Viking FGC23343

[benowicz]

Regarding the other hypothesis to explain the German examples of FGC23343+, Wikipedia has a great page that links to some interesting accounts of specific auxiliary units in the Roman army.

https://en.wikipedia.org/wiki/List_o...iary_regiments

To me, the hypothesis surrounding the Oleron recruits in de Turenne's forces during the Thirty Years' War still seems a little tidier because the skeletal unit histories presented in that link don't show one obvious chain of events linking Rheinbach/Ahrweiler (Schoenenberg) to Wallenfangen/Saarwellingen (Leinenbach) through the Romans.

The relevant Roman units would have originally been raised in Aquitania--but with later multi-national transfers/recruits. There's not really much sense of ethnic cohesion in these units maybe beyond the first few years of their existence, despite their names. Most of the available evidence links them to camps in Germania Superior, mainly on the east side of the Rhine, around Frankfurt, with some satellite camps further south in the Frankish region of the modern state of Bavaria. Obviously there could have been excursions or subsequent migrations significantly further to the west to Saarlouis or up north to the Bonn area, which I believe would then have been considered part of Germania Inferior.

Every time I have to add a new element, the scenario just appears less convincing, even though the evidence I'm working from is so thin that almost no variant can be regarded as outrageous.

I think I gave a reference earlier to the use of Basque auxiliaries to put down a rebellion in Batavia/northern Germany in 69 A.D., but the applicability to FGC23343+ is more questionable. They seem to have been recruited by Galba when he was governor of Tarragona, which is pretty wide of Poitou/Oleron, in eastern Spain, just outside of what is considered Basque country today, I believe.

https://books.google.com/books?id=hH...s%2069&f=false

[AndrewRoss]

Have you tried SNP Tracker?

http://scaledinnovation.com/gg/snpTracker.html

When I put in R-FGC23343, it showed that haplogroup dated to the Bronze age along the southeast coast of England.

The Bronze age was between 1700 BC to 500 BC.

That's more than a 1000 years before there were Vikings and of course they originated in Scandinavia.

[benowicz]

I'd be interested in knowing who compiles SNP Tracker data and their methodology. I suspect it is self-reported and not too reliable, for reasons that will be obvious when you read this reply. You seem to have a mistaken idea of the current state of my research depicted here.

While it's true that the 50% confidence intervals of the YFull and Big Tree data I've worked with place the most likely founding date at about 500 B.C., that's at the extreme edge of your estimate of the Bronze Age, and I have no doubt the expansion of FGC23343 occurred considerably later. Just when is a guess, but to say that FGC23343 dates to the Bronze Age is a bit of a distortion.

However, I'm just about 100% certain that FGC23343 did not originate in South East England. I understand how that mistake could have been made, though--a large number of closely related FGC23343+ donors, belonging to a sub-grouping known as FGC28369+, have their earliest known ancestors in London in the 17th century. But subsequent research has made it clear that they are only part of a single lineage who trace their ancestry to a well-known figure of the post-Conquest era, Vivian de Garnett, whose family came to Lancashire in the retinue of Roger 'the Poitevin' de Montgomery, comte de la Marche. To date, as far as I know, all verified (vs suspected) English FGC23343+ donors come from this lineage.

This development squares quite nicely with the distribution of the ancestral clade, ZZ40, which peaks in the Basque country of northern Spain, as there is a wealth of historical evidence for Basque or proto-Basque place names southern and western Poitou ('Poitevin' means person from Poitou). To be frank, the linked map at SNP Tracker makes no sense to me, as the distribution of ZZ40 is well known and falls nowhere near where depicted by SNP Tracker.

It is true that there is an unusually large number of surnames associated with such a young sub-grouping as FGC28369+, whose founding ancestor estimated at the 50% confidence level to have been born in the 13th century. But by far there the largest number of men in this group are surnamed Garnett, and the dating of STR haplotypes makes it clear to me that this should be considered the founding surname. Another subset of very closely related FGC28369+ donors have very good pedigrees among the gentry families of Lancashire and Cheshire from the 16th century, and have well documented financial and family relationships with the historical de Garnetts, suggesting some specific NPE scenarios, although proving any specific one of them conclusively is probably not possible. So I have little doubt about this conclusion. I know lots of donors CLAIM to be able to trace their lines to Conquest-era nobility, and that usually they are mistaken, but in this case it really is for real--two of them are easily able to trace their lines forward from the Burke's series. This is for real.

The first posts in this series did indeed posit the idea that FGC23343 had a clear Viking origin, but that should only be considered an original working hypothesis based on information available at that time, and since then it has been radically revised and improved.

That said, there still could well be a Viking 'element' of sorts within FGC23343, despite its ultimate origins in the proto-Basque regions. There is at least one donor with a solid pedigree going back to Shetland, and the Saintonge region of Poitou was the scene of one memorable incident in the Viking wars. In 845 A.D., a band of Vikings based in Ireland and the Scottish islands assisted a rival claimant to the kingdom of Aquitaine, Pepin II, and temporarily settled at Saintes. They were decisively defeated shortly thereafter, and their retreat through Ireland, Orkney and Norway has been archaeologically documented through a series of coin cashes dated precisely to these years.

It will probably be impossible to specifically connect this particular instance of FGC23343+ to the Viking era, but it is not implausible, and really should have been expected. There is solid archaeological evidence that merchants from Saintonge sold pottery in Ireland and the western islands of Scotland well before the Viking era, and the consensus among professional historians is that during the Middle Ages almost all pottery in Ireland and all wine in both Ireland and Britain until the early modern era was imported from the Saintonge area.

I would caution you against conflating 'Viking' with Scandinavian. Without a doubt the cultural impetus for the Viking culture came from Scandinavia, and almost certainly the vast majority of so-called Viking lineages should be expected to belong to haplogroups common in Scandinavia today. But as oxygen isotope data from archaeological sites in Ireland and Scotland shows, the Vikings were an eclectic lot, and clearly recruited men into their warbands from among those they encountered during their voyages. To date there is no specific evidence of this with regard to FGC23343, but it does appear more plausible than some other scenarios, such as taking slaves from Saintonge to Shetland. Historians considering the overall Viking project portray raiding and trading as two sides of the same coin, with slaves captured in Scotland and Ireland exchanged in markets in Spain, the Mediterranean, or the Carolingian empire for exotic luxury goods not available to them at home.

Other than these two groups, English FGC28369+ de Garnetts and the Shetland family, the other observations of FGC23343+ donors fall into what I consider three broad categories--a few donors living in France today, a couple with very recent origins in the German Rhineland, and a bunch of Americans whose European origins are not well documented. The men within each category may not be particularly closely related to one another, but their earliest known ancestors do share some characteristics which point to some broad hypotheses which can plausibly link them to the western Poitou area, and the island of Oleron in particular. Short of some extremely close matches with several donors with recent, well-documented connections to Poitou, there is probable little chance of proving these hypotheses, but in context they appear to be far more likely than any alternative explanation I've encountered to date.

The fact that there haven't been more FGC23343+ donors identified to date is probably due to legal restrictions on the marketing of the necessary tests in France. From what I gather it's not a law that is meaningfully enforced, but it still has had a significant impact in discouraging purchases.

Each of these donors presents an interesting history with lots of possibility, but with regard to the ultimate origins of FGC23343, the most useful evidence to date has been from the de Garnett/FGC28369+ subgroup and the simple fact that ZZ40 is their ancestral clade. There may well be a 'Viking' element to FGC23343, but if so it is distinctly in the minority here, despite the impression conveyed by early data received. It almost certainly originated or expanded in the region around the island of Oleron in France.

[benowicz]

It occurs to me that it might be useful to make some general comments comparing the sort of atomistic approach to research I've taken here w/ regard to FGC23343 and the typically statistical methods employed to analyse the history of other, larger haplogroup subclades.

To be blunt, there are simply not enough observations of FGC23343+ to even consider a statistical approach. To my knowledge, to date there have been only 11 distinct FGC23343+ lineages identified, defined as clusters of donors whose SNP status or STR haplotype data suggest they likely share a most recent common ancestor born within the last 1,000 years. In these conditions, there is no hope of obtaining a statistically significant sample size.

There are two mitigating circumstances particular to FGC23343, however, completely idiosyncratic and without any broader theoretical application to other small clades, and yet completely valid in and of themselves:

1. One of those lineages is well represented, and just by dumb luck happens to have been founded by an individual with a solid documentary trail, Vivian de Garnett.

2. FGC23343's ancestral clade, ZZ40, is so unusual with respect to the historical milieu of the founder (i.e., Conquest-era Lancashire) that the number of plausible alternative hypotheses is severely restricted.

The first point was not perfectly clear when I undertook these analyses, and only became obvious when a decisive number of diverse haplotype donors bearing the Garnett surname published their test results. Maybe other researchers knew or expected this development before I did, but I didn't, and as a result I applied the next-best analytical framework available to me when developing my hypothesis, namely geographical distribution of the earliest known ancestors.

I was aware from the beginning that this was a kind of methodological mismatch, in the sense that geographical distribution can only be relied upon to deliver accurate results when the sample size of distinct lineages is large enough to support a statistical extrapolation. Clearly I had not satisfied that requirement, but absent any other data, it was *relatively* the best method to construct a working hypothesis around. But such is the nature of research itself, an iterative process of hypothesis, test, failure, and re-formulation until the gap between the data and the explanation is narrowed to a satisfactory degree.

Only a very small portion of FGC23343+ lineages will share similar histories through the modern era, but this location in early Saintonge/Poitou is still a very useful perspective from which to approach research for the others. For example, searching for historical narratives linking Poitou to the home places of the currently identified German FGC23343+ donors, Leinenbach and (probably) Schoenenberg, did return a satisfyingly coherent explanation in the mid-17th century military campaigns of Henri de La Tour d'Auvergne, vicomte de Turenne and the largest landowner in the isle of Oleron. Now, this is far from conclusive proof, which can only be obtained by a large series of high-resolution matches with correspondingly high quality pedigrees, but I'm convinced that it is the *relatively* best explanation available, and is completely consistent with the hypothesis developed through analysis of the de Garnett analysis. No further refinement to the core hypothesis for FGC23343's ultimate origins seems necessary at this time.

There's kind of an apple-and-orange problem here, with regard to comparing the level of confidence available through these different research approaches, atomistic vs. statistical. That's kind of frustrating, but just a fact of life. My preference is to always employ statistical methods whenever possible--in fact you could say that my reliance upon the unique character of ZZ40's distribution is fundamentally dependent on statistical analyses of that far larger grouping. But with regard to FGC23343 itself, the sample size is just too small. Luckily there was also an equally unique opportunity in the Medieval history of the de Garnetts.

[benowicz]

Here's an interesting essay about regional Medieval immigration into England from France that provides some interesting statistics illustrating, on an objective, quantifiable basis, just how ayptical or 'weird' FGC23343 is in comparison to native English clades:

https://onlinelibrary.wiley.com/doi/...468-229X.12924

An historian has estimated that as of 1440, there was a total of about 30,000 French immigrants living in England, which is equal to about 1.5% of the total population. Of that group, about 2% are thought to have been of Gascon origin. That makes Gascons about 0.03% of the total population. Apparently Gascons were unusual even among the immigrant population, being disproportionately represented by the merchant and gentry classes.

So FGC23343 and its exotic ancestral clade ZZ40 sticks out like a sore thumb in Britain, requiring a special explanation.

[benowicz]

Using the de Garnett/FGC28369+ plus group as a reference point from which to develop the historical context of FGC23343 in Britain and Ireland, I tried to identify all the notable Gascon families who are known to have settled there, at least temporarily, during the High Middle Ages. There really aren't many, and they don't create a confident impression of any kind of unified network. Here they are below:

de Forz earls of Albemarle: Originally lords of Oleron, one member became admiral to the English king Richard I, who arranged his marriage to the heiress of the earldom of Albemarle, whose lands in England centered on the Scottish borders and Lincolnshire. This line became extinct in the 13th century upon the death of the celebrated heiress, Aveline de Forz. Of all the notable Gascons identified, this family seems most likely to connect to currently identified FGC23343+ donors, although the evidence is very obscure and circumstantial.

There is an interesting entry in the Testa de Nevil that could possibly refer to the ancestor of one (suspected) FGC23343+ donor, Dorey, but there is a massive gap in the documentary record making this wildly speculative. The Testa reference is to Holland, Lincolnshire, but the name has been almost unheard of there for nearly 800 years, and the donor's family knows of no heritage outside of Guernesey in the Channel Islands--although at its widest extent, the de Forz patrimony also included the Isle of Wight.

There might also be a plausible, but very oblique connection to Vivian de Garnett, whose patron, Roger 'the Poitevin' de Montgomery had a niece cloistered at Saintes, where the de Forz family were hereditary provosts. St. Vivian was the patron of Saintes.

By comparison of 67-marker haplotypes, there appears to be some level of convergence going on, as Dorey seems unusually close to the overall modal, and therefore assignment to any specific sub-lineage is very unclear. But the Dorey signature is very close to some branches of FGC28369, especially the Swifts. Maybe, at a very optimistic estimate, Dorey branched off from FGC28369 around 1000 A.D., just before the earliest certain record of the de Garnetts.

Piers Gaveston, earl of Cornwall: This man, from an obscure background, gained fame through his scandalous association with the English king Edward II. There is an outside possibility that he left an illegitimate daughter who married a yeoman of the royal household, but his line effectively ended with his death.

de Vivonne of Somerset: There was a Hugh de Vivonne from Poitou who became constable of Bristol and acquired a lot of land in Somerset in the 13th century. His family survived in the male line for only a couple generations after Hugh's death. There is some speculation that his family may somehow have been connected to the de Forz line, but apparently only on the basis of the nickname 'le Fort' being current among the de Vivonnes; I'm unaware of any direct contact between the families.

de Foix earls of Kendal: Captals of Buch near Bordeaux, one member of this family served in the English army at Bannockburn, and more than a century later another was created earl of Kendal in Yorkshire. They maintained their claim to the earldom even after they transferred their allegiance to the kingdom of France when Gascony was captured, but made Bordeaux their permanent home. I'm unaware of any English descendants, even in the female line.

Gascoyne of Yorkshire, England, and Co. Cork, Ireland: The earliest origins of this family are obscure, but I am assuming some claimed connection to Gascony based on etymology. There are a handful of records of persons of this name among 14th century royal administrators in southern Ireland, but as far as I know, genealogies of gentry families of this name become reliable really only from the 17th century in Yorkshire. It's not clear to me whether this is the same family as the early colonists in Ireland, but this branches of this Yorkshire family do seem to have acquired land in the south of Ireland by the late 18th century.

Dardis of Co. Meath, Ireland: Descendants of Jenico D'artois, a talented Gascon mercenary captain of the early 15th century. No known connection to any of the grand families of old Gascony, and never numerous even in Ireland, but the given name Jenico does seem to have become moderately popular among the Anglo-Norman gentry of the Pale among whom his descendants married.

And that is all I am aware of. There are a couple of other incidents hardly worth remarking: (e.g., the spurious interpretation of the de Cliffords' founding ancestor, whose given name was Pons, being somehow connected to the great Poitevin family of de Pons, derived from a place rather than a personal name), and the temporary administration of the estates of the Irish earl of Ormond's estates by one of his Gascon colleagues after dying while on campaign in France). No really strong, sustained presence, just the sense that occasionally, some rare one-off made it big, but never left a lasting impression.

[benowicz]

Reviewing these English families of Gascon origin presents an interesting point by contrast with the single Shetland FGC23343+ identified to date--they all came to England through military service, sometimes, at least in the case of the de Foix, fighting AGAINST Scotland. Although obviously located in the modern nation of France, in the High Middle Ages, Gascony was ruled by the English kings, and from 1295 Scotland was always allied with the kingdom of France against England.

Pre-Jacobean Scots/English relations were complicated, with families constantly defecting across the border in either direction all the time. And Shetland only came to Scotland in the 15th century. But it does seem that the forces driving early Gascon emigration worked against Scotland as a destination. So we should probably expect relatively few incidences of FGC23343+ in Scotland, and those that do exist probably date from the Early Middle Ages or earlier. I guess that is a point, tentative and speculative as it is, for positing a "Viking" origin to Shetland FGC23343.

[benowicz]

Another point that just strikes me now--the de Garnetts may well be the ONLY Gascon family recorded in England before the accession of Henry II in 1154. That's not really surprising given the typical profile of the Conqueror's supporters, but it does underscore my expectation that this type of Y DNA signature should be very rare in England, even rarer in mainland Scotland.

The overall impression of Medieval Gascon immigration to Britain and Ireland is that it started very late, tapered off very early (i.e., after the loss of Gascony to France in 1451) and that even at its height was almost negligible, comprised disproportionately of nobles and merchants. I have read historical accounts of some Gascon sailors from more modest backgrounds congregating in ports of the south-west, but this precariat don't seem to be the sort of people likely to leave a large, enduring footprint in the genetics of the local population.

The apparently longer-lasting impact of early Gascons in Ireland should probably not be a surprise, though. The Gascons had a reputation as fighters throughout Europe, and they would have found no shortage of opportunity in Ireland. Plus, the studies I've seen regarding the distribution of the so-called sixth-century E Ware pottery in archaeological sites, presumed to have come from the south west coast of France, are heavily concentrated in Ireland, with somewhat fewer in the Inner Hebrides and Argyll, and the least in south western England. There's an enduring connection here that may be one of those really under-studied areas of pre-modern European history.

I think most population estimates of the Middle Ages show Ireland hovering around a third of the size of England, so we shouldn't expect, in absolute terms, an enormous number of FGC23343+ donors there, either, but I may have come across one, based on a partial haplotype analysis. There are definitely other, currently unexplained Z209+ folks in Ireland. The only evidence I've seen for a large scale Z209+ kindred expansion in Ireland would be the de Burgo/Burke/Bourke family. There are all sorts of speculative pedigrees that give them an origin in La Manche, Normandy, but on closer examination, the oldest authenticated record I can find of them is in Norfolk, England. There are enough donors from their phylogenetic branch, S21184+, in Scandinavia to make me think they may well have had a presence there before the Viking era, but they are significantly older than FGC23343+--estimated age of 3,100 years ago. I think I have also seen a couple of other, non-Burke S21184+ in Ireland.

If many more FGC23343+ donors turn up with verified ancestry in Britain and Ireland, I think they're most likely to come from 17th century Huguenot settlers.

[benowicz]

My first thought was that there was a very large incidence of NPEs within FGC23343, but on reflection that may not be so. Quite the opposite, it may actually be considerably lower than the population average. Or maybe it's the best estimate of the true incidence available. Here's my thought process:

Of the roughly 18 FGC23343 families I've seen to date (proven & predicted by haplotype), I think there is reasonable evidence of a change of surname for 50% of them since ~1400 A.D., which I think is the consensus, if arbitrary, date usually assigned to the commencement of hereditary surnames in Western Europe. Call that an NPE or not, but that's about 3% per generation, considerably less than the 10% rate I normally hear bandied about for NPEs.

FGC23343 is kind of a weird one because it's such a small group and because the current distribution of its donors doesn't reflect at all its most likely place of origin, whether that's due to marketing problems in France or whatever. But that oddness probably provides a unique window on the broader phenom of NPEs that would usually be masked by the relative local homogeneity of Y DNA signatures in Britain, etc.

[AndrewRoss]

Some of us really are descendants of Vikings.
It is very clear on SNP tracker.

No reason to doubt the site.


R-FT40329

R-FT95849

[benowicz]

Not saying that anybody is or is not definitely a viking descendant without any more proof than their own say so. But I've just given you tons of excellent reasons to doubt that site.

At least with respect to FGC23343. It's an incredibly small subclade. Nobody should expect that the EKA of the identified positive donors to date to be a terribly reliable representation of the subclade's ultimate origins.

On the other hand, their positioning of the center of ZZ40 is obviously way off. That one is a head-scratcher.

My guess is that they're just not putting a whole lot of effort into researching the origins of subclades. They're probably just dumping in un-tested self-reported data without any effort to qualify their findings. It's probably 'good enough' for certain well-represented clades--the ones whose distribution just coincidentally matches the typical customer profile for these tests. But France is definitely a special case, given legal restrictions on marketing.

In short: indefensibly way off with regard to FGC23343.

[benowicz]

You seem to have stumbled on to this thread without much background into this specific clade. You need to understand that FCG23343 represents an incredibly tiny pool of donors--less than 50 in total, and only about 11 distinct lineages, discovered only a couple of years ago. It's not like these monstrously large groups with tens of thousands of representatives who, by their sheer numbers can reasonably be expected to yield a statistically reliable geographic distribution.

If you come from a very ordinary, very common clade whose origins are well-supported and obvious, well good for you. But I wouldn't waste my time writing about something that was so simple and straightforward.

Luckily the FGC28369 represents a rare opportunity to legitimately leverage Medieval archives. It's hard, requiring a lengthy iterative process of hypthesis and re-evaluation, but it's lightyears better than acritically extrapolating from a base of 5 donors whose pedigrees don't go beyond 1600, as SNP Tracker seems to have done for FGC23343.

But even so, ZZ40's origins are so obviously in the Basque country that there's just no excuse for SNP Tracker's mistake there. Granted, on a relative basis, ZZ40 and its ancestor Z209 are far from the most common clades in the published databases, but those that are reported clearly cluster in the Basque country and not where SNP Tracker put them. There are at least several thousand of them.

[Biblioteque]

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1305903/

AndrewRoss: Okay, if you are a Viking, who in your family has Dupuytrens? LOL

[Biblioteque]

https://www.genealogics.org/

benowicz: I re-read your entire thread and links; and I appreciate your great research, some of which was an illumination for a few of my "cast of characters."

For reference (if you have not yet found him), the work of Leo van de Pas (deceased) is greatly respected and is being carried on by others. He previously did work for the Royals.

[benowicz]

Thanks for the recommendation, Biblioteque. It's good to cast a wide net when there is so much uncertainty surrounding a case.

Which obviously held true here. There were a good many 'aha' moments that turned out to be only coincidences until the true composition, depth and haplotype diversity of the de Garnett group became apparent. Now I feel there is a solid foundational core beginning to emerge in the study of this subclade.

There are still real mysteries remaining, of course. Given the historical bias, there may never be a true resolution to the question of the extent of "viking" FGC23343. If my "de Turenne's army" hypothesis for German FGC23343 is true, it happened recently enough that there is a real, if small, chance that these men turn up with robust matches to donors with deep family histories in Saintonge. In any event, there appear to be at least two distinct lineages of them, so it would be interesting to see whether they share any SNPs under FGC23343.