Comment by tptacek
21 hours ago
This misses half the problem, which is that there aren't many intrinsic traits people care about. Your height is as biological a thing as anything else, but it's tied to your environment in the same sense as hair color. That's the point the author is making: that's it's difficult to deconfound these things, and that when we discuss "heritability", as a statistic that appears in the literature, we've always talking about confounded measures.
> Your height is as biological a thing as anything else, but it's tied to your environment in the same sense as hair color.
And you wouldn't draw a distinction between the person who is short because of poor diet and the person who is short because they lost their legs in a car accident? Both are "environmental factors" which affect the distance between the top of your head and the ground, but that's not what we are referring to by height.
> we've always talking about confounded measures.
No, we haven't. It doesn't matter that confounding factors exist in the data, we can and near exclusively do talk about abstract concepts. We live in a world where there are no perfect circles, but we can talk about things having diameters. We live in a world where people die from unnatural causes, but we can still talk about people having natural lifespans. That removing confounding factors is hard doesn't change the fact we routinely make our best effort to do just that because it is necessary for discussing the abstract concept we all refer to.