April 5th, 2011


My younger sister is somewhat obsessed with consanguinity in our family tree for a variety of reasons. I am less so, but I am interested. I've been working a little more on an outline for a possible book, listing and organizing some of the meta-issues of genealogy (notably, the motivations of those of us who engage in this activity with so much passion, and the purposes to which the results are put), and doing research in some of those areas as well. My sister, again, very interested in genetic genealogy. I am less so, but I am interested. So I spent a little time today trying to connect genetic genealogy to some issues that are interesting to me (you can use genetic genealogy to trace peoples when the records runs out, but it does seem like there's the potential to reify some really questionable categories if you do this naively; I was curious whether anyone was talking about this and the answer is emphatically YES). I stumbled across a forum that referred to a "coefficient of inbreeding" and realized I had no idea on how the math on this stuff even worked.

That was sort of a revelation.


Nice to have a really clear definition:

"Inbreeding occurs when an individual has one or more common ancestors. A common ancestor is one that is present on both sides of the pedigree. I.e., in order for an individual to have a non-zero coefficient of inbreeding it is necessary for both father(sire) and mother(dam) to be descended from one or more common ancestors."

This is an obvious corollary:

"Consider further what happens when no ancestor appears on both sides of the pedigree - F drops to zero. Outcrossing to a completely unrelated individual always reduces F to zero at a stroke."

My maternal grandmother's parents were first cousins: Irving's and Amy's mothers were sisters, thus those women's parents appear on both sides of my grandmother's pedigree. My maternal grandfather's lines do not extend far enough back (yet) for me to detect common ancestry, but I'm highly confident it is there (Kleine Gemeinde Mennonite for the win). But just because each of my mother's parents were the product of inbreeding doesn't mean _she_ was as well: possibly there was no shared ancestry between her mother and father.

In fact, unfortunately, she had significant parentage from the Netherlands including some from Fryslan, and the Mennonites my maternal grandfather was descended from originally came from Fryslan (not all Mennonites do -- the Swiss ones wound up elsewhere). There were probably shared ancestors, but equally they were probably from before, let's say, the middle of the 1600s.

My paternal grandparents came from Fryslan and Zuid-Holland. Surely, a common ancestor at some point; equally, not anytime recently. Surely, again, through the Dutch heritage all around, there is a common ancestor between my father and my mother -- but again, a long time ago, probably from before the middle of the 1600s.

A recent shared ancestor is unlikely to be a strong cause for concern for my sister or I (or, presumably, for our children). There is, however, something else we could probably worry about:


But it, too, is not particularly worrying (except maybe the frontotemporal dementia one. That, I'm a little worried about, given what happened with my father's aunt and his sister. And we would appear to be mostly in the clear on diabetes I, as well, despite his brother). Working in our favor on the Dutch issue is the overwhelming longevity of so many of our ancestors.

_Dear "Cousin"_, William Brower Bogardus

Subtitled: "A Charted Genealogy of the Descendants of Anneke Jans Bogardus (1605-1663) to the 5th Generation"

Printed by Penobscot Press, but the publisher is the "Anneke Jans and Evarardus Bogardus Descendants Association" of Wilmington, OH, which I suspect is the author and his wife.

Copyright 1996, the top level charts are available online. The book itself is sturdy and self-describes as "Printed on 60# acid-free paper". While in theory another edition of this will one day manifest down to the 7th generation, I, personally, will only believe it when my order has been fulfilled. My guess is that at some point in the distant future, something short of that will be published by someone other than the current author with a lot of apologies for the gaps that would be inevitable in such an outrageously huge undertaking. I sure hope I'm wrong.

According to Mr. Bogardus, when the lawsuit was still outstanding, there was a strong motivation to get more people in the group to help pay the legal costs, and a whole lot of lineages were invented, not out of whole cloth, but extra children were created to match up to lineages that got close. Bogardus expends a number of pages on these introduced errors.

I wanted this thing because it was becoming clearer and clearer to me that while someone (apparently this guy) had gone to a lot of detailed re-research to nail everything down, not all the old errors had been stamped out of existence (notably, people still repeating the non-Flekkeroy origin stories). I was mildly concerned that either R.'s or my line would turn out to be bogus and That Would Be Sad.

However! Our lineages are not bogus. Mine goes down to a Fonda; R.'s goes down to Peter Ceilo.

This isn't a full review, partly because it's not really a book one would read, however, a different review may follow when I read what text there is in greater detail.

Should you buy it? If you can find a copy. I paid slightly over $200 to the Goodwill of Pioneer Valley for mine. I don't think I would have had it in me to pay that much for a used book from the 1990s if I hadn't simultaneously felt like it was something of a gift to charity.

This Is Not Science: the 50th cousin sin revisited

I don't know _what_ argument Shoumateff is making in _Mountain of Names_, because I haven't received my copy yet. However, I have identified the probable basis for the assertion that humans are all about 50th cousins.


Yeah, I should have _thought_ to look there, right? Sort of a duh thing.

Here are the perpetrators:

Rohde DL, Olson S, Chang JT (September 2004). "Modelling the recent common ancestry of all living humans". Nature 431 (7008): 562–6. doi:10.1038/nature02842. PMID 15457259.

What did they do? They wrote a computer simulation. Never a good sign; computer simulations embed cultural assumptions and make them very, very not-transparent unless you have access to the code. "For conservative parameters, he pushes back the date for the MRCA to the 6th millennium BC (p. 20),". This, I would be prepared to contemplate. At least, I don't find it as obviously in-credible as I found the 50th cousin theory which is, on the face of it, ludicrous.

"but still concludes with a "surprisingly recent" estimate of a MRCA living in the second or first millennium BC (p. 27). An explanation of this result is that, while humanity's MRCA was indeed a Paleolithic individual up to early modern times, the European explorers of the 16th and 17th centuries would have fathered enough offspring so that some "mainland" ancestry by today pervades even remote habitats."

I'm not sure if that's really clear to a casual reader, so I'm going to run it through my standard de-jargonator: white colonizers didn't leave any community un-colonized. Anywhere. Again, on the face of it, patently absurd. There are still communities without white ancestry: North Sentinel Island and some groups in the Amazon spring to mind, but I'm sure with even a small amount of effort, we could come up with more. Assuming that the white men went everywhere and successfully left offspring everywhere -- even where they didn't -- is _exactly_ the kind of racism we work really hard to expose. Embedding it in a computer simulation and then getting it reproduced without any qualifiers all over the place is a fucking crime.

The wikipedia entry, amazingly, gets this mostly right.

"Other models reported in Rohde, Olson, and Chang (2004)[5] suggest that the MRCA of Western Europeans and people of Western European ancestry lived as recently as AD 1000."

At least this one requires a little more thought before scoffing at openly. Give me time, tho, give me time.

ETA: Honest to goddess, this kind of thing should give geneticists making dumb assumptions pause: