New DNA study on Sub-Roman Britain

[Walhaz]

DNA Study shows Celts are not a unique genetic group

Interesting report on this study. The study authors state that their results show that there was not one, single, Celtic grouping in Britain, and further that the Anglo-Saxon invasion did not push the various Celts out, but rather the Anglo-Saxons and Celtic groups mingled together.

[Walhaz]

Just saw that this had already been posted in the Boudicca's Last Stand discussion. Apologies.

[John1]

The paper is only referenced in the Boudicca thread because of one potential tribal boundary and will soon be lost on that swamp of a thread. This paper and early dna is certainly worthy of it's own thread and should probably be read in conjunction with Steve Bird's 2007 paper and Heinrich Harke's 2011 paper which can be found here;

http://www.jogg.info/32/bird.pdf

www.academia.edu/1178275/Anglo-Saxon_immigration_and_ethnogenesis._Medieval_Archaeology_55_2011._1-28

[Robert Vermaat]

I still have my doubt how they can research 200 people's dna and then reach such detailed conclusions about what happened to large groups of people 1000-3000 years ago.

"Likewise, the Norman conquest of England did not leave any genetic evidence.". Well, maybe the people who moved back into the depopulated nothern counties were of the same stock as the victims?

Someone remarked, based on this research, that apparently the Romans had not intermarried into the British population at all because they left no dna. But what about the probability that the Romans in Britain were mostly from NW Gaul and therefore perhaps less visible?

[John1]

200 people's dna
I think it was 2039 samples

Norman conquest of England did not leave any genetic evidence
I think this statement is because the Normans were such a mixed bag, much of the Y-DNA coming from Scandinavia.

Romans..... left no dna
Can you attribute that remark? there is a debate going on about E-V13 based on the Bird paper and E-L19 which is as yet unpublished. It would seem extremely unlikely that a 400 year occupation/contact would not leave some dna traces.

There are some heavy hitters on the authors list so I am pretty uneasy about dismissing the piece.
http://www.isogg.org/wiki/People_of_the_British_Isles

[Urselius]

I haven't seen the original paper but some of their conclusions seem strange. The closest relatives of the English would appear to be the French (40% identity), ascribed to a folk migration "sometime after the end of the last Ice Age". However, the English relatedness to the Germans (20%-30%) is ascribed to the Anglo-Saxon invasions of c. 450 AD. How did they work out this dating?

[Robert Vermaat]

200 people's dna
I think it was 2039 samples

Correct, that was a typo. :oops: But still.

Romans..... left no dna
Can you attribute that remark? there is a debate going on about E-V13 based on the Bird paper and E-L19 which is as yet unpublished. It would seem extremely unlikely that a 400 year occupation/contact would not leave some dna traces.

Yes, it was a comment made in a review of the article by NewScientist.

There are some heavy hitters on the authors list so I am pretty uneasy about dismissing the piece.
http://www.isogg.org/wiki/People_of_the_British_Isles

Dismising is a big word but it would not be the first time when either a) claims were distorted in media articles or b) claims were made because OF media articles. The number of times the house/grave/site of miracle of Jesus was 'found' by very serious scholars is staggering 9but unsurprising). A very 'heavy hitting' etymologist once had the the gall to claim that the name 'Badon' 9from the (in)famous battle of Badon Hill) had the be Germanic "because we have not a similar name in Celtic". I'm no longer surprised by any claim even though it's seems not logical at first glance.
Mind you I'm not dissing the article, I like some of the results very much, but I still have nagging doubts about the models such studies use. I mean, what are the possibilities to check the results, other than a massive study in the dna of skeletons of the periods under study? A study which, so far, seems impossible?

[Robert Vermaat]

I haven't seen the original paper but some of their conclusions seem strange. The closest relatives of the English would appear to be the French (40% identity), ascribed to a folk migration "sometime after the end of the last Ice Age". However, the English relatedness to the Germans (20%-30%) is ascribed to the Anglo-Saxon invasions of c. 450 AD. How did they work out this dating?

Exactly. Such conclusions sound awfully like the proof of a theory and not the result of a set of hard irrefutable data.

[Walhaz]

I get the validity of sampling for Celtic DNA, and seeing where the DNA is and which other DNA it matches. Basically, if the Celts were there previous to the Angles/Saxons/Jutes and Normans, then sampling for Celtic DNA markers could potentially indicate where those populations had eventually settled down after the invasions/migrations/friendly exchanges/insert-own-pet-theory-here. There are of course other reasons that the DNA markers would be present other than in-place perpetuation of the genes (later migrations, recent arrivals from abroad, etc.) but hopefully the sampling would be enough to rule out such outliers. I do share Robert's concern that the sample size is fairly small, but also I imagine that larger DNA samplings would be more costly to run.

I find the information that these genetic groups may have been different so long ago quite fascinating. I know there have been previous genetic studies done, and that none of them are satisfactorily definitive, but it does seem that each study increases our knowledge just a little.

[Urselius]

There is no reason to think that an Iceni tribesman was genetically identical to a Silurian tribesman in 43 AD. They may have been as genetically dissimilar as a modern Norfolk farmer is from a modern farmer from Glamorgan. Until large-scale work on ancient DNA is undertaken we will have no definitive answers. Projecting modern population genetics back in time is just guesswork tied up with statistics and unreliable mutational 'DNA clocks'.

[John1]

Until large-scale work on ancient DNA is undertaken
I talked to a geneticist about that once, at the time it was about $1500 to sequence a genome from a live subjects swab, but about $50,000 to do an archaeological sample. I don't know if those figures have changed but it seems modern populations are the only viable way forward at the moment. Hopefully that will change.

[Urselius]

However, sequencing a whole genome is not necessary for indications of relatedness. The real problem with ancient DNA is how to avoid getting it contaminated by recent DNA. The number of labs that can do this reliably is relatively small.

[Urselius]

I now have a PDF of the full paper. In reference to my earlier comments the following quotation from the paper is very telling, "‘Old’ and ‘recent’ here are relative terms—we can infer the order of some events in this way but not their absolute times. Although we refer to migration events, we cannot distinguish between movements of reasonable numbers of people over a short time or ongoing movements of smaller numbers over longer periods." I expect that a lot of the pronouncements in the press are exaggerations of what the authors have actually claimed, sensationalised by rather ignorant press hacks.

[Caballo]

I've also read the paper- and much of the criticism is at the journalists who have chosen headlines that are suitably provocative and often come to conclusions not contained in the paper. The 2039 people had to have all four grandparents born within 80 km of each other- effectively pushing the analysis back back to the late 19th century. Worth reading the paper!

Here are a couple of the better articles

http://www.sciencedaily.com/releases/201...145420.htm
http://www.theguardian.com/commentisfree...nglo-saxon

[John1]

a new bit on pre-Anglo Saxon dna, some Roman observations;
http://www.nature.com/articles/ncomms10326