there before the surge.

BTW on a different site I made joking inquiries about the status of the Welsh in Britain and found that there is a lot of hard feeling still.

Posted by: Zizka at October 11, 2003 10:33 PM

zizka, the idea is that the "ket lineage" started in that area of central asia, and was one of the first migrations into europe (perhaps the first) and obviously the first into the new world. that's why it survived. the ket's themselves probably preserve more of the original diversity of the original lineage, so that's why they are labelled "ancestral." i believe that the "ket" lineage is one of the major ones in northern europe as well....

Posted by: razib at October 11, 2003 10:55 PM

It will be interesting if this news report really is based on new research, as it seems to support the Weale et al. data against Capelli et al.

I'm not surprised at the lack of a distinct Cornish genetic type - Cornwall is too small and too exposed to mixture with the rest of England to remain 'Celtic'.

Posted by: David B at October 12, 2003 03:44 AM

So "Ket" is just a rather peculiar labelling for a formerly common genetic group which once covered most of Eurasia and migrated also to the New World, but has been supplanted by new groups representing most of the present populations. Would these be Gimbutas's "old Europeans" who preceded all of the Indo-European groups and (I think) built Stonehenge?

If the Welsh are part of this group, they would have had to have been assimilated to the invading Celts (?? ~ 500 B.C. ??) linguistically but not genetically. (With the Basques there's no such problem; and I'm sure that some linguist has already come up with a Proto-Ketto-Euzkadian language). Part of my dubiousness is that the info seems to come from a Welsh nationalist site. (My great-grandfather was Welsh and I have no problem with the Welsh).

Posted by: Zizka at October 12, 2003 08:33 AM


well, the reporting is on a welsh nationalist site, but the researcher is a real geneticist, and his book Y has a chapter on the welsh and central irish (the "unmixed" irish) relationship to the basque. and yes, he asserts that the "celtic" languages probably came to the "welsh" and "irish" through elite transmission.

look @ this map-the "Ket" lineage is the one that bifurcates in central siberia. and it still forms a large proportion of northern europeans and most native americans. on an aside, this lineage is also prominent among the yangboni people of tajikistan who are the last remnants of the sogdians.

Would these be Gimbutas's "old Europeans" who preceded all of the Indo-European groups and (I think) built Stonehenge?

i have a hunch-kind of a personal theory-that the "megalith builders" were a pre-literate volkswanderung of sea-people. just like africa today, pre-roman europe was a mosaic that is really hard to discuss without getting bogged down in details we are missing....

Posted by: razib at October 12, 2003 12:03 PM

In my most recent post Stonehenge was a red herring. Gimbutas mostly talked about the pre-Indo-European inhabitants of Europe. As I understand, she was on top of the archaeology, but had a lot of agendas (matriarchy) too.

Posted by: Zizka at October 12, 2003 01:54 PM

zizka, well, i have heard her described as "fantastic" archaeology ;) very entertaining, and grist for the mill of eco-feminists, but basically garbage once she got beyond pottery shards and began to reconstruct ancient societies....

Posted by: razib at October 12, 2003 01:58 PM

As I said, Gimbutas is way out there. But her "old Europeans" sound "Kettish" to me, based on what you've said. As I understand, she's good on the nuts and bolts.

My experience of archaeologists is that they are just infuriatingly bashful about coming to any conclusions at all. (This is called "processual archaeology", which succeeded the grand V. Gordon Childe stuff around 1940-1950 -- maybe someone here knows more about this. Gimbutas apparently is a survivor.)

What I've found is that archaeologists just plain refuse to interpret at all, so you end up with these long reports about the distributions of Type A pottery and Type B pottery, etc., but no conclusions. (As I remember, the Hallstatt culture appears at the places and times when historians believe the Celts appeared. So is the Hallstatt culture the Celts? Archaeologists apparently are forbidden to try to answer that question).

Posted by: Zizka at October 12, 2003 05:33 PM

zizka, the biggest problem with gimbutus is that she says that "old europe" was peaceful & matriarchal. from what i gather, the genetic evidence seems to indicate that europe has been chacterized by:
1) patrilocality
2) smaller number of males reproducing every generation than females (implied polygyny)
(this is true world-wide generally)
the two points don't contradict matriarchy, but seem rather strange in such a society.

Posted by: razib at October 13, 2003 12:10 AM

In the small amount I've read of Gimbutas it hasn't been too hard to sift out the fact from the interpretation. Before the Kettish thing came up I had forgotten her stuff, pretty much. But now the "pre-European" or "old European" inhabitants of Europe seems like a live issue again.

She also wrote a book called "The Balts", and yeah, she's a bit of a Lithuanian patriot. The old Lithuanians were apparently tougher than the Norse (pre. 800 A.D.)

Posted by: Zizka at October 13, 2003 08:12 AM

zizka, you might find this essay titled 'volkswanderung' by me from december interesting (if you weren't reading the blog last year).

Posted by: razib at October 13, 2003 11:50 AM

Actually the genetic connection between the Irish, Welsh, Siberians and Native Americans has already been established. Recent research shows that the Y-chromosome genetic markers M45 and M173 are found in many Siberian and Native American populations, those markers being common amongst the Irish and Welsh as well. Check out the paper below for further details:


Posted by: Diarmid Logan at October 15, 2003 08:12 AM