|it happened here, among other places.|
[side note :: i'm well aware of the issues that have been raised with these testing services. to the criticism that they are misleading, i don't believe that they are. the crucial part is that you have to read the fine print. i found that the 23 and me people do a very good job of explaining what tests they do, what they test for, and what limitations that those tests have. they are not and do not claim to be comprehensive dna analyses. if you want that, you're going to have to pay a hell of a lot of money to a private lab, assuming that you can find one, or you'll have to ask your doctor to test for one condition at a time. as for the warning that once they have your dna, the company can do whatever they want with it, such as sell it to third parties... well, there is something to that, but for my part, i don't have a problem with my dna being used for scientific research. in the absolute worst case, what would happen? some insane dr. frankenstein would make a perfect clone of me. sure, dude, good luck with that. the possible upside? my dna could help with research on deadly and debilitating health conditions. it's true that i don't have control over what is done with my sample once i've submitted it. but when i look at the possibilities, i personally am not overly troubled by those implications. once again, read the fine print and be sure you're comfortable with what the contract requires.]
to the surprise of absolutely no one, my ancestry is predominantly british/ irish. the 23 and me database doesn't distinguish between different locations in great britain, so whatever parts are english, scottish and welsh are just lumped together, which is a greater drawback in my mind than either of the points i mentioned above. that's where the limited database size comes in. you'd need a robust sampling to be able to point to exact locations, or even locations as broad as "scotland", and as far as i can tell, only ancestry has that.
i have to give a gigglesnort at the fact that my genealogy does turn up some specifically irish traits, since [as i've mentioned before], my scottish family has always been happy to hate on the irish. in fact, my most recent irish ancestor came to canada more recently than several of the scots. but that's been a conveniently forgotten fact since that family married into the scottish branch.
my tests also revealed a smattering of french/ german [they're grouped together, something that would likely piss off both the french and the germans] and scandinavian heritage. the amount present would indicate that i likely had a "full-blooded" ancestor from each of those places born in the late eighteenth to mid-nineteenth century. i don't know of either, but my research hasn't been exhaustive in that regard. i do know that i have a fair amount of scandinavian ancestry farther back than that. there is one french name that appears in my family tree [although it's considerably more common in the channel islands than in mainland france] but the person in question was definitely born in canada. [well, sort of. she was born in newfoundland, or possibly on the islands of st. pierre and miquelon. newfoundland wasn't part of canada until 1949 and the islands are still officially french territory.] if you want to go back far enough, nearly every english person is descended from a few dozen normans [a germanic tribe that lived in the north of france], which means that they're all basically inbred franco-germans. so all of us should expect to find some french/ german traces in our blood [spit], right?
well... not exactly. consider that we all get 50% of our dna from each parent. they got 50% of their dna from each of their parents. that can be carried back over every generation but, as you can tell, the extent to which we're related to our ancestors is diluted with each successive generation. it's cool to say that you're related to charlemagne, but it's been theorized that every single person in western europe is, and no one is going to show up with "pure" charlemagne dna in their bodies. [my research has lead me back to charlemagne via several different pathways, through several different parts of my family. every norman who came to england with william the conqueror has some genetic line back to the legendary king of the franks, which accounts for basically everyone in england now. interestingly, one line that remained nearly charlemagne-free was the scottish branch. i wasn't kidding when i said they isolated themselves.]
the dna that is passed down is gradually diluted as it's combined with other sources, until the amount contributed by a particular ancestor is too microscopic to be measured. it's like zeno's dichotomy paradox, but in reverse: if you cut the amount of dna received from one specific person in half, then in half again, an infinite number of times, the amount of data received from the original becomes infinitessimal. what gets deceptive is how quickly that happens.
did you ever hear your grandparents talk about their grandparents? there's a short enough space in time that stories can be communicated by someone who experienced them first hand. despite the changes in culture and technology that have taken place, you can almost picture them happening. but genetically speaking, when you talk about your grandparents' grandparents, you have sixteen different dna donors. that means that you're only 6.25% related to any of those people. people you grew up talking to grew up with that generation, but there's hardly any one of them in you.
if you can trace your history back a few hundred years [and you're lucky if you can!], those people whose names are in the far reaches of your family tree have contributed perhaps a couple of tenths of 1% of their dna to you. is it there? sure. but it's a very small part. most tests won't register dna contributions of less than 0.1%. that gets you back ten generations at the maximum, or about three hundred years. beyond that, everything is too muddy to get a clear read. sure, you may know the names of your relatives who came over on the mayflower, but they won't show up on a dna test.
when one of these tests shows you that you're, say, 5% scandinavian, that indicates that you probably had one entirely scandinavian relative contributing 5% of their dna to you. that would put that person somewhere around the level of your grandparents' grandparents or just a little further back. or maybe, it was a little further back than that, but you had two scandinavian relatives in different branches. you can see how this could develop. so when i read that i likely have family members from france/ germany or scandinavia born sometime from 1750-1850, that's already making the assumption that each of those contributions came from a single source.
[side note :: if you live in a place like canada, the united states, or australia, things can get even more mixed up. one branch of my family at least has been in canada, er, newfoundland, for close to 350 years. another recently discovered branch was in the united states in the early seventeenth century. but those bloodlines show as english, irish, dutch, etc. so when you hear that you likely had ancestors from england within the last hundred years, that's not necessarily the case. you could have had an ancestor from a purely english bloodline who lived in an english colony.]
so when you take all that into consideration, is there any point to doing these tests at all? well, you might decide that there isn't. but there can be some little trinkets of information that turn up that can be very intriguing. in my case, this one:
a tiny portion of my dna test indicated ancestry from arabic or north african sources. the amount would suggest that it came from a single relative who lived in the eighteenth century. now, that doesn't discount the possibility that there could be multiple donors spread across branches of the family that just appear to add up to one person, but consider the circumstances: i don't have any other contributions from outside northwestern europe from within the last few hundred years. arabs didn't intermarry with english people [or french, or german, or scandinavian people] at that time, so the chances of there being multiple donors in the bloodline are ridiculously small. so, in this case, it's likely that we are talking about one individual from either the arab middle east or north africa.
and this is where doing your own research to find out who was in your family can help. or in my case, just looking at them. i have no doubt in my mind that there are non-caucasian people in my family tree, because i've seen their influence. even that tiny percentage made itself known as recently as my grandparents' generation. three of my grandfather's siblings had surprisingly dark complexions. one sister was so dark that, when i met her as a child, i didn't even realise that we could be related. i've seen pictures, both black and white and colour, of their parents and it's very clear that this gene was part of the 50% inheritance they got from their mother. i've even seen a picture of her parents and despite the murky light, it's evident that she in turn got her looks from her mother.
so, yes, dna tests give information that can be complex, frustratingly vague, and not terribly helpful. but if you know a bit about your family history, they can also point you in a pretty specific direction.
sadly, i haven't managed to unearth enough about my great-grandmother's family to put a name to this arab or mahgrebi ancestor. and there isn't enough of their dna in me to narrow the geographical search down beyond the middle east and north africa. but something that generations of my family have been able to observe with their eyes now has some validation. that's pretty cool.