Good Intentions Are Not a Method: When Anti-Prejudice Becomes Its Own Bias
2026-07-29 · AI-generated insight
Read together, these four notes stage a quiet civil war inside the single file "Science Against Prejudice." Degler gives us the hero: Boas, whose relentless articulation of culture toppled the race concept precisely as segregation and eugenics were ascendant. That is science serving virtue. But the other three notes ask an uncomfortable follow-up: what happens when the virtue starts steering the science?
Warne's account of Gould supplies the cleanest irony. Gould accused Samuel Morton of unconsciously fudging cranial data to flatter his racial beliefs — and the 2011 re-analysis accused Gould of doing exactly that, massaging Morton's numbers to fit an anti-racist conclusion. Bias turns out to be symmetric. The man who wrote the canonical exposé of measurement bias may have been its next exhibit. Motive did not immunize method.
The Coon episode names the mechanism outright: an entire professional body voting to denounce a book almost none of them had read. The cause was righteous; the epistemic act — condemnation without examination — was indistinguishable from the dogmatism it opposed.
And notice Lewontin's move: he establishes an empirical claim (most human variation is within groups) and then glides, in the very next breath, to "no justification can be offered for its continuance." The first sentence is genetics; the second is politics wearing genetics as a badge. Edwards later called this slide a fallacy precisely because low variance on randomly chosen markers doesn't survive correlation across many variables.
The pollinated insight: Boas is celebrated and Gould, the AAA, and (arguably) Lewontin are indicted — yet all were fighting prejudice. The file's real lesson is that the nobility of the enemy you're fighting purifies nothing. A good cause is the most flattering disguise a bad inference can wear.