A Surprise when Genealogy meets DNA

By Trond Bækkevold

Earlier published in Norwegian in Tjukke Slekta 2-3, 2011. English version Dec. 26, 2014.

Modern technology provides the opportunity to solve genealogical puzzles and breaking brick walls to a degree that only a decade ago could not be imagined, but we cannot hide the fact that new problems may also surface following the use of DNA-technology. This is my story of such an incident.

I encountered this problem some years ago, while working on a family history on my maternal side. I had earlier that year had great success breaking through the brick wall on my paternal line by using DNA-technology. This inspired me to try the same for my mother’s paternal line. This line can be traced back to Erik Henriksen (1732-1832) who had settled on the Siljuberget farm in Elverum in the Østerdalen area of Hedmark in the mid 1750’s. The hypothesis was that he was descended from Finnish immigrants by the family name of Liitiäinen.

Obviously my mother has no Y-chromosome, so I contacted my cousin Terje (son of my mother’s late brother). He was easily persuaded, and the test was taken and sent to Family Tree DNA in Houston. Eight weeks later the result arrived.

Since several members of the Liitiäinen clan had been tested during the last years, we knew that they, as most Finnish immigrants to Norway, belonged to the N1c-haplogroup. It was therefore a big surprise that the haplogroup of my cousin was R1a, even though we had to admit that his phenotype looked more Norwegian than Finnish, and he certainly did not look like a Liitiäinen.

Emil Uthus

Emil Uthus – R1a

Unexpected results of a DNA-test may have several explanations. The most likely would of course be genealogical mistakes and false hypothesis. However, the Siljuberget descendants have been thoroughly investigated, and the weakest link in the chain would be the link between Erik Henriksen and his supposed father, Henrik Mortensen Liitiäinen. Two other sons of Henrik Mortensen had, at that time, already had their DNA tested through the Forest Finn project at FTDNA and their STR-profiles were available from the project site. Both these lines matched, and, as expected, were haplogroup N1c.

We thus had two possibilities. Either our hypothesis was wrong, and R1a would be the correct haplogroup for this line back to Erik Henriksen, or it could be a Non-Paternal Event (NPE) somewhere along this paternal line. That means that one of the alleged fathers in this line was not the biological father. One way to find out would be to test another patrilineal descendant. By going back to our common great grandfather Emil Uthus (1877-1972), we were able to locate a descendant of his younger brother Søren Uthus (1878-1962). He was also willing to do a test, and to make a long story short, his result was N1c and matched the other Liitiäinen with a genetic distance of 2. It was now obvious that it had to be a NPE in my great grandfather’s generation or later. This came as a surprise on both of us, but on reconsideration, it should not.

The Family Myth

During a visit to my mother’s aunt Sigrid in 1998, she told me that my great grandfather Emil in reality was not the biological son of his father, but of a «prominent Swede with blue blood» (i.e. The King of Sweden). At that time I did not find this story particularly realistic, and, in fact, I felt it unpleasant and inappropriate on behalf of my great-great grandparents. There are several stories like this, some of them may be true, but most are very unlikely. I pushed it all away and forgot the entire story until now. Maybe it really was true? After all Emil had a different appearance than his siblings, being taller with a more slender body and long face, whereas his siblings were shorter and had wider face and body shape. Also, one would presume that the Swedish Royal Family would have a more continental haplogroup than the typical Nordic R1a «Old Scandinavian» group that my cousin had.

Emil Uthus 1928

Emil Uthus 1928

I must also admit that we, at this point, were not able to pinpoint in which generation the break had occurred, and it felt unpleasant. It could be narrowed down to the last four generations, where I was a descendant of three, and I had personally known two of them.

Could we find out?

To find a solution to this mystery appeared to be almost impossible. We would at least need a «suspect» to test against, and to locate such a person would require so much information of persons, time and location that it seemed impossible. Could there be another way?

My cousin had through his Y-DNA test at Family Tree DNA 3 matches, one with genetic distance 2, and two with distance 3, one Norwegian and two Americans. I contacted the matches, and the Norwegian and one of the Americans responded quickly. They could not give any helpful information, other than it was no obvious connection. I knew of course that even a GD 2 match can be before the genealogical timeframe. FTDNA has a calculator that can calculate the probability of how many generations it is back to a common ancestor. In this case it came up with these numbers versus the Norwegian with GD 3:



4 27,24 %
8 57,78 %
12 77,85 %
16 89,03 %
20 94,77 %

When it came to the American with GD 2, it was a greater probability that the relationship could be closer, but even this could not be certain given that different positions on the Y-chromosome have different mutation rates. Still, the calculator came up with these numbers:



4 57,59 %
8 88,15 %
12 97,14 %
16 99,36 %

In other words, the probability of finding a relationship within a genealogical time seemed reasonable. It was in fact even a great possibility it could be quite close. It was a pity the American didn’t respond.

About two weeks later he finally answered and when I read the answer I felt the hair raise on my arms. Richard wrote something like: «I cannot see any common ancestors, but my line is from Elverum in Østerdalen in Norway», and then he named his paternal line back to Anders Embretsen (1830-1894) from Bernhus i Elverum. His son emigrated and was the grandfather of our match Richard. Even though this didn’t give an immediate solution, we clearly had a family tree we could search for possible candidates.

I contacted Ronny Rismyhr Haugen who is working on a «bygdebok», i.e. genealogical history of the parish, for Sørskogbygda in Elverum, as he is known for having an almost complete genealogical record of all the people in this area covering the 1800’s. He responded immediately, and that was when I really got the goosebumps. The aforementioned Anders Embretsen had a brother Nils Embretsen Bernhus (1837-1922), and he had been a neighbour of my great-great parents! Nils had been married to a widow, 17 years older than him. They had no children together, but she had several children from her first marriage. When I take into consideration that my great-great grandmother was married to a man, 15 years older than her, a man who was known for having a drinking problem and could be away drinking for days, the idea of two unhappy marriages can easily be visualised, although this is nothing more than a speculation.

Oline Persdatter 1835-1923

Oline Persdatter 1835-1923 mor til Emil

The fact remains that my great grandfather was not the biological son of the man given as his father in all official records. It is also quite sure that he must be of the same paternal line as the men from Bernhus. A simple Y-DNA-test had led to a situation where 35 years of genealogical work had to be discarded, and 1/16 of my ancestry had to be replaced. It must be said that a certain identification of the real father cannot be done without testing of the two candidates against each other. This will be difficult today without exhuming both of them, which, of course, is out of the question.

I have also done additional testing to make certain we had identified the correct point where the NPE occurred . After all, it could have been my great grandfather that was not the father of my grandfather, or my grandfather not being the father of my uncle, or my uncle not being the father of my cousin. Several Family Finder tests of my cousin, second cousins and even some third cousins confirmed that the break occurred between my great grandfather and his paper trail father. The probability is quite high that we have found the correct father.

Skigard vinter 2

To confirm this I had to find a descendent of the Bernhus-family. Unfortunately Nils E. Bernhus had no other children, so I had to test a descendent of his brother, a man living in my hometown called Ola. He turned out to match my mother as third cousin on the Family Finder autosomal test, which would be just right. He also matched my cousin with a distance of 2 at the Y37-test, as well as an exact match with Richard. They did not match on the Family Finder though. That is not that strange, as the chance of having a match with a third cousin is about 90%, the chance of a match with a fourth cousin is between 40 and 50%. The actual relation in this case was third cousin once removed. Absence of a match in the Family Finder does not exclude a relation in this case.

I find the combination of autosomal tests and Y-tests a very useful tool when it comes to solving problems like this. The limitations of the tests must be kept in mind though. For more distant relationships like fourth cousins the autosomal tests quickly become useless. The amount of intermarriages in a population must also be taken into account.

I must mention that my own match with Ola was stronger than the match my mother had with him. The reason is that I was related to Ola and Richard on my father’s side, which made me «useless» in proving the particular relationship we were researching. Fortunately, my mother and Terje had no other known close relationships to Ola. This means that, so far, Nils Bernhus is the best candidate I have. Finding out that my great grandfather had another father was quite a shock, but I have come to term with the situation, and cannot see any reason why the rest of the family should not do so also.

Ethical considerations

Unexpected test results like this can lead us into situations where we may hurt people. When it is so far back in time that not even the grandchildren are alive, I do not find it a problem to publish this information even though some in my family may have second thoughts. I would not have published this information if any of the siblings of my grandfather were alive.

In the Norwegian State archives, adoptions are closed for 100 years, and I think the 100-year rule could be a guidance for publication of such findings. Of course, each case must be subject to individual consideration, but I recommend that DNA-genealogists should be careful in such cases. Cultural differences between different countries should also be taken into account.

In this particular case, I find the descendants’ right to know about their biological origin more important than the privacy of a great-great grandmother who has been dead for more than 90 years. I also find this topic so interesting for genealogists in general, that it will justify publication of this article.

Comments are closed.