Jun 22, 2011

Scientific vs. Traditional Genealogy

Update: If you're following this discussion at all, you have to read Michael Hait's article on the subject. He also has a second, shorter post on it.

I haven't posted in a while, because I was "camping" for a couple weeks. Catching up on my blog reading, I found a new controversy in the genealogy blogosphere. Never a dull moment...

It all started in the comments to Randy Seaver's Saturday Night Genealogy Fun. The activity was to post your "most recent unknown ancestor." Dutch genealogy blogger and computer scientist, Tamura Jones, replied with a comment that could be literally true in many cases, but somewhat insensitive:
The scientific genealogy truth is simple: for most of you, your most recent unknown ancestors are your parents.
Another commenter was offended, and the discussion became a bit heated. So Randy wrote a separate blog post Monday with his own thoughts on the issue. As usual, Randy's response was measured and thoughtful; he could see the merit to both sides of the argument. Then yesterday, he allowed Jones to guest post on his blog, further explaining the purported distinction between scientific and traditional genealogy.

I've read many of Jones' articles distinguishing between biological, official, and legal genealogy. (Start here, here, here, and here. They have links to other related articles.) He tends to be a bit blunt, but his articles are always interesting, and often very thought provoking. It's good to take a step back every once in a while and think about what it is we're actually doing in genealogy–a philosophy of genealogy, if you will. But just as few scientists spend much time contemplating the philosophy of science, these genealogical distinctions aren't important to every genealogist (the same goes for practitioners in many fields).

First, even with modern DNA tests, it's still impossible to definitively prove any relationship. Two men may share the exact value on every Y-DNA marker tested, but all it means is they probably share a paternal ancestor. Based on the test alone, they're just as likely to be father-son, uncle-nephew, grandfather-grandson, paternal fifth cousins, etc. Mitochondrial DNA (mtDNA) is even more problematic. It's less useful than Y-DNA since maternal surnames typically change every generation. And because mtDNA changes so infrequently, large numbers of people can match without sharing a common ancestor in thousands of years.

An autosomal DNA test of over 1 million SNPs can reduce the likelihood of non-parentage to infinitesimal odds, e.g. 1 in a billion or more, but they're still just odds (and that's without considering the possibility of identical twins, who have the same DNA). Plus, the usefulness of autosomal DNA is reduced each generation further back. After about five or six generations, we're just as likely to inherit DNA as not from any individual ancestor. This raises another distinction between genealogies I don't see on Jones' website (he may have written about it somewhere, I can't claim to have made a reasonably exhaustive search).

There's a difference between our genetic family tree and our genealogical family tree. The latter includes all of our ancestors, but our genetic family tree doesn't include each individual ancestor. It certainly includes one or two of our family lines straight back for about as far as they go, because of the unique qualities of the Y chromosome (in men only) and mitochondrial DNA. Unlike the rest of our DNA, sections of these don't recombine, so they're passed down virtually unchanged for hundreds or thousands of years. However, the remaining DNA does recombine every generation, so we don't inherit any detectable DNA from some of our ancestors after only a few generations. And to top it off, it would be almost impossible to determine how much of our DNA came from which ancestor, unless there were multiple, surviving relatives from every single branch, and they were all tested.

Second, official and biological genealogy track each other pretty closely. The rule of thumb derived from various genetic studies is about a 2-5% rate of non-paternal events (NPEs) per generation. So at the highest estimate, 1 out of 20 "official" paternal relationships is possibly incorrect. The overwhelming majority are the same as what's written on the birth certificate. Besides, while the identity of the father can sometimes be called into question, it would be rather difficult to fake motherhood. There are usually several witnesses to a child's birth: parents, grandparents, doctors, midwives, nurses, etc. And those are just the people physically present in the delivery and waiting rooms. Plenty of other family, friends, and neighbors saw the pregnant woman before she left for the hospital, and came back with a baby. Barring the exceedingly rare case when the baby is switched at the hospital, the parentage of most children is a known fact.

Third, biological family isn't the only type of relationship researched by genealogists. There are several adoptees in my extended family. We may not be related by blood, but I consider them just as much family as my other relatives. (Of course, there are times I'd rather not admit I'm related to certain blood relatives.) It's quite possible some people who've researched their family history got it all wrong (biologically), because of an unknown adoption or other NPE. While biology is undoubtedly important to how we turn out, the people who raise us also pass down cultural and social values.  To most genealogists, the people they know as their parents and grandparents are the ones whose ancestry they're interested in. I know some adopted genealogists who research both their biological and adopted families. And I know of at least one genealogist who knows they're adopted, but isn't interested in their biological family. To each their own.

Finally, in his guest post on Randy's blog, Jones made this bold statement:
We have a scientific basis for genealogy that enables us to leave traditional genealogy behind.
That's simply not true. I'm a major enthusiast regarding genetic genealogy, and I can imagine even more uses for it in the future. But a DNA test is merely another tool in the kit for genealogists. It can't replace traditional methodology, only supplement it. Earlier, I wrote that a Y-DNA test alone gives no indication what the relationship between two men would be, only that they're paternally related. It's only in conjunction with other sources that we can determine the actual relationship.

Ultimately, the title I gave this post is a misnomer. Scientific and traditional genealogy aren't at odds, they're two aspects with lots of overlap. There are probably some traditional genealogical practices that are outdated and unnecessary. For my own research, I try to skip the whole paper step, and go straight to digital (I've got enough mess without printouts of all my genealogical data). On the other hand, many people have thoroughly researched, well-documented family trees that could be considered "scientific" genealogy in every sense, without ever taking a DNA test. I think the distinction should be more about the methodological rigor than the particular set of tools used in the process.

No comments:

Post a Comment