A series of reasonable probabilities chained together often creates an improbability of result. However much we were taught this in math class – that 90% times 90% times 90%… very rapidly becomes a less than 50-50 chance – we resist the concept. We like to apply it to history, social policy, and why our relatives act the way they do. Just about everything. If Churchill hadn't mentioned that day...
This becomes a touch humorous when discussing genealogy. The general rule is that 5-10% of births every generation are “non-paternity events.” That is, the father of record is not the actual father. If you trace your line back and keep track of how many males you had to trace through to get to a particular person, after about 6 guys you are talking about folks who have a less than 50% chance of being a blood ancestor. 12 guys, 20%.
The worse news in this is that the surname line, which is by definition father to son and the one which folks are most interested in, is thus the most fragile genealogically. We are fond of saying things like “Well, we Wallaces came from Ayrshire in Scotland, mostly, so…” Yeah. Ancestors last in Scotland in the 17th C, via Ulster in the 18thC? Sorry Douglas, you might not be a Wallace. You might still be a Scot, of course, as they were the bulk of who would have been around for those non-paternity events.
The mother to daughter line, hardest to trace, is of course more reliable because even when the occasional mother or sister gets “credited” with a birth, it goes back into that line in a generation anyway. True, there are always unrecorded events of early widowhood and remarriage that obscure the biological line, but it’s still way better than the male percentage. You get some extra advantage if you are tracing back through heavily Puritan lines, or Cohen lines in Jewish heritage, because there’s evidence that they were down even below that 5% line. Everyone else, plan on 10%.
I’m sorry to insult your ancestors like this and knock the stuffing out of your DAR line, but that’s just the statistical reality.
As DNA evidence improves, we begin to be able to track a little better. In every surname line there seems to be numerically dominant genetic group which would have the strongest, though not indisputable claim to be the father-to-son surname line. There are usually a few other large lines, suggesting that those nonpaternity events occurred in the comfortably distant past, so you can at least claim your family has been Harrisons for several centuries (though not the Original Harrison).
We got lucky in our surname evidence. A picture of my great-grandfather Charlie Wyman in Nova Scotia looks enough like me, and even more like my brother, that we can safely start counting our six males back from there. And a good thing for future generations, actually. Though anyone who spent 10 minutes with my Dad and any of his sons would be overwhelmed by the resemblance of gesture and personality (even though two of us did not grow up under him), looking at a photograph wouldn’t give you much indication we were related. None of us looked like his father either, until the last year or so, when I started seeing him in the mirror. (You guys are next). But tentative conclusions I had reached about my general ancestry I am rethinking. My confidence in having a fair number of East Anglian forebears has waned recently. I doubt that many of you will be interested in the specifics of that reasoning unless one of the surnames jumps out at you as one in your own family tree; the reasoning itself and the collateral information may be of some interest. I’ll likely write that up over the weekend.