The Welsh according to the Financial Times are stone age. It's hard to disagree until you realise the paper isn't referring to the Assembly or local government or the state of our national media. Instead it's a response to the latest scientific offering regarding British DNA, albeit research that's a step-up from the Dafydd Iwan is descended from an old Welsh King of England hooey currently entertaining the public on S4C.
Meanwhile according to the Guardian the Oxford University study reveals that "30% of white British DNA has German Ancestry" (No it doesn't, it merely shows that in the dim and distant past many present day Germans and Britons shared common ancestors.) The Daily Mail says something similar about the French while the BBC gleefully trumpets the fact that the "Celts are not a unique genetic group."
Let's take a layman's look at the sample that allowed the researchers to conclude that the UK's population could be divided into 17 distinct groups, while recognizing that genetic variety in Western Europe is both very homogeneous and very recent. Here's a large scale map.
It looks to me that some of the sampling was designed to prove a point. The large number of samples in Devon and Cornwall, in Pembrokeshire and in North East Scotland (Picts) for example. At the same time other interesting possibilities are ignored. Mid Wales, Carmarthenshire and the South are hardly covered at all, a large area north of London - which stood out in Victorian examinations of negrescence - is empty of samples, as are large areas of Scotland and significant areas of the Welsh border (West Herefordshire and all of Shropshire).
What does the map say about Wales? Well the North Wales grouping is certainly distinct but how far south does it spread? This might have told us something about the extent of the Ordovici lands and there are also no samples from Lleyn (possible Gangani). Blood groups long ago told us something about the differences within Pembrokeshire and no doubt the current study will revive the Little England meme, but is it true? In reality the S Pembrokeshire cluster doesn't appear to coincide with S Pembrokeshire at all, it spreads north and west. There is no obvious link with Devon or Flanders, the usual suspects in the populating of the area and I wouldn't be surprised if the two groupings ie North and South Pembrokeshire both predate the Roman never mind the Norman invasion. If a wider sample had been taken in North Ceredigion, Carmarthen, Glamorgan and Gwent we might have a better understanding of the actual demographic history.
The Welsh border grouping seems heavily weighted to the Forest of Dean while excluding sampling in West Herefordshire and Shropshire. It's a puzzle why the southern (Dobunni?) area should then reappear in Cheshire. Again Wales has been somewhat short-changed by limited sampling in a survey which has been described as perhaps the richest genetic survey of any country to date.
I'm surprised that anyone is surprised that there is no single Celtic grouping. The Anglo-Saxon invasion lasted no more than a few decades yet it has left a distinct 10-40% trace - according to the survey - amongst the central and southern English. The period from the re-populating of Britain following the Ice Age to the Belgae arriving just ahead of the Romans was around 9000 years, for sure there would have been many other population movements during this timespan and subsequently various groupings waiting to be discovered - a single "Celtic" grouping would be highly unlikely.
A large part of the sampling was carried out in Continental Europe, see map, but again there seem to be some omissions. Why nothing from Friesland, after all Frisian is the language most closely related to English? Why no testing in southern Ireland, population movements are not all one way, see the Deisi. There was also limited testing from much of Denmark. Still the absence of what the survey calls FRA17 from all three Welsh groupings, and only the Welsh groupings, does indeed seem significant and nails the South Pembrokeshire is Little England meme - the absence of FRA17 and GER3 suggesting that all three observed Welsh groupings were amongst the earliest inhabitants of post-Ice Age Britain.
As far as Wales and the Welsh are concerned there is still much to learn about historical demographic events and this large survey is far from being the last word.
Let's take a layman's look at the sample that allowed the researchers to conclude that the UK's population could be divided into 17 distinct groups, while recognizing that genetic variety in Western Europe is both very homogeneous and very recent. Here's a large scale map.
It looks to me that some of the sampling was designed to prove a point. The large number of samples in Devon and Cornwall, in Pembrokeshire and in North East Scotland (Picts) for example. At the same time other interesting possibilities are ignored. Mid Wales, Carmarthenshire and the South are hardly covered at all, a large area north of London - which stood out in Victorian examinations of negrescence - is empty of samples, as are large areas of Scotland and significant areas of the Welsh border (West Herefordshire and all of Shropshire).
What does the map say about Wales? Well the North Wales grouping is certainly distinct but how far south does it spread? This might have told us something about the extent of the Ordovici lands and there are also no samples from Lleyn (possible Gangani). Blood groups long ago told us something about the differences within Pembrokeshire and no doubt the current study will revive the Little England meme, but is it true? In reality the S Pembrokeshire cluster doesn't appear to coincide with S Pembrokeshire at all, it spreads north and west. There is no obvious link with Devon or Flanders, the usual suspects in the populating of the area and I wouldn't be surprised if the two groupings ie North and South Pembrokeshire both predate the Roman never mind the Norman invasion. If a wider sample had been taken in North Ceredigion, Carmarthen, Glamorgan and Gwent we might have a better understanding of the actual demographic history.
The Welsh border grouping seems heavily weighted to the Forest of Dean while excluding sampling in West Herefordshire and Shropshire. It's a puzzle why the southern (Dobunni?) area should then reappear in Cheshire. Again Wales has been somewhat short-changed by limited sampling in a survey which has been described as perhaps the richest genetic survey of any country to date.
I'm surprised that anyone is surprised that there is no single Celtic grouping. The Anglo-Saxon invasion lasted no more than a few decades yet it has left a distinct 10-40% trace - according to the survey - amongst the central and southern English. The period from the re-populating of Britain following the Ice Age to the Belgae arriving just ahead of the Romans was around 9000 years, for sure there would have been many other population movements during this timespan and subsequently various groupings waiting to be discovered - a single "Celtic" grouping would be highly unlikely.
A large part of the sampling was carried out in Continental Europe, see map, but again there seem to be some omissions. Why nothing from Friesland, after all Frisian is the language most closely related to English? Why no testing in southern Ireland, population movements are not all one way, see the Deisi. There was also limited testing from much of Denmark. Still the absence of what the survey calls FRA17 from all three Welsh groupings, and only the Welsh groupings, does indeed seem significant and nails the South Pembrokeshire is Little England meme - the absence of FRA17 and GER3 suggesting that all three observed Welsh groupings were amongst the earliest inhabitants of post-Ice Age Britain.
As far as Wales and the Welsh are concerned there is still much to learn about historical demographic events and this large survey is far from being the last word.
Good write up, DNA can prove anything the person paying for it wants, just like Opinion Polls.
ReplyDeleteThe genetic research has probably been going on for a while, but thier results are handy for a paranoid Britsh establishment and rabid rightwing press wary of the contined rise of of the SNP and all things Scottish.
Im sure in not the only one whose noticed every small thing that can be, being used to promote Britishness and Englishness while erasing or undermining anything Scottish, Welsh or Irish.
I'm beginning to wonder where this relentless promotion of Britishness is taking us. None of the destinations I see look inviting.
ReplyDeleteIt was a big mistake for this survey to exclude the Irish Republic.
ReplyDeleteThe media spin goes against the findings eg Little England, Landsker nonsense. In reality the "south" Pembrokeshire grouping is one of three oldest - all in Wales - and also has none of the Saxon marker.
The Saxon marker is present in most of England, between 10-40%, which suggests a sizeable folk movement. The argument has been between those who say the Saxons slaughtered the Britons and those who say they integrated. The survey suggests a combination of the two. It would have helped if samples had been taken in Brittany.
Britishness? This is very interesting, I've thought that Britishness was for the little people and that the elite were developing into a supra-national grouping above the nation state. I still think that but at this stage of history we do seem to have an Anglophone empire plus a few vassals eg the international community ranged against unpeople who can be attacked with impunity.
As for the media spin Welsh and Scottish nationalists can see the spin against their own countries just as the left in England can see the spin against trade unions, public ownership etc. What the majority fail to see is the even greater, almost total propaganderising of foreign affairs
As to regions of Wales lacking sampling, this may have something to do with stringent criteria imposed by the study (and connected to the statistical analysis).
ReplyDelete"Approximately 4000 rural samples from throughout the UK have so far been collected using the criteria that all four grandparents were born in the same rural area, defined as lying within 60 km linear map distance of each other. "
The nitty gritty of sampling is here (open access I hope)
http://www.nature.com/ejhg/journal/v20/n2/full/ejhg2011127a.html
Figure 2 suggests there were people from mid-Wales in the collection
I am interested to know how feasible it is to identify Radnorians that conform to these criteria?
Of course the populations in central Wales are small and there has been a fair amount of movement even within the farming communities but still I would say that there were many who met that 60km criterion.
ReplyDeleteOh I should have added that one shouldn't be too precise about county borders. Indeed the old Welsh territorial divisions would seem to me to have more cultural relevance, even today. What I miss from the study are things like: where does the North Wales grouping end, what about the Silures of SE Wales, what about Shropshire etc
ReplyDeleteIreland is also a big gap in this study. Would be interesting to see if mixing of Welsh and Irish populations is detectable genetically.
ReplyDelete