Then why is he, in your telling, interpreting "race is a social construct" as "there is no genetic difference between English people and Native Americans"??? If he were not claiming race is biologically/genetically 'real' then there would be no confusion on that point. Well, I would say that any genetic study that deals in racial categories is inevitably doing this. Like, what you describe here:
...
Is precisely "starting from the variants and shoehorning them into racial categories", specifically "people's racial self-identification".
I think we're talking past each other, but we have an opportunity to find common ground. What it boils down to is that identifications are geographically correlated and so is genetic variation. For that reason, and that reason alone, we find correlations between identifications and genetic variation. This is not intended to bolster old racial theories.
We could just
not say things like white and black, and say Gambian, Northern European, and so on, with varying levels of granularity as we see fit. But given the granularity that exists in common parlance, people's identifications, the census, educational research, public policy, and so on and so forth, we don't have much choice but to point out the correlations as they exist in the existing widespread terminology.
Reich is coming from a background where people observe "African Americans" as having a higher risk of prostate cancer. They then say "African Americans" have this higher risk because of healthcare disparities. Reich writes a
paper showing it's actually a genetic effect--"African Americans" as a population genetically have this difference. Then everyone becomes outraged, saying, "but, no, don't you understand, race is just a social construct. There's no underlying basis for which you can claim anything about it." And then Reich says "Yeah, look, what I'm saying is that geographic correlations lead to correlations in variants among certain populations, and... just look at the data." And then everyone's like "But it's an arbitrary categorization devised by white supremacy! You can't count the races! You need to work closely with social scientists to fix your fraught terminology and..."
Do you see what I'm saying?
Like, really: look at the images Samson posted upthread. There is simply no way to pull racial categories out of those maps without shoehorning at some point in the process. I honestly don't see how it is difficult to grasp the concept here: human genetic variation is clearly A Thing, but it doesn't map to racial categories because racial categories are socially constructed, never indicated "neutrally" by the genetic data. And that is the real objection to use of "race" in the context of genetics, not some kind of sacred taboo on the whole concept of human genetic variation.
Yeah, I did. One of the images supported what I've been saying this whole time. Another showed poor geographic correlations for tiny portion of the genome, namely one loci, which is also something I've been saying this whole time. But a few loci don't matter that much in what we're talking about (though in some cases they do, as in the case of HbS). Most traits are
polygenic or omnigenic, involving dozens to millions of loci. Take a look at the studies I posted upthread.
Large numbers of loci vary in geographically correlated ways. In some cases, small numbers of loci vary in geographically correlated ways. That means we should expect
traits to vary in geographically correlated ways. And that is, in fact, what high throughput sequencing and genetic clustering analysis algorithms are showing.
Again, what it boils down to is that our identifications are geographically correlated and so is genetic variation. I am not in the business of claiming there is a white race or a black race or whatever and neither is Reich. The first sentence of this paragraph is not doing anything nefarious. There is an important nuance here.