GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2009-10 > 1255140666
From: Gary Felix <>
Subject: Re: [DNA] R-L21 vs. R-U152
Date: Fri, 9 Oct 2009 19:11:06 -0700 (PDT)
<It appears to me that the L21 hit rate is going to make part of the
continent (northern France) a higher L21 area than the vast majority of
the isles, geographically speaking (England).>
More evidence for L21 being a Mesolithic Doggerland marker.
Mexico DNA Project Admin.
--- On Fri, 10/9/09, Alan R <> wrote:
From: Alan R <>
Subject: Re: [DNA] R-L21 vs. R-U152
Date: Friday, October 9, 2009, 2:45 AM
It is certainly looking like L21 is far too common on the continent for an origin in the isles to still be a serious possibility given the lack of evidence for a huge out of isles demographic movement into large chunks of NW Europe west of the Rhine at any period. The hit rate during L21 testing in France is so high (probably higher than England) that an isles origin is not even worth considering any more. So, L21 prevalence is no longer a simple case of isles=high and continent=low. It appears to me that the L21 hit rate is going to make part of the continent (northern France) a higher L21 area than the vast majority of the isles, geographically speaking (England). Also, now that we know that L21 is exceptionally common in the parts of France where boats would have actually sailed the shortest crossings to the isles in prehistoric times then its hardly surprising that this would have easily led to a heavy representation of L21 along the
seaways of the isles. I suspect that southern England pre-Saxons, Vikings etc would have been genetically very like northern France with a pretty high L21 count that later was somewhat diluted. I think the only national peak that significantly outshines northern France will be Ireland and maybe the Irish Sea areas of Britain where I suspect a founder effect took place and created a peak. Part of the total L21 peak in Ireland is down to M222 which is usually explained as a mushrooming of a lineage due to a greatly successful clan lineage in relatively recent times. Take M222 away and the L21* in Ireland is reduced to a level that is not much different from the north French L21 hit rate. I know some like to filter out non-native Irish names etc to raise the totals but I think its best not too unless the same level of surname sorting is done for all national totals.
The big picture distribution of the main L21 block is most easily perceived if you look at the R-L21 plus project continental map at a scale showing western Europe. This shows it is very much a NW European clade that is almost entirely located in Europe north of the Alps and Pyrenees with only a very few exceptions. Within NW Europe it does bear a striking resemblance to the borders of the northern half of the Gaulish Celtic lands with a substantial drop as one heads east of the Rhine and/or north of the Main into the historically more solidly Germanic lands. it seems incredibly common in northern France (even when Brittany is excluded), west/south Germany and I suspect too from the very decent showing of L21 in Holland south of the old Lower Rhine line (which lay well to the north of today's river) that the gap on Belgium is purely due to very little testing there. Gaps within this block are often likely down to sampling. Compare
the L21 map with the S116* map (which between them show most of the individuals who were L21 tested) and you will see that gaps are sampling gaps.
I would think that L21 is best described as something like a NW European Celtic clade which kind of implies the northern Gaulish lands west of the Rhine/south of the Main,plus the isles. It, so far, is fairly rare outside the NW European former Celtic lands so it seems apt and I have no doubt it was a very big clade among the Celts there, probably the modal. Although I think NW Celtic is a good shorthand description, I only mean that it was probably common among the Celts, not necessarily that it was the Celts of the Iron Age that first spread it. I suspect that it must have mainly spread much earlier than that, in the Neolithic, as indeed (I suspect) the Indo-European roots of the Celtic languages did.
The high outlier L21 presence in Norway is puzzling but any number of founder effect or elite dominance scenarios are possible to explain this and perhaps a stepping stone 'pocket' will be found somewhere like lightly tested Denmark (an area usually thought Germanic but that saw major Celtic influence and perhaps presence in late prehistory). It is also possible that there was simply a founder effect at an earlier date when the sea to Norway was crossed by a small group of settlers.
I suspect that S28 is more common in the southern and Alpine former Gaulish lands like southern France, Switzerland and north Italy while L21 was dominant in the more northern Gaulish lands. This may explain why S28 but not L21 appears in Celtic settled northern Italy. The Celts who invaded northern Italy were mainly those nearest to southern France and to the Alpine area where L21 has a poor showing so far but S28 is common. There are of course other possible explanations.
One other concept that L21 has shattered is the idea of a major genetic split between continental/NW/central European Celtic/Gaulish/'real Celts' 'core' on the one hand and a genetically different Celticised/'not real Celts' Atlantic 'fringe' stretching from Portugal to the Orkneys on the other. This idea was very common when there was only S28, S21 and the the rest was lumped together as M269. L21 has destroyed that idea as its prevalence very much links the isles fringe with the Gaulish heartlands of France and SW Germany and neighbouring areas. Conversely, L21's relative rarity in Iberia and SW France divides this southern part of the proposed Atlantic area from the L21 peak in the northern part of the proposed Atlantic area in the west and north of the isles.
As for date and deeper origin, the jury is still out. However, the idea of a link with early Neolithic Cardial culture that David proposed for S28 really does not work for L21 at all. The project map of L21 and a map of Cardial are almost diametric opposites with the two distributions practically mutually exclusive. A far better distribution correlation for L21 would be the Early Neolithic Linearbandkeramik (LBK) culture (and descendant later post-LBK cultures), the westernmost part of which after passing up from the Upper to Middle Rhine tuned sharply west into the Paris Basin of northern France and then further west as far as the east of Brittany. Clearly later in post-LBK times the same people would have fanned out somewhat and infilled. Unfortunately there are no good modern LBK maps on the web. There are only old ones which show LBK stopping around Paris. More modern LBK maps can be found in Cunliffe's latest books. Anyway,
the west end of LBK and L21 have similar distributions allowing for sampling issues. Perhaps an S116* individual following the rapid spread of this culture along the Danube was the originator of L21, with the mutation perhaps mushrooming as the Rhine was reached. Exactly where the mutation first occurred is unclear but distribution would indicate it must have been north of the Alps, perhaps somewhere along the Danube. The two main routes east-west across Europe were the Danube and the Mediterranean. The distribution of L21 effectively rules out the latter option.
To unsubscribe from the list, please send an email to with the word 'unsubscribe' without the quotes in the subject and the body of the message
|Re: [DNA] R-L21 vs. R-U152 by Gary Felix <>|