First, I didn’t claim that there wasn’t any relationship at all between race and genetics. Clearly there is a relationship, though an attenuated one. But when I said that the dividing lines between races aren’t based in science, I meant exactly that — that the racial categories in use today predate any of the scientific claims that are made for their salience, and thus map only partially to subsequently invented or discovered “racial” genetic groupings.

A couple of examples will indicate what I’m getting at here:

First, the geographical borders of various racial groups’ ancestral homelands are by no means fixed or settled, as the debates over the racial identity of the accused Boston bombers recently showed. There are no bright lines on the globe or in our DNA that mark the borders between racial categories, and yet bright-line distinctions have always been essential to American racial ideology.

Those fictitious bright lines are even more central to racial ideology when we consider people of mixed ethnicity. To claim Barack Obama or Malcolm X as “black” is culturally commonsensical, but ancestrally reductive. Given our country’s history of the color line and the one-drop rule, a divergence between cultural and ancestry-based racial taxonomies is inevitable.

That should make my own position a bit clearer. Now onto the NIH…

The first thing that needs to be noted about that study is that it wasn’t based on a random or representative sample of people. The participants were drawn from relatively homogeneous communities, and individuals who self-identified in ways that diverged from the racial taxonomy of the study’s authors were excluded from the sample.

Secondly, it should be made clear that the “clustering” seen in the NIH data was clustering around an observed norm rather than a predictive one. The NIH didn’t go in with a genetic profile for each race, in other words, and measure how well individuals matched up. Instead they examined groups of people according to their self-reported racial identity and conducted a genetic analysis intended to determine how cohesive those groupings were.

Nobody is denying that people from the same ethnic background are likely to share more genetic commonality than people of different ethnicities. That’s neither a novel nor a controversial claim. The question at hand is whether such ethnic “clustering” can be productively understood and deployed as evidence of unambiguous and salient racial distinctions, and that’s a question on which the NIH report is silent.

Here’s an illustration of what I mean.

In the US, it’s common to think of sickle cell anemia, a genetic condition, as a “black disease,” and in fact statistics on prevalence bear that out — black Americans are far more likely than whites to carry the sickle cell gene. But that fact, it turns out, is a result of ethnicity and history, not race.

Sickle cell is common in some parts of Africa, and some parts of Europe, but not others. As it turns out, most American blacks have ancestral origins in areas of sickle-cell prevalence, and most American whites do not. But if the geographic distribution of Americans’ ancestors were different — if, for instance, the country had been settled by South African blacks and Sicilian whites — the incidence of sickle cell in the white population would be higher than the incidence in the black population.

Race is a form of shorthand, in other words. It’s an approximation. In some situations, for some purposes, it’s a useful approximation. If you’re trying to tell someone which of your several friends named Jim you’re referring to, specifying that you mean “the white Jim” may be helpful, and if you’re trying to get the most bang for your buck in a sickle-cell awareness media campaign, targeting black media may have merit.

But the fact remains that Nelson Mandela is at less risk of sickle cell than Al Pacino.

“Nothing” might be a bit of an overstatement. If you happen to let facts bother you.

“Subjects identified themselves as belonging to one of four major racial/ethnic groups (white, African American, East Asian, and Hispanic) and were recruited from 15 different geographic locales within the United States and Taiwan. Genetic cluster analysis of the microsatellite markers produced four major clusters, which showed near-perfect correspondence with the four self-reported race/ethnicity categories. Of 3,636 subjects of varying race/ethnicity, only 5 (0.14%) showed genetic cluster membership different from their self-identified race/ethnicity.”
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1196372/