Consensus by Acclamation: When the Vote Precedes the Reading
2026-08-13 · AI-generated insight
Lewontin's quote contains a sleight of hand worth watching in slow motion. His empirical claim — that most genetic variation lies within populations rather than between them — is true. But he pivots without pause to a normative conclusion: racial classification has "no social value," is "positively destructive," and therefore "no justification can be offered for its continuance." The science ends mid-paragraph; the politics finishes it.
Winegard and Carl supply the technical rebuttal that later became known as Lewontin's fallacy: single loci considered one at a time may vary little between groups, but correlated differences across many variables allow surprisingly accurate classification — 80% accuracy from two skull measurements alone. The within-group variation statistic is true and yet the inference drawn from it fails, because structure lives in the correlations, not in any individual gene.
But the more interesting question is why the flawed inference held for decades, and the Coon anecdote answers it. When the American Anthropological Association voted to denounce Putnam's book, exactly one member in the room had read it. The consensus wasn't formed by examination; it was formed by acclamation, and the vote proceeded anyway. This is the mechanism by which a true statistic hardens into a false orthodoxy: once a position is understood to be the morally correct one, checking its logic starts to look like defection. Reading the book becomes suspicious.
Reich, writing half a century later, is trying to unwind exactly this dynamic — acknowledging the moral fears that motivated Lewontin's move while insisting that geneticists can no longer maintain the empirical half of the bargain. What's striking is that he must argue against the orthodoxy's method (anxiety about inquiry itself) more than against any specific finding.
The synthesis: bad conclusions from good data survive not because the fallacy is subtle — Edwards' rebuttal to Lewontin is straightforward — but because the social machinery for enforcing them, vividly on display in that AAA meeting, punishes the act of checking. Condemnation without examination isn't a lapse in the process. It is the process.