Is race just a social construct?

(Hat tip: John Ray.)

An unrealistic typical inside-the-dogma-boundaries debate about race, this one at The New Republic, ellicits a response from Razib at Gene Expression. Razib says race is not simply an ideological construct.

Race is a social construction. But it is not one constructed purely from human ideology. That many perceive Greeks as white and Turks as non-white is a reflection of social axioms (Christians are white, Muslims are brown). That may perceive Greeks as white and Thai as non-white is not a reflection of social axioms (Greeks exhibit physical characteristics of the white race, Thais do not). Humanists are well schooled in the interplay between ideology and facts in generating a narrative of the world. To pretend as if there is no factual basis in the outlines of an ideology is a denial of reality, which would less concerning if not for the fact that most Americans parrot this very line about race as if it was universally accepted.

The whole argument over "race is a social construct" gets very interesting in places. One of the more curious turns is when it mutates into "race is only a social construct."

To which the response must be: "Therefore what?"

"Therefore," it seems to me, the whole notion of race is a phantasm, and should be completely in any discussions of social policy. Yet it seems the folks most likely to insist that race is only a social construct are the ones most likely to insist that this social construct be considered in such things as university admission, job qualification, and even in the criminal justice system (e.g., the death penalty is unfair because members of one social construct are executed more often than members of some other social construct.)