Hi Alex,

The "not the parent expected" possibility is remarkably common. Even in well-attested genealogies like those of medieval royal dynasties, the rate of unknown cuckoldry is about 1%, perhaps a little more. So "well-to-do" people are by no means immune. If we go back 1000 years, the probability of such events therefore becomes about 30% in any given genealogy.

And that's for situations where we know the genealogies involved. There are plenty cases where the genealogies have been hidden, either on purpose or by accident. Having a bastard child by one partner, and then later getting married to a different partner was very common: if you have a child born before the parents were married, this is a good warning sign of this. The child may inherit the mother's surname, the biological father's surname, or the adoptive father's surname, depending on who they ended up living with. Add to this adoptions from other family members, foundlings and remarriages, and you have plenty opportunity for "not the parent expected" events in your family history. There are plenty examples of this, we all have them somewhere in our family trees, and I can even point to examples in my own family that cross the Catholic-Protestant divide.

In DNA testing, particularly if you invest heavily in autosomal DNA, it's good to think of two family trees: the social one that represents the parents that actually did the upbringing, and the biological one that represents the genes that were passed on. The further back in time we go, the more these trees diverge - for everyone. In the six generations that lived before me, I can clearly point to two instances among my 126 ancestors where autosomal DNA has indiciated a discrepancy between the parents that everyone has assumed and the real biological parents.

I would not totally discount the possibility of Viking heritage. R-DF98 accounts for a small amount of Scandinavian DNA, so there were probably R-DF98 Vikings arriving in the British Isles, but there would be very few of them, and they would be more common in the Danish Vikings that arrived in England than the Norse Vikings arriving in northern Scotland and Ireland. Haplogroups I, R1a and R1b-P312 make up much more of Viking DNA than R-U106, and those R-U106 in Scandinavia tend to be from haplogroups other than R-DF98.

The Viking paper you refer to is by a well-known team - their results formalise a lot of what we were already seeing in our less-rigorous testing prior to the paper's publication. I know and have talked to some of the authors about this paper, which is part of a wider campaign to look at DNA in Ireland. However, look at the paper's conclusions again: they find a greater contribution from Vikings in autosomal DNA, but note that other studies have found a relative paucity of Viking-contributed Y-DNA. Identifying which haplogroups can and cannot be considered Viking is in itself a difficult challenge, and the results may say an equal amount about what we don't know about what Viking Y-DNA consisted of than the relative fractions of Viking Y-DNA present in Ireland today. Either way, there is relatively little Viking Y-DNA detectable in Ireland today.

Any scenario for your family remains a likely consideration. However, I think the "NPE" possibility and the Viking possibility are probably less likely than a family from Great Britain migrating to Scotland some time between the Romans' arrival in England and the Reformation. But that's not a clear winner either: we really can't tell much from a R-DF98 call alone, beyond guessing from the amalgam of data we've found from other R-DF98 families and other nearby haplogroups.

Cheers,

Iain.