> [DNA is not known to correlate] with race. DNA obviously correlates with race. What do you think makes the skin color of black-skinned people black? Or are y

> [DNA is not known to correlate] with race. DNA obviously correlates with race. What do you think makes the skin color of black-skinned people black? Or are you claiming skin color doesn't correlate with self-identifying as black, or being identified as black by the rest of the society? > You can't tell which race someone is by their DNA, because (...) race is a social construct based on appearance, not a biological reality. So is gender. However, the biological reality is that you can tell which gender someone self-identfies with by their DNA correctly at least 99 times out of 100. How hard it is to believe you can do at least as well (success rate >0.99) with race self-identification? As a data point, some people claim they can do it, and would bet their lifes on results. http://archive.wired.com/politics/law/magazine/16-01/ps_dna http://archive.wired.com/politics/law/magazine/16-01/ps_dna > Sure, if you assume from the start that skin color genes are correlated with intelligence genes then it follows that skin color is correlated with intelligence. I'm saying there's no reason to make that assumption to begin with. Nobody makes this assumption, though (well, except for the sake of argument). There's no reason to assume the opposite either. Science works by formulating a hypothesis and testing it.