PDA

View Full Version : DNA and surnames



Jon
11-15-2014, 09:31 PM
There's a debate started at the Yahoo group about possible Norman origins to L513 surnames, and therefore the HG generally...I have to say, I don't quite follow the logic. Firstly, surname adoption generally started around the Norman times anyway, which means that many natives would have ended up with Norman sounding names randomly, or through land/tenure loyalties to landowners, many of whom in southern Scotland were Norman post-1100. We see the same thing with big clan names: very few MacDonalds can trace their ancestry back to the original chiefly line, for instance: most were unrelated clansmen. Secondly, we all know how dangerous it is to equate surname history with genetics, given the high rate of undetectable NPE, adoptions, etc etc. Finally: am I wrong, or has L193 for example not been roughly dated at far before the age of significant Norman settlement in Scotland?

There are surely enough examples to show that such a clear-cut conclusion is hard to justify...how would a MacVicar on South Uist or a MacDonald on Skye (both L193), for instance, feasibly trace back to Norman lines?

If I'm missing information here, I will be the first to admit my error...but right now I'm having trouble seeing how this adds up.

Jon

AnnS
12-21-2014, 07:38 PM
Hi Jon

I agree with you. People work within the frame of reference they bring with them. Surnames are handy, especially when the haplotype seems to be the core haplotype of the original family (e.g. Glendennings, Elliott's, Littles, McLeans and L193; Mackenzie and Group E). But surnames alone don't tell the story. I think there was an unrecorded adoption that happened in 1400 in the Mackenzie line, as I've said several times on the Yahoo group forum. The location of the surname and its proximity to others who have neighboring surnames and share the same DNA is a good indicator of on old cluster in the area. There certainly is a large diverse group of L193 folks living in the general area of Ayrshire, extending from Galloway and Carrick up through Lanarkshire and the Clyde river area. These people were there before surnames and therefore before the Norman influx. If there ever was a doubt about that, that doubt should be erased now, based on SNP evidence. L193's cousins are the A7 SNP people, including the D1 and E groups. D1 and E look decidedly Scottish, in spite of the branchlet of D1 that ended up in Ireland. There is no scenario that has L193 originating in Scotland, migrating to France, and then coming back over to the same area of Scotland in the form of Norman knights.

Now if the question is did L513 originate from northern France or migrate through the Normandy area of France on its way to Britain, that answer could be yes. I think L193 originated in Scotland. On the other hand, I think the Ros and Massie families of Group C have a common ancestor in Normandy and that their Ros and Masssie's came over with Wm the Conqueror.

Ann

Salkin
12-21-2014, 08:47 PM
I envy people from long-time surname cultures sometimes. Around here, surnames in the usual sense were not adopted until around 1900. Before then almost everyone (who wasn't noble, anyhow) used patronymics.

Of course, it turns out the surname I inherited was a polite lie, so it wouldn't have been useful much further back than when it turned up anyway, but still.

Dubhthach
12-22-2014, 09:30 AM
It wouldn't surprise me if L513 arose in what is now France, however this was probably 1,500 years+ before the time of the Norman's.

AnnS
12-22-2014, 03:18 PM
It wouldn't surprise me if L513 arose in what is now France, however this was probably 1,500 years+ before the time of the Norman's.

I agree with you, Dubhthach. Some of them stayed on the Continent, in that general area west of the Rhine, and some went west to Britain. It amazes me that they were so successful in Britain, arriving, as they did, well after the metal-working L21 guys. According to Michael Hammer's report of 2013, they are about 10 percent of the R1bs in each of England, Scotland and Ireland.

SearchSeeker
12-22-2014, 05:19 PM
Interesting, that may explain that the L-513 O'Sheas are actually the Normans (or arriving with Norman conquest) and not the L1066 O'Sheas. If that turns out to be the case, some have a lot of editing to do on quite a bit of information put out there on the internet and in other documents.

SearchSeeker
12-22-2014, 05:25 PM
Jon, where specifically is this debate going on, in which Group, a L513 group? I would imagine there are some that are fiercely opposing any connection to the Normans or an element that came over with William the Conqueror.

dp
12-22-2014, 05:30 PM
...
Welcome to Anthrogenica.
dp :-)

dp
12-22-2014, 05:31 PM
...
Welcome to Anthrogenica.
dp :-)

SearchSeeker
12-22-2014, 06:06 PM
Welcome to Anthrogenica.
dp :-)

David, thank you x2.

Jon
12-25-2014, 07:44 PM
Hi All,
Yes, welcome SearchSeeker!

This debate was started, although not taken up for long, on the old Yahoo 11-13 group. I agree with what Ann said above - it wouldn't surprise me if L513 had originated in France, although clearly before Norman times. L193 seems so completely exclusive to Scotland (despite low testing rates) that I can't see beyond an origin there. Dave Vance posted a real interesting chart with color codes for L513 (also on Yahoo), and observed that the 5668 branch seemed to focus more on Ireland and Scotland (including L193) while the S6365 branch is a lot more mixed in terms of origins. Given that L513 is likely to be very old, the 5668 and 6365 could therefore represent two quite different groups of people in more recent historical times. I'm pretty convinced that L513 in the Isles must in its majority represent 'indigenous' DNA - Brittonic, Irish/Scotti, Pict, whatever. But I'm also sure the Normans will also have brought some in with them as well.

Jon

rms2
12-26-2014, 02:29 PM
Low testing rates? For L193? As I recall, L513 and its descendant, L193, were among the very first SNPs to be discovered downstream of L21. They have had several years of testing, and, unless my memory is failing, L193 was included in one of FTDNA's early "Deep Clade" arrays. I haven't counted the number of L193 tests listed (and I won't), but I know there is quite a list at the L21 Project of men who have been tested for L193. I would guess that it is one of the most tested of all the L21+ SNPs.

If L193 is showing a predominantly Scottish distribution, that is probably significant, because there really has been quite a bit of testing for it.

Jon
12-26-2014, 03:37 PM
Thanks - yes I guess of all available, it has actually been offered for a long time. I was more referring to the under-testing generally, and more specifically non-isles. I'd love to see more testing in France, especially with regards to L513 for instance. However compared to other groups, for instance DF21 or M222, L193 really does seem to be stubbornly Scottish. In terms of population groups, I guess it could qualify as Brythonic, Irish/Scotti or Pictish (or indeed some or all of these). The major expansion occurred around 400AD or so I've read, in SW Scotland, right when the Romans left. So there are plenty of possible candidates.

rms2
12-26-2014, 03:46 PM
You are right about low rates of continental testing. It would be nice if we could go back and test Busby's samples for the downstream clades. That would be something.

Jon
12-26-2014, 06:43 PM
Do you think that the current distribution patterns (e.g. L193 being 'Scottish') are nevertheless significant?

rms2
12-31-2014, 01:51 PM
Do you think that the current distribution patterns (e.g. L193 being 'Scottish') are nevertheless significant?

I think they could be, but it would be nice to have a scientific study. In terms of the commercial dna testing market, and relative to other L21 subclades, L193 is not under tested.

Jon
01-01-2015, 11:57 AM
I just looked again at the FTDNA maps of 11-13. It's so tempting to put down L513 generally as ancient Britonic (with typical representation in central Europe); then L193 representing the Scottish branch (Strathclyde?). It's not that simple, I know, but humans naturally seek out patterns and meaning! Happy new year!

TigerMW
01-22-2015, 02:37 AM
I just looked again at the FTDNA maps of 11-13. It's so tempting to put down L513 generally as ancient Britonic (with typical representation in central Europe); then L193 representing the Scottish branch (Strathclyde?). It's not that simple, I know, but humans naturally seek out patterns and meaning! Happy new year!

I am just speculating but I think the majority of L513 is ancient Brythonic, even those in Ireland with Gaelic names may just been part of groups coming in.

I think Breton types could be part of this and we can't forget some of our most unusual haplotypes are from Benelux and Germany. L513 could easily have arose somewhere between Bretagne and the lower to even mid-Rhine.

Jon
01-22-2015, 04:49 PM
I agree with you Mike; although it's encouraging to hear it from you, as I know you have much more of an overview of the DNA aspects than I. I suppose that could explain why it's found at high frequency in (western?) coastal areas - those being the remnants of the Brits pushed outwards by incoming groups - as well as around the old Norman March areas and others (possibly also actually including the descendants of Breton-Norman settlers).

Dave-V
01-22-2015, 08:30 PM
At the heart of this discussion is the continual question of whether the British Isles concentrations in L513 represent a major initial migration to the Isles (or for some, even an origin there for L513) with individual lines back-migrating to the continent, or whether they are a result of a series of separate migrations to the Isles. My own speculations are more the second than the first.

For example, you've got the "fringe" branches (hey I'm one of them) like Z23519, Z23532, etc. which separated from the rest of L513 3000+ years ago but are still found in Ireland and England. Some would point to those as evidence of the age and diversity of L513 in the Isles. I would argue they are evidence of separate migrations, although age of the branches has nothing to do with timing of those migrations. But if I'm honest I'd argue that mostly for the non-scientific reason that I can't make myself believe that my male ancestors for most of 3000+ years of major population shifts couldn't manage to move away from their former relatives.

If L513 was in Brythonic Britain, it's not hard to look at the patterns of MDKA origins and identify early subclades - especially A7 under S5668, or perhaps BY16 under S6365, that have early branching, good diversity, and strong British Isles connections as the most likely to have been in an early perhaps Brythonic-era migration to the British Isles. But that's highly speculative of course for all the usual reasons; Isles-biased data and self-identified recent origins with no accounting for NPEs, etc. Some of the surnames within even those subclades have other origins (Norman, etc) which could still be either NPEs or individual lines back-migrating and coming back in a later migration. I do believe that across 3000+ years we will find cases of individual lines circling out and back to the Isles, although it wouldn't be the norm.

Clearly I'm interpreting the data to fit a theory; but that's pretty much what everyone does when they speculate. The data can obviously support other conclusions too. This won't be scientific (IMHO) until we get more evidence of L513 patterns on the continent.

Jon
01-23-2015, 04:54 PM
Well put Dave, especially regarding the need for more continental testing.

Does anyone know the relative frequency of L513 in Wales (i.e. the modern-day Welsh population?). If we're looking at ancient Brythonic, one might expect a healthy frequency in Wales. Most of what I read concerns Scotland and Ireland...I guess the relatively lower frequencies in England are akin to the slightly lower rates of L21 there?

SearchSeeker
02-02-2015, 01:26 AM
I might speculate to say the majority of ancient Brythonic is not L513, since we're all speculating here.


I am just speculating but I think the majority of L513 is ancient Brythonic, even those in Ireland with Gaelic names may just been part of groups coming in.

I think Breton types could be part of this and we can't forget some of our most unusual haplotypes are from Benelux and Germany. L513 could easily have arose somewhere between Bretagne and the lower to even mid-Rhine.

Dubhthach
02-02-2015, 09:39 AM
Should be noted that in context of Iron age (300BC-100AD -- up to arrival of Roman's into Britain) that the differences between Celtic dialects between Britain and Ireland would have been relatively small. One can only talk about "Brythonic" as a distinct language group in context of Roman contact (eg. before this the written language appears to match with Gaulish heavily). Obviously Proto-Goidelic underwent a number of sound changes which is evident in switch between Archaic Irish and Old Irish (Ogham stones vs manuscript tradition).

So in context of period 300BC-100AD the difference between two area's would be probably akin to difference between modern Scandinavian languages.

Jon
02-02-2015, 05:07 PM
I've read some interesting stuff regarding the age of Gaelic in Scotland. Given the massive extent of Gaelic place names (easily the biggest single linguistic source of them in Scotland), many are now concluding that a form of Gaelic must have been spoken in Scotland pre-Irish settlement. Seems only logical, given the proximity, especially of the west Scottish coast. L513 seems so widespread that it also must have been shared widely among the various groups and tribes in Britain and Ireland? The subclades (e.g. L193; L69 etc.) seems to be more geographically specific however.

TigerMW
02-03-2015, 05:04 PM
I might speculate to say the majority of ancient Brythonic is not L513, since we're all speculating here.
I don't think you'll get much argument on that as L513 is not that large of a subclade anywhere, including in Great Britain. The largest L21 subclade found in Great Britain might be DF21.
I think Wales and NW England might be important places to look for people with ancient Old Briton descent, just because those are a long the edges of old Roman controlled and then Anglo-Saxon controlled regions.

TigerMW
02-03-2015, 05:10 PM
Should be noted that in context of Iron age (300BC-100AD -- up to arrival of Roman's into Britain) that the differences between Celtic dialects between Britain and Ireland would have been relatively small. One can only talk about "Brythonic" as a distinct language group in context of Roman contact (eg. before this the written language appears to match with Gaulish heavily). Obviously Proto-Goidelic underwent a number of sound changes which is evident in switch between Archaic Irish and Old Irish (Ogham stones vs manuscript tradition).

So in context of period 300BC-100AD the difference between two area's would be probably akin to difference between modern Scandinavian languages.
This also correlates with Mallory's suggestion in the "The Origins of the Irish" where he cites O'Reilly (I think) with the thought that much of the descendant population in Ireland came there through ancestor groups in old Great Britain.

Dubhthach
02-03-2015, 09:12 PM
This also correlates with Mallory's suggestion in the "The Origins of the Irish" where he cites O'Reilly (I think) with the thought that much of the descendant population in Ireland came there through ancestor groups in old Great Britain.

I wouldn't rely on O'Reilly he's dead over 60 years, been alot of increase in knowledge since in terms of archaelogy as well as even when it comes to stuff such as manuscripts/sources/linguistics.

I should note of course is to get to Ireland the closet route is via Britain and then via Brittany, thence even today you have regular car ferry service from Ireland to Roscoff in Brittany.

From linguistic point of view there's some debate with regards to Eoghanachta of Munster having been recent enough arrivals (eg. coming to dominance in 5th century), some of this is tied in with mention of name "Nia Segamain" (nephew of Segamain -- Segomo is a Gaulish god equated by Roman's with Mars and Hercules )

There is an ogham stone from Waterford that appears relevant:

http://ogham.celt.dias.ie/stone.php?lang=en&county=Waterford&stoneinfo=inscription&stone=263._Ardmore_I

R-FGC11134 (parent of CTS4466) does seem to have some men (who are CTS4466-) with origins in northern France/Cornwall, the high presence of Z253 in Munster (L1066 and of course L226) might also point to connections between Munster and Northern France.

What's notable from archaelogical point of view is that the contact with Britain that resumes after 300-200Bc (there had been 500 year "dark age" before this) is restricted to northern half of island of Ireland. In context of Irish historical narrative a trope that is contuinously recycled is concept of division of Ireland into two halves.

Leath Cuinn (the half of Conn -- Dál Cuinn == Connachta and Uí Néill)
Leath Mogha (the half of "Mug Nuadat" (slave of Nuadha) -- titular ancestor of the Eoghanachta)

SearchSeeker
02-20-2015, 02:00 PM
I wouldn't rely on O'Reilly he's dead over 60 years, been alot of increase in knowledge since in terms of archaelogy as well as even when it comes to stuff such as manuscripts/sources/linguistics.

I should note of course is to get to Ireland the closet route is via Britain and then via Brittany, thence even today you have regular car ferry service from Ireland to Roscoff in Brittany.

From linguistic point of view there's some debate with regards to Eoghanachta of Munster having been recent enough arrivals (eg. coming to dominance in 5th century), some of this is tied in with mention of name "Nia Segamain" (nephew of Segamain -- Segomo is a Gaulish god equated by Roman's with Mars and Hercules )

There is an ogham stone from Waterford that appears relevant:

http://ogham.celt.dias.ie/stone.php?lang=en&county=Waterford&stoneinfo=inscription&stone=263._Ardmore_I

R-FGC11134 (parent of CTS4466) does seem to have some men (who are CTS4466-) with origins in northern France/Cornwall, the high presence of Z253 in Munster (L1066 and of course L226) might also point to connections between Munster and Northern France.

What's notable from archaelogical point of view is that the contact with Britain that resumes after 300-200Bc (there had been 500 year "dark age" before this) is restricted to northern half of island of Ireland. In context of Irish historical narrative a trope that is contuinously recycled is concept of division of Ireland into two halves.

Leath Cuinn (the half of Conn -- Dál Cuinn == Connachta and Uí Néill)
Leath Mogha (the half of "Mug Nuadat" (slave of Nuadha) -- titular ancestor of the Eoghanachta)

I think there is a very good possibility Z253 was right there in the mix along with, or as part of the Eoghanachta. My own research keeps hinting back to a connection there. I've posed this before but if it is accepted that L226 are Dal g'cais to understand Z253 one only has to work back from the Dal g'cais origins. Sometimes it feels like there is an effort not to make this simple connection as to who the most likely progenitor is, meaning what higher level group/tribe/dynasty, etc. However I believe the data shows an existence in the Isles well before 500 BC, referring back to some variance/age estimates done a while back by MJOST.

Dubhthach
02-20-2015, 02:52 PM
I think there is a very good possibility Z253 was right there in the mix along with, or as part of the Eoghanachta. My own research keeps hinting back to a connection there. I've posed this before but if it is accepted that L226 are Dal g'cais to understand Z253 one only has to work back from the Dal g'cais origins. Sometimes it feels like there is an effort not to make this simple connection as to who the most likely progenitor is, meaning what higher level group/tribe/dynasty, etc. However I believe the data shows an existence in the Isles well before 500 BC, referring back to some variance/age estimates done a while back by MJOST.

As the current "The O'Brien" is confirmed L226+ (Baron Inchiquin) it's fairly obviously than yes the Dál gCais (well at least the Uí Thairdelbaig the direct line that Brian was on). Bart Jaski has some slides from a talk he gave about the origins of the Dál gCais up on academia.edu

They only cover the genealogy tree plus a map of distribution of relevant dynasts.

The (legendary) rise of Dál Cais
https://www.academia.edu/6671376/The_legendary_rise_of_D%C3%A1l_Cais

Was part of a talk he gave at a 1000 year anniversary academic conference on the Battle of Clontarf last year in UCD (University College Dublin).

In general it's regarded that the Dál gCais are an offshoot of the Déisi of Munster (Waterford area) who over time migrate northwards, basically conquering Clare after the south Connacht kingdom of Uí Fiachrach Aidhne fell into decline. They only start getting mentioned in the likes of the annals in the early 10th century.

Of course in context of Munster the Eoghanachta look to have only really taken control in the late 5th/early 6th century. Their rise parallels that of rise of the Dál Cuinn in the northern half of Ireland.

Dubhthach
02-20-2015, 02:53 PM
I think there is a very good possibility Z253 was right there in the mix along with, or as part of the Eoghanachta. My own research keeps hinting back to a connection there. I've posed this before but if it is accepted that L226 are Dal g'cais to understand Z253 one only has to work back from the Dal g'cais origins. Sometimes it feels like there is an effort not to make this simple connection as to who the most likely progenitor is, meaning what higher level group/tribe/dynasty, etc. However I believe the data shows an existence in the Isles well before 500 BC, referring back to some variance/age estimates done a while back by MJOST.

As the current "The O'Brien" is confirmed L226+ (Baron Inchiquin) it's fairly obviously than yes the Dál gCais (well at least the Uí Thairdelbaig the direct line that Brian was on). Bart Jaski has some slides from a talk he gave about the origins of the Dál gCais up on academia.edu

They only cover the genealogy tree plus a map of distribution of relevant dynasts.

The (legendary) rise of Dál Cais
https://www.academia.edu/6671376/The_legendary_rise_of_D%C3%A1l_Cais

Was part of a talk he gave at a 1000 year anniversary academic conference on the Battle of Clontarf last year in UCD (University College Dublin).

In general it's regarded that the Dál gCais are an offshoot of the Déisi of Munster (Waterford area) who over time migrate northwards, basically conquering Clare after the south Connacht kingdom of Uí Fiachrach Aidhne fell into decline. They only start getting mentioned in the likes of the annals in the early 10th century.

Of course in context of Munster the Eoghanachta look to have only really taken control in the late 5th/early 6th century. Their rise parallels that of rise of the Dál Cuinn in the northern half of Ireland.

SearchSeeker
02-20-2015, 04:17 PM
According to the legendary rise of the Dal Cais, it would then follow logically that Z253 is in there, or snps between Z253 and L226 as including Ailill Olumm and the Cormac Cass, assuming there is legitimacy to that pedigree. I guess it would have to be snps between Z2534 and L226 that would encompass Ailill Olumm. Am I reading something wrong or would this not make the Éoganacht something under Z253?

Apparently all of this would all be well below the date for the Z253 origin and even my L1066 origin since this pedigree begins around the 2nd Century ad. L1066 would have had to split off well before all those guys came along, if it's really upwards of 3000 years old. A question might be where were these guys' ancestors between 1500 and 1000 BC, I think they were in Ireland, unless they came over as Milesians, which no one likes to hear as anything other than a fairy tale (some parts are of course).

TigerMW
03-04-2015, 02:29 AM
There's a debate started at the Yahoo group about possible Norman origins to L513 surnames, and therefore the HG generally...I have to say, I don't quite follow the logic. Firstly, surname adoption generally started around the Norman times anyway, which means that many natives would have ended up with Norman sounding names randomly, or through land/tenure loyalties to landowners, many of whom in southern Scotland were Norman post-1100. We see the same thing with big clan names: very few MacDonalds can trace their ancestry back to the original chiefly line, for instance: most were unrelated clansmen. Secondly, we all know how dangerous it is to equate surname history with genetics, given the high rate of undetectable NPE, adoptions, etc etc. Finally: am I wrong, or has L193 for example not been roughly dated at far before the age of significant Norman settlement in Scotland?

There are surely enough examples to show that such a clear-cut conclusion is hard to justify...how would a MacVicar on South Uist or a MacDonald on Skye (both L193), for instance, feasibly trace back to Norman lines?

If I'm missing information here, I will be the first to admit my error...but right now I'm having trouble seeing how this adds up.

Jon
Going back to L513, it seems we still can't get away from the Normans. I'm not saying that L513 is a Norman paternal lineage but we somehow have many elements of L513 mixed in with them. We just keep finding more folks with some connection.

Here is the latest example which by the way is a demonstration of the value of NGS testing. In L513 there is an STR signature that has been very consistent and we label it variety "B2". B2 appears to be about a 1000 years old, maybe as little as 800 or as much as 1250.

I'll go through this inside-out (from me).

We have our tenth Big Y result in now. I'm in as Walsh and I have found my closest match who's MDKA turned out to be surnamed Welsh. True enough, we have a couple of SNPs no one else seems to have. My type of Walsh had a 19th century priest who researched us and said our type was Cambro-Norman (his words were Anglo-Norman.)

A surprise from Big Y testing was that the next closest person in B2 was surnamed Jackman. Jackman says their type is not English but Cambro-Norman. I had been expecting B2 group of Barrett's to be closest because an old poem has that Barrett and Walsh were brothers in the Cambro-Norman Invasion. Barrett is solidly B2 but his closest match by SNP determination is an Adams of West Midlands, possibly Hereford.

We also have a Morris and Wilson tested and then the seemingly odd man out, Sunneson from Ostergotland. Morris said he had a Walsh inter-marriage somewhere way back. I didn't think too much of the Morris' surname until he pointed out some Morris's think they are a variant of the name Maurice or FitzMaurice. Morris and Sunneson kind of stand alone up until today.

Here's a picture of the SNPs derived for "B2".

https://dl.dropboxusercontent.com/u/17907527/R1b-L513_SNP_Tree_Discovery_B2.pdf

The new news is the person I didn't have assigned. He must not have had 67 STRs, but he is most significant, so far, origin-wise. Up until now we had a set of 18 point SNPs that I had graded C or higher that everyone in B2 seems to have. This must represent some bottleneck or lineage struggled through. The phylogenetic block of 19 was cut to 16 with everyone having the 16 except the last guy, Rogers. Rogers is the far right column. Welsh, Walsh and Jackman are the first three.

Rogers looks like another name with the Norman connection. I realize the Normans had a heavy influence on English society and surnames, but we just can't seem to get away from them.

Dubhthach
03-04-2015, 11:20 AM
Case though of the Walsh's of Ireland is that you are looking at the everyday men at arms/archers who arrived in, they were generally of Welsh origin but obviously sworn to a "Norman" lord. (Such as the obvious Cambro-Norman Barry (de Barry), Fitzgearld and Gogin (de Cogan) families). My feeling is that L513 results showing up is actually more due to presence of L513 in west Britain since at least the Iron age.

This would tie in with L513 been present in Scotland (L193) and spread over into Ireland (Airghíalla II) during the Iron age.

TigerMW
03-04-2015, 11:31 PM
Case though of the Walsh's of Ireland is that you are looking at the everyday men at arms/archers who arrived in, they were generally of Welsh origin but obviously sworn to a "Norman" lord. (Such as the obvious Cambro-Norman Barry (de Barry), Fitzgearld and Gogin (de Cogan) families). My feeling is that L513 results showing up is actually more due to presence of L513 in west Britain since at least the Iron age.

This would tie in with L513 been present in Scotland (L193) and spread over into Ireland (Airghíalla II) during the Iron age.

L513 is probably over 3000 years old so I don't if we can have them restricted too much in territory by the Classical timeframe. L513 could have been fairly decentralized by the time the Roman Empire invaded Britain. We have early branching or unnassigned continentals like de Crombugghe and Ammerlan, Swedes like Jonnson and even a couple of Germans. We have few Frenchman, but they show a closer affinity to later branching and could well have been some kind of Irish or Brit movement be it from the Anglo-Saxon era or later.

I agree the case is quite good for the "B2" or CTS6621+ in western Britain being part of some peoples that integrated or allied with the incoming Normans. I don't necessarily agree that the CTS6621+ Walsh's had to be archers and foot soldiers. The two earliest known Walsh's in Ireland were mounted soldiers or what I call knights. They were also "dear relatives" of some of the true Norman Marcher Lords.

My second cousin, Dennis, the true genealogist thinks we come in to play here,
'There is an interesting statement which I believe comes from "Notes on the Ecclesiastical Remains at Runston, Sudbrook, Dinham, and Llan-bedr", by Octavius Morgan and Thomas Wakeman, 1856. In a discussion of a fellow named Guy de Sancto Wallerico and Dinan/Dinham castle, it mentions...
"Guy may have had another younger son or sons (besides a son named Reginald), from whom perhaps descended a family whose name we find written Le Walleys, Wallens, Walshe, and finally after some other variations of orthography, Welsh, who were seated here and at Llanwern down to the early part of the XVII century-they bore the same arms as those given by the Heralds as the coats of Sancto Wallerico, viz: Ermine, a bend sable, and the corruption of Wallerico to Walsh is no more strange than many others that could be named. However this may be, two of the family accompanied Strongbow to Ireland in or about 1170; these were David Le Walleys, and Phillip Le Walleys, younger sons of Ralph and brothers of William Le Walleys. It may be right to observe that Dinham, and Llanwern were mesne fees, the former certainly, and probably the latter, at that period held under Strongbow, (Gilbert de Clare) as Lord of Chepstow. It is probable that at one time Dinham was their principal seat, as several of them are found described as of this place, thus Adam De Dinham in the reigns of Henry III (1216-1272), and Edward II (1307-1327). William De Dinham, or William Le Walsh in the Reign of Edward III (1327-1377). Christopher Welsh, who was High Sheriff in 1569, was the last of the family who held Dinham'.
http://www.castlewales.com/dinham.html

I wonder if the CTS6621 were some kind of Brit who allied with the March Lord regimes early on as Robert FitzHamon conquered the area. There are some others L513's (not this type) with names like Ross/Rose and Massey/Massie that appear to be true English at the time although not necessarily Old Anglo-Saxon... but even the Norman surnames creep in their stories.... and they are related in their story as they are in their genes. All I can figure is that everyone needed a Norman surname or association at the least.
It literally paid off, perhaps as land in Ireland in some cases.