DNA trumps oral history?

+11 votes

Hi Cousins,

I was a member of Clan Macneil Association of NA for several years and bought the two books by former chief Robert Lister MacNeil, one using the Irish Annals and oral history to show the descent of the Macneils of Barra from Niall of the Nine Hostages. 

Recently (in the past couple of years) a DNA study was done of male Macneils on Barra and found they they did not carry King Niall’s markers. Instead, they found Viking markers, disproving, it seems, the claimed descent.

Kind of a downer in a way. We Americans of Scottish descent seem more attached to the romance of our heritage, the facts be ‘darned.’ I know there has already been a large impact on genealogical research by DNA. Has there been a reaction against the “proof,” given that DNA is not always certain? 

Wandering thoughts late at night with Jimmy Fallon keeping me company.


in The Tree House by Pip Sheppard G2G Astronaut (2.2m points)

5 Answers

+10 votes
Best answer
Y-DNA is far more reliable and definitive than atDNA (autosomal). You either have the identifying marker(s), or you do not. If the Barra Macneils don't have Niall markers, they are not his descendants. That is a disappointment Pip, and I can identify in the reverse.

My family long believed that our McDaniel ancestors were descended from Scotland's Clan Donald, but two male McDaniel cousins had Y-DNA tests and were proven not to be Clan Donald but Niall descendants. That was a big letdown at first, but Niall has his own appeal.
by Loretta Layman G2G6 Mach 3 (32.3k points)
selected by Pip Sheppard
“Niall has his own appeal.” Agreed. I noticed on the official Clan Macneil website that they still make this claim (descent from Niall, and I recall RL Macneil’s history of Clan Macneil had a father to son descent to the present clan chiefs. I would counter that Viking descent has it’s own appear, too!

My association with the clan association came from ,y maternal grandmother who was a Neal. Of course, I’ve nit been able to go back to Scotland on that line, and Neal is such a common name there that there might be no connection to the Macneils. But, I knew the president, and thought is was a good way to honor my grandmother’s heritage, even if here was no connection.

An aside: RL Macneil, an architect by profession, bought back Kisimul Castle, restored it over many years. The present chief, his son, leased it to National Trust (Scotland) for a bottle or two of Scotch for 99 years.
Pip, some of the clan associations are very reluctant to alter their stories because of DNA results, or even historical documentary evidence. Heaven forbid they'd let facts get in the way. Of course, I don't know if that is at the heart of Clan Macneil sticking to their story.

Very nice about a Macneil buying back Kisimul Castle and restoring it. The Stewarts did likewise with Castle Stalker in 1908.
It’s the romance of the thing. It’s hard to romance a story if the facts keep getting the way. That said, dna has its uses and its limitations,  as I’ve learned reading the responses to my original post. Scotland is a peculiar case, it seems. It’s only been in the last couple of years that I’ve gotten into some recently published works that shed a whole new light, not so much on research, but on the genetic, migratory, and political aspects of early Scotland. It makes one consider a wider genetic inheritance than just going by surnames. I mentioned before that I had a lot of Scottish surnames in my ancestry, but I recognize that this is not as monolithic as I once though. So, in celebrating my Scottish heritage, I realize that’s it’s not all Gael. It’s also Pictish, Brythonic, Norse, and Angle-ish!
Exactly so Pip. Here's a good illustration of the diverse origins of Scots. As administrator of the Lynn surname project at FamilyTreeDNA, I have 15 men whose ancestors were Scottish, and there are between them 4 distinct Y-DNA haplogroups.
+6 votes

Here is an interesting statistic from ISOGG:

Table: Cumulative effects of NPE

Generations p = 1% p = 2% p = 5% p = 10% p = 30%
n = 5 5% 10% 23% 41% 83%
n = 15 14% 26% 54% 79% 100%
n = 25 22% 40% 72% 93% 100%
n = 35 30% 51% 83% 97% 100%
by Helmut Jungschaffer G2G6 Pilot (544k points)
Actually I don't buy those numbers.  Presumably for this purpose an NPE is any case of somebody taking the surname Plant who didn't get the name from his biological father.  This would include not only the products of extra-marital liaisons, and people taking the name of a stepfather or adoptive father, but also illegitimate children taking their mothers' names.

Illegitimacy was never that rare, so you'd expect a higher NPE rate.  The problem is not to explain away the number of yDNA non-matches but to explain why it isn't bigger.

But as I said somewhere else, when there's no population explosion, the natural tendency of male lines is to die out.  This applies to both surnames and yDNA.  Sometimes they hang by a thread for a long time and then die out.  Sometimes they seem to have become securely established, but then they die out anyway.  Occasionally, somebody wins the lottery and the line proliferates, but this is a random fluke.

These patterns are very common in English genealogy.  Since 1400 most peerages and baronetcies have been limited to heirs male, and most are already extinct.  The relevance of this fact to surname DNA studies is ignored.

Surnames like Sykes and Plant are still around in significant numbers because they've had a lottery winner, even if he only won a minor prize.  That makes then untypical. Nobody studies the typical cases - the extinct ones.

If 68% of Plants descend from a common ancestor, he was the lottery winner.  The existence of so many Plants with the same DNA is just the result of a random combination of events.  But it's no surprise, because the surnames where such a combination has occurred are the ones that lend themselves to this sort of study.

You can't infer that the name Plant only originated once, because there's nothing in the data that tells you how many unrelated Plant families are already extinct.

And the size of the lottery prize is random, so you can't infer anything about NPE rates just by looking at the number of people not descended from the lottery winner.

But the other thing is, you don't know how the lottery winner relates to the surname-progenitor, the original Plant, if there was one.  For all we know, the Plant lottery winner was some Tudor bastard who took his mother's name.  There aren't amy grounds for the glib assumption that the majority DNA type is the DNA of the first of the name.
My intention was not to make a claim about the actual numbers, just the concept.
Yes I know.  But I don't buy the concept.

Henry de Bloggins had one son John de Bloggins who wasn't his.  Everything straightforward from then on.  All the Blogginses have John's DNA but none of them have Henry's.
Very good R J and let us not forget that surnames perse, or at least surnames passed down the generations via the male line, were not a "thing" until the poll tax of 1377, at which time they became mandatory.

Nobility had surnames of a sort, but even those can't be used to trace patrynomy. For instance the surname Lacy is derivative of de Laci, a Norman appellation that designated the origin of the lineage,which could be a place in Normandy , or the de could designate the first fiefdom in Angleland.

It often happened that such lines went extinct in the male line, or an up and coming scion of another minor nobility married into  a more prosperous or pompous line and adopted the name as his own, such as happened to, for example, the de Laci's,

Yet ths "nobles" ostensibly of Norman ancestry, but not necessarily, would name there children as (ex;) William FitzJohn), Fitz is an anglo contraction of fils de or "son of"


Commoners, per se, did not have surnames, but were known by their occupation,  (copper, potter, smith)place of residence(hill, forest, whidby,York, dale-a valley, physical characteristic (tall, short, whitehead) or patronymic (ex: son of David, or Davis, Davidson, Davidsson, depending on region Welsh, Scotch, Saxon, Norse or Dane in which the first of the name lived).


By the way, any town or name which ends in "by" is of Norse origin. For instance Gouldsby was Goulds farmstead. And these farmsteads became vills and towns


Personally I abhor the term illegitimate,as it has an unsavory and derogatory cultural connotation, for me non paternal event is sufficient, and NPE's can be from many different events other than intercourse outside the blessings of a religion or state.

Slaves in America were not allowed to marry, so they would have practices like "jump the broom". In frontier communities and especially immigrant communities, where a religious authority was not available,a couple could unite in a handfast ceremony.

As an aside, "illegitimacy" was more prevalent amongst royalty and nobility than we are informed.  Stability of succession was important, for stability of the realm or estate.

Lacking knowledge of sperm and ovaries, these people believed that in the ejaculate was a "mini human" a homonculus.

Males of the era enaged in many activities that had an adverse effect on sperm carrying the Y chromosone, or were simply away from the castle too often,or some other cause, to play their role in reproduction.

As a consequence, certain ministers of court, would even promote a liaison with the Queen, just to ensure stability in the realm by succession.

Such a cause might be the reason that YDNA testing of Richard III revealed a different Y haplogroup than other descendants of Edward III.

It is questionable if any of the descendants of Edward III are actually his descendants. It is thought by some that Edward III was homosexual and possibly did not father any children.

At least his wife Isabella of France, took up with Roger Mortimer, had Edward III's purported lover, Piers Gaveston, executed, along with Hugh Despenser elder and younger. (all of the above are also my purported ancestors, including Edward III).

Henry VIII had difficulty producing a male heir, and most certainly for that reason his court went out of their way to introduce healthy young females into his presence, and when they failed to produce a male heir, it is my opinion that they manufactured or exposed infidelity as a means of getting rid of her and introducing another. There must have been succession panic in the court, as Henry was getting on in the years,and suffered at least one debillitating infirmity from his youth.

Had they not failed, this country (and us) would not exist, and the continent would be Spanish/French.
RJ, statistics apply to populations, on the individual level it's either 100% or 0%. Inferential statistics allow us, as the table shows, not to be surprised by the Blogginses' result. That's really all.
Yes.  But a surname isn't a population, and people are trying to apply the arithmetic to individual surnames.  That's the bit I'm not buying.
Jennifer, you are talking about Edward II and not Edward III.  I just read a recent, well-reviewed biography of Edward II ... exact citation escapes me at the moment ... and there is no reason to believe he did not himself produce children.  But that could be a very long and involved discussion, that does not fit in this thread.
Hi David:


Books are books, written by persons with their own agenda. I don't discount books, but read them with discernment.I can't count the number of Family History (Genealogy) books that are full of errors.

My bad misattributing Edward II as Edward III. Tis hard to keep up with all of the 1st, 2nds, 3rds, 4ths

Richard III great great  great  grandson of Edward II, if Richard's YDNA shows cuckolding. When did the cuckold occur?
Best guess is Edward II

Line of Descent as follows
Richard III
Richard of Coningsburg, 3rd Earl of Cambridge
Edmund of Langley, 1st Duke of York
Edward III, King of England
Edward II, King of England

The notion that Edward II did not father children can be quite offputting to those who claim descent from Edward II, which purportedly is my case, reality has a way of spitting one in the eye. If I am not a descendant of Edward II, then I would like to know who carried the G2 (G-P287)haplogroup found in Richard III. This haplogroup happends to be a dominant haplogroup amongst the North Ossetians who are of Sarmatian ancestry.

G2 is found in high frequency amongst the North Ossetians, who are of Sarmatian Ancestry. A Sarmatian tribe, the Alani settled in Brittany/Normandy during the governorship of Aetius

An ancillary question is what is Edward II's haplogroup Edward III had three sons who had issue, Edward The Black Prince, Lionel of Antwerp andJohn of Gaunt, has anyone traced down those lines and tested YDNA of descendants.?

How much cuckolding was there in the line of European nobility and Royalty? Quite a bit I suspect.
The DNA mismatch is between two different lines of descent from Edward III, so the NPE is certainly to the south of Edward III.

Further back we can't say, because we don't have any other lines to test.

Edward II appears to have had an obscure but acknowledged illegitimate son.  Mistresses can lie, but obviously he can't have thought it was impossible.
Most certainly to the south of Edward III, however does not preclude Edward II

Edward II's bones are still interred, I wonder if any English group is interested in pursuing testing his DNA, or if the authority would assent.
+6 votes
I believe many other Irish clans have had similar results. Many men with the surname had the same old Irish descent as the clan chief but many did not and they were often clustered geographically. I haven't read about any serious pushback but I'm sure some have been very disappointed.

But, as you know, Irish clans often adopted people and sometimes surnames were adopted matrilineally. Of a thousand years, you'd have to expect some disruption here and there. People should be more shocked if they didn't see results like this.

Besides, decent from a Viking adopted into an Irish clan is still pretty darn cool.
by Davis Simpson G2G6 Mach 2 (21.8k points)
My DMA shows a smaller percentage of Viking descent, about 8%. I’m thinking that is might have been in the Galloway and Dumphries area as I have a Galloway (surname) line. Just a guess, I can’t trace that line back that far. My Beatys May have been from the same area. And, you’re right. I do think it’s pretty darn cool.
This is autosomal.  The genes spread out through the population.  The entire English and Scottish population is 10% Scandinavian.  The Viking blood had spread itself through the population further back in time than most lines can be traced.  So don't even bother thinking about which line you got Scandinavian autosomal from.

Y-DNA can tell you about distant ancestry, possibly, but not so much about more recent ancestry.  Which means you aren't much wiser, because your Y line only accounts for a tiny percentage of your ancestry.

Puting it another way, if Niall really did account for all those male lines, then it's unavoidable that we're all descended from Niall many times over, through lines that aren't all-male.  It's not like those with the Y type are the only Niall-descendants.

RJ and I may disagree a tiny bit about the value of yDNA testing, but he's spot-on about the Scandinavian genetic influence in the British Isles, particularly Ireland and Scotland. A quick blog post read: "Autosomal NGS Study Reveals Information about Early Ireland and Scotland."

+6 votes
There's no great certainty that "Niall" ever existed.  Legendary heroes are often fusions and confusions of several different people.

Certainly there are no adequate grounds for thinking we know his DNA.  Sykes is a geneticist, not a medieval historian.

But in any case, the notion of an Irish or Scottish clan being all descended from a single progenitor is impossible.
by Anonymous Horace G2G6 Pilot (568k points)
Agreed, RJ. Oral history (even if written down, ex. the Irish annals, Y Gododdin, etc.) be iffy. Turns out that the oral history of my Underwood line had just enough truth to it to lead in the right direction to discover the facts, but it took some digging. Alistair Moffat (The Faded Map, 2010) tries to take a more balanced view than either/or. Tim Clarkson (The Makers of Scotland: Picts, Romans, Gaels, and Vikings, 2016 3rd) maybe less so. Both are great background on the mix of folks along what are now The Borders.
There's no doubt, RJ, that much of the Niall story is a mix of myth and legend. One thing is certain though. A 2006 DNA study by Trinity College, Dublin shows that over 20% of all men in northwestern Ireland (at the time of the study) were descended from a single ancestor in the early medieval period. That certainly points to some man of great power in the region. The founder of the Uí Néill dynasty is the first such man who comes to mind.
Europe was full a long time ago, so it hasn't had an exploding population like America.  The population has been much more stable.

In that situation, male lines easily die out.  Many men have no grandsons.

Most male lines do in fact die out.  A minority hang on and mulltiply.  Only a handful will have a very large number of y-descendants.

But this is a fluke.  It's winning the lottery - a combination of random events.  It can happen to anybody.

Mathematically, the man who buys a thousand lottery tickets is more likely to win.  But he probably won't.  Mathematically, whoever won the lottery, you can say that statistically they're most likely to be the person who bought the most tickets.  But don't bet on it.

Dating the common ancestor doesn't help, even if you could, which you can't really.  The progenitor might have 3 living y-descendants after 300 years and then it mushrooms from there.

RJ you might be correct for most common haplogroups, but there is at least one specific haplogroup which might not be able to identify the actual first of the line in England, but research is coming close, and the very possible origin of that man has been identified , via ancillary research, as a Greco Persian, of Sarnatian ancestry, living in Syria around the time of the 1st Crusades https://www.yfull.com/tree/R-YP5578/ hauled back to Northumbria (Yorkshire) to upgrade English Steel, along with stone masons, to upgrade motte and bailey, timber, mud and wattle castles to stone.

The first English ancestor of the line, very well could have been a Muslim captive or voluntary emigrant from, say Damascus. Then again there were Christian dhimmi's living in the Levant, under the various caliphates.

Or he could have been aSeljug(Seljuk) Turk.


The point is that there are some rare and unique SNP's, that can be traced back to a putative origin.


Not actually traced.  It's all just handwaving and guesswork, like the Niall thing.

I have an alternative theory.  They took women on Crusades, to do the laundry.  Perhaps one of the women came home pregnant.
Actually Traced.

While there are  couple of unrelated  lines that share the same surname (it is an occupational name). One line in specific is YP5578, that appeared 750 ybp. Most persons who carry that  SNP share the same surname, the few that don't are either descendants of NPE's or persons who had no surnames until the poll tax of 1377.

The surname in it's original spelling is occupational, the ancestral home is in the north of England (Northumbria as known by Saxons and Danes).

Appearing about the same time, in the same  region was a brother SNP YP1451 (more of an "uncle") a sub of R-Z93, YP5578 is a sub of YP5585 (the only known person tested to date is an Indian=the subcontinent), and YP5578 is a sub of YP5585.

The YP1451, has as an only known match, an Iranian. He traces his ancestry to West Riding Yorkshire, as do the YP5578's. His surname is Mason, an occupation used to build castles. The earliest known rendering of the YP5578 surname is Ferror, a ferror is a smith.

Addendum: Approx 375 ybp, a boy was born YP5578,Y28816,YP5905

There are 10 descendants of him that tested Big Y, all trace their ancestry to the same individual who migrated to Jamestown aboard the Neptune in 1618.

The pregnant laundress is a cute idea, but quite farfetched considering the mores, ethics, prejudices of the time and especially place.

And the fact that the two occupations a mason and an iron worker were highly prized in England and used to upgrade armor and defenses.

Then again isn't genealogy the art of hand waving and guesswork?

It is lamentable that some lines are not so fortunate, but maybe their should be a surge of Big Y testing.

There is nothing like YDNA SNP's, beats the heck out of STR's.

A lot of money and time could be saved by simply ordering Y37 and BigY, alas FTDNA has seen a marketing opportunity, upgraded their Big Y to Big Y 500, and upped the price by $100 as well. They caught on that people were going straight to BigY from Y37, saving money by skipping Y87 and Y111
+2 votes
I've had a similar story with the Elliot surname project. Although 73% of people tested proven to be brythonic cell in origine the rest or something else. Many members of clan Elliot are not tested & all assume they MUST descend directly from the great border reiver families. I got tested to prove I did also but then I got the results, E-M35!  Oops, definatly NOT R-M269.
by Jesse Elliott G2G5 (5.1k points)
The 73% won't all be related within the timescale.  Their common ancestors might have lived in the Stone Age.
Many testers have a much more narrowed down subclades that R-M269. Plus fair geneology linking back to the Scottish borders. The point was, I & others are not. We have E-M35, R1a,I, J1,J2, G, all in the Elliott DNA project.  Many surname projects are finding this out, as the original poster has.
I was looking at a surname DNA project that had posted a load of total twaddle about the origin of the name.  Doesn't bode well.

They need to look at numbers and distribution maps and real history.

Since the studies are mostly American, they tend to cluster round the immigrants.  This gives a skewed picture - the earlier the immigrant, the more descendants, on average.

In 1700 the population of the American colonies was 98% English/Welsh.  Not a lot of Scottish and Irish immigrants had arrived, so they had a lot of catching up to do.  Elliott is a common name across England, so presumably there were a lot of English Elliotts in America before any Scottish ones.

It seems to be unusual for a DNA project to produce any information about genealogical relationships between immigrants, apart from debunking claimed brothers.
Yes, as far as my Elliott's go,we are tied to Somerset Elliott's who came to New England 16--1700s. IF at all. The E-m35 positive plus the the unsure connection past our ggg  grandfather born 1793 in New Hampshire leaves the the possibility of an adoption of the name right at that point here in America. Looking hard for possible NPEs around that time.

What so many don't realize or know is that surnames per se are fairly modern. In fact the Nederlanders did not have surnames until forced on them by Napoleon.

In England surnames were made mandatory by edict during the poll tax of 1377, and were taken or assigned based on four criteria:

physical characteristic (armsstrong whitehead, tall short

 location Sander(s),Forest, Hill, Warren, and farms, towns  and vills(names ending in :"by" are Norse for farmstead (ex: London, Gouldsby, York)

occupation (cooper, smith, ferror.potter,

 patronymic.Davis, Williams (Scotch or Welsh for Davidson or Wiliamson, Williamsson, Johnsson (Scandinavian), Johnson, Wiliamson (Saxon). FitzWilliam, FitzJohn (son of William son of John or fils de William fils de John) John Fitzgerald Kennedy was named for his maternal grandmother Mary Josephine FitzGerald

Nobility went by the name of their fiefdom or estate, hence Wm Berkeley, Gov of Virginia in 1643 was of the Berkeley's whose dungeon held Edward II, and in whose death they were complicit or executioners.

Normans called themselves (ex) Robert FitzGilbert de Laci Robert son of Gilbert of Laci. However as you pointed out most of these lines went extinct, the Laci's certainly did However they were cadet lines that emerged, from the heiress whose husbands often assumed the "de Laci" or some such appellation, which eventually became simply Lacy.

John of Gaunt,1st Duke of Lancaster, son of Edward III, would have been known as Lancaster, not Gaunt (he was born in Ghent Belgium, rendered Gaunt in English)


Perhaps worth mentioning here but the Berkeley family is one of the few if only that supposedly has an ubroken line of Male descent from Sazon nobility https://en.wikipedia.org/wiki/Berkeley_family, However the surname came from the families fiefdom of Berkeley.

The Berkeley family descends in the male line from Robert Fitzharding (d.1170), 1st feudal baron of Berkeley, Gloucestershire, reputedly the son of Harding of Bristol, the son of Eadnoth the Constable (Alnod), a high official under King Edward the Confessor.[1]

Names such as Somerset also came from fiefdom's, some Norman names like Montgomery had origin in France or the low countries.

Sorry can't help myself.



Related questions

+18 votes
5 answers
154 views asked Mar 8, 2019 in The Tree House by Kim Williams G2G6 Mach 3 (38.3k points)
+5 votes
1 answer
+9 votes
7 answers
132 views asked Mar 7, 2018 in The Tree House by Sherri Rohm G2G Crew (560 points)

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright