We’re going in circles a bit here — we disagree only on the definition of “race,” not on anything of substance.

Just a few quick loose ends.

First, I don’t think it’s consistent to say that race “is only identifiable insofar as other people perceive it,” yet ethnicity can be “immutable.” This is only true if you define race as “whatever people perceive race to be,” and de-link it from the necessity that people of a race must be of “the same stock” (as per Merriam-Webster).

Also, on levels of specification — this is why I like Sailer’s definition. The level of specificity needed depends not only on where one is in the world, but on context of the discussion. If a very specific tribe is susceptible to a certain disease, it’s useful to look at race at that level. If we’re making broad generalizations about people from various continents, that level works.

Is race a vague concept? Sure. Can we “socially construct” different ways of classifying people, and of naming those classifications? Of course. But under any usable definition of race, such classifications are rooted in common ancestry, which is genetic in nature — thus, it is false to say DNA is “not integral” to the concept of race.