3 ms·

And what we do know, doesn't necessarily mean what those searching for descent from antiquity think it means. All tracing of such descent assumes that pregnan

And what we do know, doesn't necessarily mean what those searching for descent from antiquity think it means. All tracing of such descent assumes that pregnancies involved known paternity, whereas we know that in general, there is a lot of error in recorded paternities (one of the more interesting side effects of DNA sequencing, both in consumer offerings, and in medical and official use, is demonstration that far more children are not the children of their named or assumed father than was suspected). Add to that, the fact that you could potentially have, say, Charlemagne in your direct line of descent, and still have none of his genetic material in your makeup. This is so because at the remove of Charlemagne, you're 48, give or take a few, generations, removed from the trillions of leaves on your family tree, but you get roughly 1/2^-n of your DNA from an nth generation progenitor and have only 2^36th or so base pairs of DNA in your entire genome. So, Charlemagne can be your umpty-great grandfather, and yet you in be in no genetic sense, a descendent of Charlemagne. In other words, past a dozen or so generations, any data we've got on bloodlines has a lousy signal to noise ratio for proving direct descent.

OkayPhysicist 4y ago
Matrilineal or patrilineal lines give you much stronger signals than that, though, thanks to mitochondrial and Y-chromosome DNA respectively.
- tsimionescu 4y ago
  Sure, but unbroken patrilineal or matrilineal lines of descent are rather improbable over such long times - any family that had only sons or only daughters breaks the line, and that's not particularly uncommon. Add some infidelity into the mix and it becomes even harder to believe.
- walnutclosefarm 4y ago
  I don't see how. Sure, if you had mitochondrial DNA from your long-dead male ancestors, you could suss out your true patrilineal line of descent, but it's extremely unlikely that such data is available. You might be able to get it by inference by analyzing the fanned-out descendents of a relatively recent putative paternal ancestor - back 8 or maybe ten generations, but even that's unusual. And, if you're targeting a specific ancestor, it requires an unbroken line of paternal descent - as soon as your path to, say, Henry VIII goes through a female descendent, at any generation, you've lost the Y chromosome relationship. Same argument applies with the sex roles reversed for females and mitchondrial inheritance. At ten generations back, you've got 1024 grandmother leaves on your tree. Only one of them is your mother's mother's .. mother's mother. At 20 generations (say, 500 years, so the actual age of Henry VIII), you've got a million grandmother or grandfather leaves on your tree, and only one is purely matrilineal, and one purely patrilineal. (That doesn't mean, of course, that you've got a million distinct grand^20 fathers - it'll be way less than that, because the some men will appear on many different leaves, and the same for mothers. But the odds that the path you claim to Henry VII is purely maternal or purely paternal are still very small.
jlawson 4y ago
Non-genetic descent is actually much stronger than that even. When genes are sexually recombined, they don't recombine randomly per-base pair, they're cut and recombined in big strips. The chromosomes actually only 'snip' in a small number of places. This means that after a relatively small number of generations, it's quite easy to have no genes at all from one of your ancestors. I don't recall the exact numbers but it's something like 8-12 generations where you start having serious numbers of ancestors with no genetic connection.
- walnutclosefarm 4y ago
  Of course. Absent recombination, you'd start having ancestors dropping out after only 5 generations. With recombination, it's a few more. I didn't make the argument that way, because when I have in the past, I get a bunch of "yeah, but ..." rebuttal. In addition to the complicated nature of recombination, there is the additional complication that when you get back 6 to 8 or more generations, the same progenitors generally start appearing in multiple branches of your family tree - sometimes in many, many branches. But it is nevertheless the most correct. We all probably have great, great, great, great, great, great grandparents from whom we have no genetic inheritance at all. Add a couple more greats, and it's a near certainty.
bjourne 4y ago
> far more children are not the children of their named or assumed father than was suspected The origin of this factoid is one British study about paternity tests for fathers on some street. However, the methodology of that study was deeply flawed and the fathers self-selected (i.e. they already suspected they weren't the biological fathers). Later studies has put the rate of "false fatherhoods" at about 1%. https://www.iflscience.com/false-paternity-isnt-actually-widespread-34920 https://www.iflscience.com/false-paternity-isnt-actually-wid...
- walnutclosefarm 4y ago
  I've never heard of the British study you propose being the source of my observation. But it doesn't matter. If we're talking about deep ancestry, like say tracing your lineage back to Charlemagne (which seems to be a popular anchor), you're talking about 40 to 50 "fathers" between the good old King and a modern descendent. If there is a 1% probability in each generation that the paternity recorded was wrong, then you're looking at at least a 1 in 3 chance that your paternity chain is broken. If the rate is even as high as 2%, it's 1 in 3 that the chain isn't broken. One break, of course, and your claim of descent is invalid.

OkayPhysicist 4y ago

Matrilineal or patrilineal lines give you much stronger signals than that, though, thanks to mitochondrial and Y-chromosome DNA respectively.

tsimionescu 4y ago
Sure, but unbroken patrilineal or matrilineal lines of descent are rather improbable over such long times - any family that had only sons or only daughters breaks the line, and that's not particularly uncommon. Add some infidelity into the mix and it becomes even harder to believe.
walnutclosefarm 4y ago
I don't see how. Sure, if you had mitochondrial DNA from your long-dead male ancestors, you could suss out your true patrilineal line of descent, but it's extremely unlikely that such data is available. You might be able to get it by inference by analyzing the fanned-out descendents of a relatively recent putative paternal ancestor - back 8 or maybe ten generations, but even that's unusual. And, if you're targeting a specific ancestor, it requires an unbroken line of paternal descent - as soon as your path to, say, Henry VIII goes through a female descendent, at any generation, you've lost the Y chromosome relationship. Same argument applies with the sex roles reversed for females and mitchondrial inheritance. At ten generations back, you've got 1024 grandmother leaves on your tree. Only one of them is your mother's mother's .. mother's mother. At 20 generations (say, 500 years, so the actual age of Henry VIII), you've got a million grandmother or grandfather leaves on your tree, and only one is purely matrilineal, and one purely patrilineal. (That doesn't mean, of course, that you've got a million distinct grand^20 fathers - it'll be way less than that, because the some men will appear on many different leaves, and the same for mothers. But the odds that the path you claim to Henry VII is purely maternal or purely paternal are still very small.

jlawson 4y ago

Non-genetic descent is actually much stronger than that even. When genes are sexually recombined, they don't recombine randomly per-base pair, they're cut and recombined in big strips. The chromosomes actually only 'snip' in a small number of places. This means that after a relatively small number of generations, it's quite easy to have no genes at all from one of your ancestors. I don't recall the exact numbers but it's something like 8-12 generations where you start having serious numbers of ancestors with no genetic connection.

walnutclosefarm 4y ago
Of course. Absent recombination, you'd start having ancestors dropping out after only 5 generations. With recombination, it's a few more. I didn't make the argument that way, because when I have in the past, I get a bunch of "yeah, but ..." rebuttal. In addition to the complicated nature of recombination, there is the additional complication that when you get back 6 to 8 or more generations, the same progenitors generally start appearing in multiple branches of your family tree - sometimes in many, many branches. But it is nevertheless the most correct. We all probably have great, great, great, great, great, great grandparents from whom we have no genetic inheritance at all. Add a couple more greats, and it's a near certainty.

bjourne 4y ago

> far more children are not the children of their named or assumed father than was suspected The origin of this factoid is one British study about paternity tests for fathers on some street. However, the methodology of that study was deeply flawed and the fathers self-selected (i.e. they already suspected they weren't the biological fathers). Later studies has put the rate of "false fatherhoods" at about 1%. https://www.iflscience.com/false-paternity-isnt-actually-widespread-34920 https://www.iflscience.com/false-paternity-isnt-actually-wid...

walnutclosefarm 4y ago
I've never heard of the British study you propose being the source of my observation. But it doesn't matter. If we're talking about deep ancestry, like say tracing your lineage back to Charlemagne (which seems to be a popular anchor), you're talking about 40 to 50 "fathers" between the good old King and a modern descendent. If there is a 1% probability in each generation that the paternity recorded was wrong, then you're looking at at least a 1 in 3 chance that your paternity chain is broken. If the rate is even as high as 2%, it's 1 in 3 that the chain isn't broken. One break, of course, and your claim of descent is invalid.