I've posted this before, as have others, but I really think
this NYT article by Harvard geneticist David Reich is required reading. It's cfc, we're all smart people capable of deploying our Reason to arrive at Certainty without doubt and Truth without error. But it'd be nice to see more nuance beyond the classic 1970s consensus on social constructionism. I sometimes see that whole "video game NPC effect," where it seems like everyone has learned something and added a little nuance to their views, but the next time the topic comes up nothing's changed.
1. Does Race exist and has its reference/basis in reality?
Parroting a post I wrote a few months ago, "race" has a lot of baggage and doesn't lend itself to good discussions. "Ancestry" and "populations", however, are better defined and lead to discussions that move past the racism-vs-1970s-constructionism debates that have dominated this topic for decades.
In roughly 1500, many pairs of populations had been geographically, and therefore genetically, isolated for tens of thousands of years. E.g., Western Europeans and Bantu peoples. Tens of thousands of years is enough time for SNPs to build up and create population-level differences in many distributions of phenotypes. Melanin levels and eye color were not the only things affected. Americans who have most of their ancestors in Europe 500 years ago are more likely to have multiple sclerosis. Americans who have most of their ancestors in Africa 500 years ago are more likely to have sickle cell anemia. These are real differences in the distributions of diseases between these two populations.
Of course, a population is an artifact of a statistical procedure. There's a popular view that it makes no sense to construct populations based on whatever "race" means. I pretty much agree with that. However, it's unreasonable to think it's meaningless or harmful to construct populations using broad geographic area of ancestors several hundred years ago. It makes sense because without modern transportation, genetic divergence is geographically-determined and 40,000 years of separation is enough time for some differences, including medically relevant ones, to build up (to preempt an objection: using Africa as a region means you're trying to clump the Bantu with the San, which are highly divergent groups. But this objection has less force in practice because the larger, more similar groups, are hundreds of millions of people--dwarfing, say, Pygmies and San. In large part, this is because of the Bantu southward expansion several thousand years ago and Bantu lifestyles more conducive to population growth). In any case, we don't even need to construct populations geographically a priori. If you run an unsupervised cluster analysis by genetic similarity with k = 5 or so, you'll obtain clusters roughly corresponding to Europeans, Sub-Saharan Africans, East Asians, Native Americans, or something like that. E.g., figure 2
here. Obviously this changes as you increase the granularity. But as you do so, you'll still find more similarity between clusters that had been grouped at lower resolutions. But this whole point seems weird to me. I think a lot of people would essentially be like "if you run k means with k = 1, that's ok. If you run it with k = 5, that's bad. If you run it with k = 100, that's ok."
There's been mention of how genetically divergent populations have historically tended to mix. This is totally correct and is a major blow to classic racial theories. Most modern Europe-dwellers are the descendants of Europe-dwellers and ancient steppe people who migrated into Europe in several waves. There is no primordial European race that's been isolated and static for tens of thousands of years. But that doesn't mean there aren't important genetic differences between ethnic English and Bantu peoples because you still need to go back tens of thousands of years to find common ancestors. Admixture is a fundamental part of the human story. But that even being possible implies the existence divergent groups that can admix in the first place.
Overall, the idea of a "black race" seems particularly bizarre. The San and Bantu are both parts of the "black race," but are as divergent as Europeans and Han Chinese. How does it make sense to call the Bantu and San members of the same race?