Read Our Article in the March Issue of BBC WDYTYA Magazine
FGB Free Clinic - Case no. 6 - Surname To Place Name To Surname

When Your Ancestors Disappear From Your DNA

Savoie Poster

Some years ago, we returned to the homeland to visit our mother. She was something of a social butterfly and her house generally was  filled with revellers every evening. On this particular visit, we greeted one evening's round of strangers with the aplomb we had acquired from an early and thorough training in graciously welcoming the latest dozen of wacky characters our mother had discovered.

On introduction, one fellow firmly refused to believe that we were our mother's daughter. Initial good humoured assurance on our part gave way to some annoyance as the man continued to assert that, not only were we not our mother's daughter, but that she had no daughters. "She has two daughters, actually," we said, our sardonic tone moving toward the acid. In the end, we failed to convince the ill-mannered dolt, but what shocked us more was our sense of outrage at having our rather obvious blood relationship negated.

So, we fear, may be, at least in part, the feeling of Monsieur B. who has written asking for advice:

"I just received my Ancestry DNA evaluation and it presents me with a puzzle.  My great-grandfather came from Thures, near Cesana Torinese, Circondario of Susa, in present-day Torino.  His ancestry goes back as far as the parish records go.  He married an English or Scottish woman and their third grandchild, my father, had a 100% French mother (verified).  My mother’s side is all Swedish Finn.  My DNA results show origins in Great Britain 44%, Scandinavia 27%, Finalnd/Russia 8%, Ireland 8%, the Iberian Peninsula 6%, Eastern Europe 3%, European Jewish 3% and Italy/Greece 2%.  Ancestry says not to put much importance in the smallest percentages.  No Western European at all, where I expected my French ancestry to appear!  So, what am I to make of this?  I wonder if my family, and perhaps many of the families in the Susa with French names, adopted a French identity centuries ago but are really from elsewhere.  I identify the Duchy of Savoy with French origins and even the Ecartons that predated Savoy in the region.  Has anyone else ever encountered similar results?  What do we really know of the origins of the French-named families found in Susa (other than the Waldensians who were a distinct population)? I have another French line, Lalange, that stayed put in Indre for centuries yet nothing shows up!"

Dear Readers, might any one of you be an expert on DNA and genealogy? Can anyone suggest the solution to this mystery? We await your response with optimism!

 The comments received to date, some of which refer to the subject of the legality of DNA testing for leisure in France, as we discussed some years ago here:

While it would seem unusual to find no trace at all of a grandparent’s ethnicity (¼ of your own genetic makeup), there is one common misconception about how genes are inheritied: although you’ll get 50% of your DNA from each parent - half of their own genes - you don’t know WHAT half you’ll get. I think most people imagine something like a circle that’s getting cut in half and handed to you, but it’s not that clean-cut. The 50% you get is speckled all over that circle, a bit here, a bit there. If your parent is half Irish and half French, you might expect you’ll get 25% of each, but you might not; maybe you just end up with all the Irish-origin genes (and a sibling could end up with a different set.)  Zeph.


One must also take into account that precious few French people have taken DNA tests for ancestry. So many more people with English, Irish, German, Polish, or Scandinavian ancestry have taken the tests, that ethnicity charts tend to skew in those directions. Unless and until many thousands of French take genealogical DNA tests, ethnicity charts for those us with French roots are little more than entertainment. Peggy


It seems there could be several factors at play.

Perhaps primarily,’s ethnicity estimates (underline estimates), even though a nice, solid-looking number is presented, should be viewed with significant skepticism. They might be somewhat correct, but they might be very wrong. These numbers are averages of their 40 current schemes of deriving a number when comparing your DNA to a reference panel of 3,000 tests (samples taken roughly within the last 10 years) which they believe are representative of 26 regions. Each of these average numbers are derived from a range, a big range (click on any region to expand the box to reveal them—a solid-looking 37% nearly crumbles away when revealing a 4%-65% range). Their Scandinavian estimates have often been singled-out as being significantly over-estimated, though yours might not be. Curiously, even though their Western Europe region is solidly France and Germany, their map of their Great Britain shows secondary and tertiary regions which include most of France and Germany.

Offering one of my questionable estimates as an example, my mother’s ethnicity estimates include 31% Irish, with a range of 17% to 45%, and she has no known Irish ancestry, at least back into the 1700s, so it seems couldn’t be more than about 3%, and possibly much less, so, what to think?

As Zeph pointed out, beyond parents, we might not receive an arithmetically perfect division of DNA from the preceding generations. Though parents are always be 50%, grandparents, from who we might expect to receive 25% each, have been observed to vary between about 17% to 33%, though it is probably a bell-shaped curve, so usually closer to 25% than the extremes. Further, chromosome recombination is chunky, there usually being 1 or 2 breaks per chromosome, but sometimes 0 or 3, so you might get 100% or 0%, or anything in-between, of any particular chromosome from each of your parents.

Of course, all genetic genealogists learn to incorporate the possibility of unexpected fathers or adoptions, low as it may be, when doing analysis and interpretation. Perhaps it should be pointed out here that paternity testing in France is currently illegal without signed consent forms to avoid the chance of revealing indiscretions which could disrupt families, almost suggesting it is a bigger concern here than elsewhere, though it may not be.

Overall, I would mostly ignore the current ethnicity estimates, unless you really suspect there’s something more there. Hopefully, these estimates will improve in the future, but they all seem to rely on the notion that people did not move around much--that current residents are representative and they all remained in the same regions for centuries or millennia--but move around many did. Human history is very messy. I suspect there is greater possibility is that broad population Autosomal DNA analysis may eventually provide better genealogically relevant information than any ethnicity estimate, but these are still very early days.   David C


It is my understanding that DNA testing in France is prohibited by law so there is a giant hole in Western European DNA results. Terri Meeks


When it comes to genealogy one cannot rely on DNA results alone. They must be supported by documentation. Apart from the important point already made. Any of the following scenarios will cause a probalem. These may have happened many generations past and the current family could be completely unaware!
• an illegitimate male child passed off as mother’s brother.
• an illegitimate male child within a marriage.
• a husband adopting his wife’s surname.
• stepchildren adopting their stepfather's surname.
• an adopted male child who takes the surname.
• a foreign name altered to resemble an existing local surname.
• a male purchaser of property adopting the seller’s surname.
• a mis-spelling at some point that switches to a new surname entirely.

Monsieur J, by e-mail


In her comment below, Roberta Estes refers us to her excellent blog post on this very subject. Highly recommended!


I had a similar issue with my Ancestry DNA report. But when I uploaded the genetic file from Ancestry to DNA LAND, I had much finer results. As I understand the process, different DNA assessments focus on different areas. DNA LAND interpretation of your genetic test results is free and is frequently updated as more data from areas are collected. It's operated by geneticists affiliated with Columbia University and the New York Genome Center. You can read about their research and purpose on the website. At first I thought my French DNA wasn't showing up, because I assumed it would be placed in Western Europe, when in fact it is described by DNA LAND as included in the area they call southwestern Europe. Most of these genetic assessments don't really assign countries, rather areas where certain results are common. My experience is that each company has its own way of describing genetic heritage as well as what they look for or focus on. It makes a difference. One surprising result to me was the affirmation of English heritage in northern France, that verified the area with my ancestors who had the surname L'Anglois. Another was the result that verified my grandmother's claim of African ancestry. At approximately 6%, that is a size that can put autosomal DNA into the right time frame for my family. and my grandmother who no one believed when she said, "We're Black." The reason many people like to use autosomal DNA is that it gives more recent time-frame stability. Professor Henry Louis Gates gives an excellent explanation in several resources on the internet. If I've made an error or left something out here, one of his articles might help clarify the issue. He has many references to follow through on, too. DNA LAND has already matched me with several Québécois cousins. Those of us who chose to connect, with one exception, found one or more common ancestors in our genealogies. The one exception is clearly related, as our genealogies show, but we haven't found the common ancestor yet. I suspect one of us might have made an error that will become clear with more work. Some of us charge ahead without the good documentation that would clarify many relationships. I'm always surprised when someone is reluctant to "give up" an ancestor who isn't one of their own.   Madame C


Goodness! This is topical. These are all incredibly helpful! Many thanks. There are more comments below than we could put here. Please do read them as well.


©2016 Anne Morddel

French Genealogy