I think you may be missing the argument tptacek is making. He is conceding your definition. However he is distinguishing between the concepts of "heritability

I think you may be missing the argument tptacek is making. He is conceding your definition. However he is distinguishing between the concepts of "heritability" (variance that can be statistically attributed to genetic factors) and "genetic determinism" (variance that is caused by genetics). To drive a wedge between these two concepts, he is drawing on examples like the ones put forth in Ned Block's paper, like how wearing of earrings is (or at least used to be) almost perfectly correlated with having XX chromosomes, despite the fact that being biologically female does not directly cause the wearing of earrings. Thus, according to him, your interpretation of heritability is bogus, unscientific, race science, etc. Is this a good argument? I don't think it is. While XX chromosomes may not directly cause the wearing of earrings (or lipstick), genetics clearly are a link in the causal chain that ultimately creates the variance observed when heritability is measured. Women wear earrings because they are biologically women and we live in a culture where women wear earrings. The two factors combine to produce the effect. The most charitable interpretation of tptacek's point is that there are other, non-genetic inputs into the directed graph of causes, that these non-genetic factors combine with the genetic factors to produce the variance, and that the non-genetic factors are morally more important. For example, clearly culture plays a role in why only women wore earrings when Block wrote his paper. This is even more clear when we see that this is no longer the case, and now it is reasonably common for men to wear earrings too. The logical conclusion is that non-genetic factors can interact with a genetic factor to produce a difference that the genetic factor itself could not have produced. For example, you could imagine a scenario where racial prejudice and discrimination caused teachers to give up on black kids based on nothing but their race, causing them never to achieve the IQ they could have achieved if they had been given the same educational attention as a white kid. This would be a case where an IQ difference is heritable, even though the true blame lies with how people are treated based on their race. But is that the world we live in? Given the strong heritability of IQ, and the enduring disparities that are frustratingly durable even through decades of remarkable social change, I think the onus is on the people proposing such an extra non-genetic factor to identify and measure it, in a way that can be isolated and empirically evaluated. Otherwise the argument is isomorphic to: "you're a race scientist if you think that height differences between population groups could be caused by genetics." The extra, non-genetic factors that people would propose (socio-economic status especially) can be controlled for, and do not seem to explain the gap. Most teachers seem to value the success of racial minorities as highly, if not more highly, than other students. I think that some cultural anti-patterns in black culture could certainly play a part (like cultural pressure against being a good student), but that has been decreed to be just as verboten as the genetic hypothesis.