No announcement yet.

Viking FGC23343

  • Filter
  • Time
  • Show
Clear All
new posts

  • #46
    I guess there's room for doubt about this conclusion because, just based on the visible differences between haplotypes, Group B is only just barely more remotely related to Group A than Donor C is to Donor A. Like only 100 years more remotely, not the ~350 that I concluded.

    I figured the reason companies don't offer clear guidance about predicting relationships past GD of 6 at 67 is because of an increased chance of masking mutations when the number of mutations >10% of the total number of markers. That is, the larger mutations are as a proportion of the number of markers, the higher the chance that the next mutation over-writes an existing one, or mimics the haplotype of a cousin's lineage.

    So I calculated the probability distribution of X number of masking mutations for each generation's removal from the common ancestor, and recalculated a confidence interval for the relationship accordingly. That's reflected in the figures I quoted in my last post.

    This narrow, 15-year overlap in the 50% to 95% confidence intervals of the B-A and B-C comparisons that I talked about was only enabled by this assumption of a number of masking mutations. In other words, I'm talking about an only 5% scenario. Very tight. Not implausible--one in twenty are not crazy odds. But very tight.

    To me, this suggests a fairly specific mutation scenario, where the ancestors of the Group A experienced 3 mutations which coincidentally mirrored those experienced by Group B.

    I'm thinking these were probably DYS390, DYS464C and DYS 476, experienced by Group A in the years after they branched off from Dorey but before they branched off from one another--roughly between 1350 and 1600 A.D. A little fast, but plausible.

    The time frame for Group B experiencing these same mutations might be inferred by both members' identical genetic distance to Donor C, roughly between 1000 A.D. and 1600 A.D.

    I wish there were a wider margin of error in this case, but this is just the way the numbers line up. I guess this is why the company doesn't express much confidence about genetic distances in this range.


    • #47
      Here is a curious document that may be of particular interest to the FGC28370 people.

      It seems to be a very large pedigree of the d'Aubigny family whom I've discussed several times in relation to the early history of the Garnett and Dorey families.

      The Garnetts definitely descend from them in a female line (depicted here), but their heraldry suggests they may also descend from them in the direct male line. The earliest record of a Dorey family, spelled with a "y", relates to some lands held in the neighborhood of Valognes in the 11th century by the Garnetts' connections, including the d'Aubignys at Huberville.

      I can't vouch for the document's accuracy, but it is very interesting that one line depicted (pages 4 & 5) shows a Ralph d'Aubigny born on Jersey (just like the earliest confirmed ancestor of the Doreys). Among Ralph's descendants were the Hungerford family of Farleigh, whose estates included Tellisford in Somersetshire--the earliest known residence of the Swift family, who form the single largest cohort within FGC28370.

      I could go on and on about the coincidences, but this Ralph's 3X great grandmother was a Sibylla de Valognes, hearkening back once again to Dorey's family history.


      • #48
        As far as I can tell, the earliest known ancestor of the Swifts was a London merchant named Flower Swift, born in the late 16th century at Tellisford or the adjacent parish of Norton St Phillip in Somerset. Both places were part of the estate of the Hungerfords of Farleigh who married into the d'Aubigny family as per the link in my last post.

        There is some speculation as to whether the unusual given name "Flower" reflects a connection to the Flower family of Norton St Phillip, who seem to have held a lease on the manor grange and been successful enough to entertain some pretensions to being gentry. One member of this family, Jeffrey Flower, "gent.", had a small armorial plaque near his tomb in Bath cathedral.

        From the mid-1500s at least there was an industrial cloth mill at the adjacent Farleigh Hungerford. Could this have been the line of business that the first Flower Swift was engaged in?

        It seems Tellisford--and maybe some of the other surrounding manors?--came into the possession of the Hungerfords in the late 15th century, right around the time of the marriage with the d'Aubigny family.

        Could the Swifts descend from an illegitimate son of this branch of the d'Aubignys? Commerce would have been a typical career path for other non-inheriting sons of the gentry in those days.


        • #49
          Originally posted by benowicz View Post
          Here is a curious document that may be of particular interest to the FGC28370 people.

          It seems to be a very large pedigree of the d'Aubigny family whom I've discussed several times in relation to the early history of the Garnett and Dorey families.. . I can't vouch for the document's accuracy. . .
          I've had a chance to kick the tires a bit and I see some problems with this document.

          The consensus of academic historians is that there were two distinct d'Aubigny families in England immediately after the conquest--one headed by a fellow nicknamed William 'Pincerna' (i.e., "butler") and another headed by William 'Brito' (i.e., "the Breton"). It's not an unchallenged consensus--some have favored the interpretation that they were two branches of the same family, or even that the two Williams represented one and the same person.

          Inconveniently, the Garnetts descend from the 'Pincerna' family and the Jersey connections relevant to the Dorey family belong to the 'Brito' family.

          I'm a little weirded out by the coincidence of the Tellisford connections of the 'Brito' descendants, though. As I mentioned in an earlier post, Tellisford definitely belonged to Geoffrey de Mowbray, bishop of Coutances, at the time of the Domesday book.

          The passage of the bishop's estate to his nephew Robert de Mobray, and through his widow to the d'Aubigny 'Pincerna' descendants of her 2nd marriage is clear.

          I think the pedigree in that first link may simply be too ambitious. I think the author may have just speculatively tried to incorrectly force in the Elizabeth d'Aubigny who married William Botreaux into the 'Brito' family. Tellisford and Petherton, the main seat of the 'Brito' family in Somerset, are at opposite ends of the county, and the other marriages in these lines seem otherwise to bear more geographic coherence. I believe that Tellisford did come to the Hungerfords from a d'Aubigny 'Pincerna', but that her precise parentage is not known.

          Unfortunately, even if this is the case, this still leaves the connection to Jersey and the Dorey family unresolved.


          • #50
            Originally posted by benowicz View Post
            I've had a chance to kick the tires a bit and I see some problems with this document. . .
            Though maybe the problems are not completely insurmountable. There is a very interesting piece of circumstantial evidence linking the 'Pincerna' and 'Brito' families. William 'Pincerna' and William 'Brito' definitely were married to sisters--Maude and Cecily Bigod--suggesting that perhaps 'Brito' and 'Pincerna' were close relations as well, maybe paternal cousins. Both seem to have been born around 1100 A.D.

            Much is made of tracing the 'Pincernas' to St Martin d'Aubigny in La Manche, and the 'Britos' to St Aubin d'Aubine in Ille-et-Vilane, but this is a little bit like putting the cart before the horse. Just as often, maybe more often, the places are named after the family, not the family after the place.

            These places would have fallen within different feudal jurisdictions in the 11th century, but that's doesn't necessarily mean anything, either. The 'Pincernas' are often speculatively made a branch of the Neel family of Saint Sauveur in La Manche, but it is an historical fact that several members of this family spent a number of years in exile in Brittany. The Saint Sauveurs even gave the characteristically Breton name 'Eudes' to several of their members. It is easily possible that one branch of d'Aubignys from La Manche permanently set up shop in Ille-et-Vilane.

            There is not, I think, enough information known about the early 11th century origins of the 'Pincerna' or 'Brito' families to preclude them sharing a common agnatic origin.

            A scenario reconciling these findings could be as follows:

            The 'Brito' family acquired Tellisford from their 'Pincerna' cousins. We've already seen such a swap for geographic convenience with regard to the de Montgomery family lands in Montaigu-la-brisette and Lancashire.

            Elizabeth d'Aubigny Botreaux may really have been a 'Brito', and brought Tellisford to the Hungerfords upon her marriage, and the Swifts may descend from an illegitimate son of one of her brothers or uncles.

            This would make the Swifts close relatives to Sir Phillip d'Aubigny, warden of Jersey in the 13th century, leaving plenty of scope for a direct relationship to the Dorey family.

            The observed genetic differences between the Swifts, the Garnetts and Dorey are probably consistent with this hypothesis, although they don't leave much margin of error.


            • #51
              Originally posted by benowicz View Post
              Here is a curious document that may be of particular interest to the FGC28370 people.

              It seems to be a very large pedigree of the d'Aubigny family whom I've discussed several times in relation to the early history of the Garnett and Dorey families. . . .
              Well, I guess this is off. By 1356 at least, Tellisford was in the hands of a family called de Pavely. No chance of proving a Swift/d'Aubigny connection through this tree.

              Too bad. It really seemed like there was a solution at hand.


              • #52
                Without completely dismissing the other theories I've worked on here, I think there is another, more statistically plausible scenario that I never even considered: That the true stock to which all these different surnames belong is actually the Edgeworth family first noticed in Gloucestershire during the Domesday Book.

                I was unable to resolve the network of relationships between identified clusters (i.e., Knuckles, Swift, etc.=Group A; Edgeworth/Garnett=Group B; and Group C=Dorey), which show Group A almost equally related to Groups B & C, although Group B & C show a likely relationship maybe 300 years later to one another). My problem was likely that I was moving in the wrong direction--I wanted to avoid any assumed relationships that were below the 50% confidence interval given resolution and observed GD.

                If instead I work downwards towards a single MRCA date which applies between all groups equally, I get an likely compromise of 25 generations or MRCA about 1250 A.D., with a confidence level of about 70% between Groups A & B, and 25% between Groups B & C. Together they average out to nearly 50% and imply no scenario nearly as restrictive as the crazy 5% scenario I had been complementing early to force the Garnett theory to work. This is far more satisfying.

                And not just from a numbers perspective. The variety of surnames within Group A could suggest that its geographic locus around Bath, England represents the cradle, so-to-speak, of the wider kinship group. The name Edgeworth first appears in the 11th century not too far away, and was still the primary home of this name into the late 19th century.

                The very close relationship between Edgeworth and Garnett is interesting, but probably not too surprising. The earliest documented ancestors for these families lived around 12 miles apart in the late 16th century, on the border between Denbighshire and Cheshire.

                Unfortunately, the implication of this theory would be that our Dorey sample from Normandy more likely reflects relatively recent migration from southern England, and therefore does not help us discover the Edgeworth's Norman homeland.

                There is a vague geographic coherence to the early associates of the Edgeworth family, including the great de Lacy family, that suggests their roots are to be found in the south or the south-west of the Cotentin peninsula, near the border between La Manche and Calvados. Maybe the place name 'Le Mesnil-Herman' is a clue--the earliest known Edgeworth was named Herman. But I have no idea how I could build on this.

                Knowing whether the Scots Henderson and/or Chalmers are also FGC28370+ could be helpful in filling out the Edgeworth ancestors' route from (presumably) Scandinavia to Normandy. A lot of the Cotentin was settled by arrivals from Scotland and Ireland, but they're best attested further north on the peninsula. There was also apparently some small contingent in the Cotentin who arrived in Normandy from the east English Danelaw--though they mostly seem to have settled much further east in Normandy.


                • #53
                  Following the Edgeworth lead is taking me in a direction I didn't expect. Decidedly non-viking. I don't know how to sync all of this up. But I started here.

                  Those references that I've followed seem to be good, and they may have led to tracing this line back another one or to generations to a town called Dreux, in the department of Eure-et-Loir. Strictly speaking, I don't think it is part of Normandy, although in the 10th and 11th centuries it fell within the political orbit of both Normandy and the kingdom of France.

                  The key seems to be nailing down the identity of Herman de Edgeworth.

                  The first completely securely identified ancestor, as far as I know, was a Sir Peter Edgeworth, who served as sheriff of Gloucestershire in the 13th century. Documents referenced at the first link can be found on publicly funded heritage sites, so I'm pretty sure the author is correct to be confident that he has traced the family to Peter's father, one William in the manor of Wick aka Painswick, from a document dated 1137/1138.

                  If the transcriptions of that document are correct, William's father is given as "Herem.", which some sites have assumed is an abbreviation for the very rare name "Herembald", but more assume is "Herman".

                  Given that a fellow named Hugh Puher appears right next to him in the document, it seems reasonable to interpret it as "Herman", since the Domesday book and subsequent transactions show Hugh's father, Walter "Pontherius" and Herman holding interests in lands adjacent to one another around modern day Himbelton.

                  Here's where it gets interesting: As the author of that original linked document states, there are a number of other entries for men named Herman in the area. It's not clear if these are the same man or two or more different men.

                  Places in the Domesday Book associated with the name Herman

                  In several of these entries--although not related to any of the specific places associated with the Edgeworths--the Herman bears the surname "de Dreux".

                  As it so happens, Herman de Dreux appears on several versions of the list of companions of William the Conqueror at Hastings, along with an "Amauri" or "Amaure" de Dreux--there seem to be some variations in this list.

                  "Amauri" or "Amaure" are almost certainly phonetic spelling variations of "Almeric"--the exact same name as one of the grandsons of Herman (Edgeworth?) of Shell, Worcestershire from the Domesday book, per that history of Himbelton parish linked above. Neither name is totally unique, but in combination, they do give a strong impression--if the identification of Herman at Shell with the Edgeworth family is correct--that we should be looking for the origins of the de Dreux family.

                  As it turns out, the name Amaury de Dreux does appear in the genealogy of a family instrumental in some ecclesiastical foundations in Normandy. This Amaury's grandfather, Baudry aka Baldric, son of Gaudo aka Wado, could well have been the father of either Herman or our Amaury de Dreux, or both.

                  All circumstantial, but maybe reinforcing this set of associations is the fact that back in England, an Almaric de Dreux (almost certainly Wm I's companion) and Durand, sheriff of Gloucester, apparently combined to unjustly deprive a fellow of his property in Chedglow, Wiltshire around the time of the Domesday Book.

                  Particularly interesting about this is the fact that Sir Peter Edgeworth held the office of sheriff of Gloucestershire later, in the early-to-mid 1200s. And that some researchers say that Durand's family came from Pitres on the Norman side of Eure, across from Dreux.

                  Interesting, but why would Herman's descendants not simply retain the surname de Dreux if it had already been established?

                  Also note: The de Dreux I'm talking about are NOT to be confused with the cadet branch of the Capetian dynasty founded in the 12th century--too late to have any relationship.


                  • #54
                    Some additional information that could add a bit of color to the relationship between the early Edgeworth/Dreux family and their de Lacy overlords: One of the foundations to which Guado aka Wado de Dreux donated was the Abbey of Saint Evroul--re-founded by William FitzOsbern.

                    The de Lacys acquired some of their early English properties from FitzOsbern, leading a few researchers to hypothesize that de Lacy was a client or captain of FitzOsbern's affiliation. I guess this makes sense given the de Lacy family's humble origins.

                    So I guess a plausible scenario would be that the Edgeworth/Dreux family were vassals of FitzOsbern both in Normandy and in England, and that when FitzOsbern's son forfeited after his 1075 rebellion, lordship over his vassals was granted to de Lacy, in consideration for his assistance in putting the rebellion down.


                    • #55
                      Originally posted by benowicz View Post
                      I've had a chance to kick the tires a bit and I see some problems with this document.

                      The consensus of academic historians is that there were two distinct d'Aubigny families in England immediately after the conquest--one headed by a fellow nicknamed William 'Pincerna' (i.e., "butler") and another headed by William 'Brito' (i.e., "the Breton"). . . .
                      Well, it seems both theories are back in contention. That is, that the currently identified clusters of FGC28370+ can be attributed either to the de Dreux or the d'Aubigny families. Some researchers have clarified that the Lord Daubeney family do NOT descend directly from William 'Brito', but rather from a nephew of his.

                      They seem to take the view that the relationship is through a brother or sister of Brito, rather than through a brother or sister of his wife, Ceclia Bigod, though I can't quite see why this must be, given the nature of the evidence offered.

                      If the Lord Daubeney family descended instead directly from William 'Pincerna' I (his eldest son also bore that epithet), this would satisfy the objections of those authors (I think), and leave open for a d'Aubigny explanation for the Garnett, Swift and Dorey matches.

                      Both theories have their problems. From what I can tell, Tellisford was occupied by a family called de Pavely in the mid-1300s, though I guess the d'Aubignys may still have retained ownership. And the closer I look at the Edgworth pedigree, the more frustrated I am by my inability to account for the generations linking Francis Edgeworth b. ~1560 to Sir Peter Edgeworth fl. 1221. In fact, the heraldry associated with the what I believe is the donor's line seems to put them very near "Edgworth" in Lancashire from the early 1500s at least, and much closer to the orbit of the Garnetts of Lydiate and Warrington who appear to have claimed agnatic descent from the d'Aubignys at one point.

                      If a coherent set of TMRCAs can be worked out.


                      • #56
                        Originally posted by benowicz View Post
                        Some researchers have clarified that the Lord Daubeney family do NOT descend directly from William 'Brito', but rather from a nephew of his. . .
                        I'm not sure if this link below provides much proof one way or the other, but for the record, the heraldry of Brito's proven descendants (i.e., the so-called Belvoir line) does not match that of the Lord Daubeny (aka "Ingleby" or "Jersey") line.

                        While the Lord Daubeny line doesn't contain the "gules, a lion rampant argent" motif that underscores the Garnett family's d'Aubigny connections, either, that is possibly explained by the fact that that branch of the d'Aubignys had adopted the de Mowbray surname (and arms) on succeeding to their estates.

                        So I guess it's still a toss up, though the question had to be asked.


                        • #57
                          Just found some fascinating new information which I think supports the hypothesis that the currently identified FGC28370+ cluster has a viking origin through the d'Aubigny family of the Cotentin--although indirectly, leaving room for challenge.

                          I called a group of high-resolution haplotypes I "the Swift group" (aka "Group B") because the many surnames appearing in it are bracketed by two donors who trace their origins to a merchant family named Swift from the Somerset/Gloucestershire areas in the late 1600s.

                          One of these, TJM9P, seems to descend from the fellow named "Flower Swift" I mentioned earlier, a merchant from Tellisford, Somerset who did a lot of business in London. According to this researcher, my suspicion that they were in the cloth fulling industry at Tellisford seems correct. This is an important detail.

                          The other, 4Q7H9, descends from one of the original settlers of the Quaker colony at Philadelphia--quite independent of the "Flower Swift" line who arrived in Baltimore and spread southwards.

                          The tree that donor presented can be carried back two generations to a Dr. Samuel Swift of Moreland township, a member of Benjamin Franklin's circle of friends, and back from there to a John Swift, a justice of the peace noted for his involvement in the Keithian controversies that challenged the proprietary party in early Pennsylvania.

                          However, the information most relevant to the English origins of this family probably comes from a related line of Swifts, settled at Croydon Lodge on the Delaware River, about twelve miles away from Moreland: their coat of arms. See page 7 of the link below.

                          Now, the relationship of this new line to 4Q7H9 is not explicitly stated, but in the context of everything we know seems most likely. In addition to the Croydon Lodge line being a merchant family from Bristol, Gloucestershire, they, like the Dr. Samuel Swift line, intermarried with the Matthias Keen family.

                          Anyhow, the arms born by this family, "or, a chevron barry nebulee argent and azure between three roebucks courant proper", is that of the family of "the rich merchant Swift" originating in Rotherham, Yorkshire.

                          Probably not coincidentally, from 1561 at least, the Rotherham family had a branch at Northleach, Gloucestershire (aka "Norlache"), an important center of the cloth fulling trade. See page 364 of this link.

                          The author of that piece contests the claimed gentry origins of the Rotherham family, and he may or may not be correct about that. However, it does, if I am correct in linking these families, move the center of gravity of these matches away from the southern homeland of the de Dreux/Edgeworth family towards the northern base of the Garnett/d'Aubigny family.

                          We've already been incredibly lucky to get this far, but it would take an extra-special amount of luck to find the specific MRCAs between the Garnett/Edgeworth and Swift groups. So, for the time being at least, we may have to content ourselves with the general observation that, as some historians have claimed, William the Conqueror pursued a specific policy of settling the northern counties with Normans from the Cotentin, presumably because their experience in that remote, border province better prepared them to deal with the Scots and unruly native elements.


                          • #58
                            This was unexpected, but here is another piece of circumstantial evidence that may support the theory that the currently identified FGC28370+ people may have a viking descent through a cadet branch of the d'Aubigny/de Mowbray family: Neel, vicomte du Cotentin, who frequently appears in lists of the companions of William the Conqueror, settled very near the Garnett family in England.

                            Places in the Domesday Book associated with the name William (son of Nigel)

                            Nigel, of course, is a Latinized version of Neel, and this William Fitz Nigel appears to be the son of a vicomte du Cotentin.


                            The widely repeated theory that the d'Auginys of Saint Martin d'Aubigny are a branch of the Neel vicomtes du Cotentin appears to be based primarily on their shared use of this unusual name, but is not unreasonable, as theories go.

                            But it's also interesting that the arms ascribed to their English barony of Halton appear to be a variation of the Lord Daubeny arms discussed elsewhere (i.e., gules four lozenges in pale, instead of in fess).

                            It may not be immediately relevant, but it is also interesting that this Garnett DNA donor mostly likely descends from the William Garnett who married a daughter of Sir Thomas Gerard--and lived in Warrington, where William Fitz Nigel owned a lot of land.

                            Last edited by benowicz; 11 April 2018, 11:26 AM.


                            • #59

                              For giggles I ordered FGC28370 from YSEQ. I'm not sure how long it will take but I'm guessing it'll be several weeks to get the results.

                              Curiously, I checked y-search for the first time in ages.
                              My best match (although not very good) is a Thacker from Somerset.

                              7ZQAU Thacker Somerset, England R1b1a2 (tested) Family Tree DNA 94 17

                              Secondly, have you seen anyone named Gendron or anything similar during this exercise. I ask as I've just gotten my 1st and only 111 marker YDNA match on ftdna from someone who's name is Johndro (with other spellings documented). My only guess as to how we're related is via my Gendron matches as we all lived in Quebec at some point. How we have different surnames with a reasonable probability of a more or less recent ancestor I've not yet tried to figure out.


                              • #60
                                Another piece of circumstantial evidence that leans closer to the currently identified core FGC28370+ group possibly representing cadets of the Neel vicomtes of Saint Sauveur:

                                Originally posted by benowicz View Post
                                Another detail that I think lends weight to the Garnet/d'Aubigny interpretation of one branch of FGC23343: Y Search 46CE6, another closely matching profile, seems connected in some way to the Gerard family of Bryn, Lancashire.

                                I'm not the first person to make that observation. I saw it in an old social media post by a person I believe is the donor for the Edgeworth entry. But I did encounter a very detailed documentary account that I think goes a long way to closing some of the gaps in the versions of the Garnet of Haughton/Bunbury trees I've seen to date.

                                The Gerards were closely intermarried with a branch of the Garnets of Lydiate located at Warrington since the early 1415, when a William Garnet married a daughter of Sir Thomas Gerard. Probably a son or grandson changed his name at some point for an inheritance.

                                I'm pretty sure this doesn't represent the Gerard family genetic signature, as they're clearly a branch of the same patriline that gave rise to the FitzGeralds in Ireland, and for all the various haplotypes that appear in that DNA project, none of them appear to be FGC23343. . . .

                                Revisiting the pedigree of the Gerard family, I see that the version of their descent from the FitzGeralds in Ireland has been falsified. The earliest records give their original arms as azure, a lion rampant ermine ducally crowned or--identical to those of the Garnetts except for their tincture and a bordure, typical of the ways cadency was indicated in early heraldry. Page 23.

                                That author hints that there is documentary and further heraldic evidence suggesting the Gerards--and now we can presume the Garnetts as well--were cadets of the de Montalt family of Hawarden in Flintshire, whose name was often rendered "de Mohaut".

                                The earliest known ancestor of the de Mohauts was a fellow called Ralph fitz Norman, "dapifer" (i.e., steward) of Chester. I don't think I mentioned it in my last post, but the backstory behind the settlement of Neel du Cotentin in Cheshire was that he was not granted his lands there by William the Conqueror, but instead received them as tenant-in-chief of his cousin, Hugh d'Avranche, Earl of Chester. So there is definitely a family connection here.

                                In fact, the de Mohauts intermarried with the family of the Neels.

                                The genetic distance between the Gerard and Garnett and Edgworth samples is closer than I would expect if these donors had a MRCA in the 11th century, but Gerard did only test to 37 markers.
                                Last edited by benowicz; 11 April 2018, 04:00 PM.