2 ms·
> This isn't really clear at all and there is plenty of evidence to the contrary. For example, consumer DNA testing pretty reliably agrees with self-reported ge
by astrange 3y ago
> This isn't really clear at all and there is plenty of evidence to the contrary. For example, consumer DNA testing pretty reliably agrees with self-reported genealogical claims.
The DNA testing is associating you with a cluster; that cluster could be any "size" and isn't based on specific patterns. (And it's fundamentally not causation; you can more or less move anywhere you want and have children with anyone you want.)
There's also a selection effect that they don't show you the clusters that they can't make work.
Also often not true; for fun, look at /r/23andMe for white people from Oklahoma and you'll see it's surprisingly common that their parents told them they were part Cherokee when they aren't even 1% native. Famously happened to Elizabeth Warren, but it's not just her.
- User23 3y agoThe causation is in the past. Up until about 500 years ago there was little to no gene flow between a number of major population groups. Obviously there has been subsequent hybridization, but we can detect that too. The largest pre-1492 hybrid groups I know of are central Asian for obvious reasons, like the Uyghur people. Last I checked you can’t legally have children with anyone you want in any jurisdiction. You have to get consent and that’s often withheld on the basis of apparent ancestry. Cultural and social factors also make it less common than it would be by random chance. Elizabeth Warren tested out something like ~1/64th Cherokee I thought? And thanks to the wonders of meiosis you could genealogically have relatively recent ancestors of some group without any genetic markers.