Never Say Never

With well over a thousand matches for each of the family's DNA kits I manage, I had long ago promised myself I'd limit my frustration by only pursuing those matches whose relationship was likely to be fourth cousin or closer. After all, it would have to be a dedicated researcher who could come up with a reliable, documented tree stretching back to the fourth great grandparent which would be necessary to go beyond that point of relationship. Granted, there are plenty of serious genealogists out there with trees numbering in the tens of thousands of relatives, but the great majority of my matches who don't have the dreaded "No family tree" label have trees with less than one thousand members—often, less than one hundred. Not much of a potential for locating most recent common ancestors among those sparse familial branches.

In this season of reconsideration, when I'm revisiting my protocols for sorting through this haystack of disparate matches, I'm beginning to see my way around that dictum. At Ancestry DNA, at least, segmenting my results by using their "Shared Ancestor Hints" filter has demonstrated the value of including the more distant cousins in my viewfinder.

In fact, using this approach has led to a few observations. Easiest to see, among those new insights, is that quite a few of my distant matches actually share their own surname with a surname prominent in my family tree. Of course I'm going to check out the results for a distant cousin who himself bears the surname Broyles or McClellan, for instance, since those names are part of my own genealogical heritage. While that approach might seem naive for someone searching their Jones or Smith heritage, it is a gimme for people researching hard-to-find names like Taliaferro or Aktabowski. Not too many of those around, you know.

Two other observations popped up, now that I've shifted from my old policy of never going beyond fourth cousin, but they seem to contradict each other. The one thing I had long ago noticed was that, at least at Ancestry DNA, the estimated level of relatedness usually seemed spot on—and when giving a range (say, "third to fourth cousin"), the end result usually settled with the closer relationship.

However, these two observations broke from that pattern. In the one, my kit would be estimated to compare with a match at a certain level, and then fall short of that. In the other, my kit would turn out to match the other person at a more distant level than predicted.

Several of my husband's matches were predicted to be more distant relationships than they actually turned out to be. For instance, one match slated to be within the range of fifth to eighth cousin turned out to be a third cousin, once removed. Another "distant cousin" proved to be his fourth cousin, rather than the fifth to eighth cousin it was predicted to be.

Granted, some of those were near-misses. But in exploring this new world of distant cousins—which, previous to this point, had been off limits by my own policy—I observed something. In these lines, which all happened to be related to my husband's mother, looking at the details of the matches' trees confirmed that they, like my husband, had more than one route to relatedness. Yes, that means they were their own cousins. But it also means that there is something in their genetics which foils the algorithms for deducting which level of relatedness the two parties should be.

On the flip side, my own mother's lines have inter-relatedness, solely by virtue of being early colonists in a vast—and vastly empty—new continent. Repeated inter-marriages among extended family members over the generations should predictably produce results similar to those of my husband's family. But in some cases, I see the opposite result.

Particularly in the lines involving my Lewis, Meriwether, Gilmer and Taliaferro iterations, the relationship turned out to be more distant. One match, for instance, was predicted to be of a fourth cousin level, but turned out to be fifth cousins, once removed. Another supposed fourth cousin turned out to be a seventh cousin.

I'm sure we'll all see surprises in this new world of genetic genealogy. Some will come with scientific explanations in hand to help us understand the dynamics behind the aberrations. Other situations may end up, hanging around on the street corner, waiting for someone to come up with the reasons why it turned out that way.

One thing is sure, though: if I had never bothered to explore these far distant relationships, I probably wouldn't have noticed such anomalies hiding in my genetic outliers. With the more finely-honed tools, though, there needn't be any reason to shy away from exploring these even more remote relationship possibilities.