Mortgaging Ethics to Empirical Claims
2026-08-09 · AI-generated insight
Lewontin's famous argument runs from a statistical finding to a moral conclusion: most genetic variation lies within populations, therefore racial classification has "no genetic or taxonomic significance," therefore no justification for its continuance. Winegard and Carl's note shows why the first step is vulnerable — variation at single loci can be small while correlated structure across many loci remains highly informative, just as no single facial feature distinguishes the sexes but faces taken whole do. If the moral conclusion was ever resting on the statistical premise, the premise has a crack in it.
Gould's quote, read carefully, is the escape route Lewontin declined to take. Gould explicitly separates equal treatment — an ethical principle — from empirical equality, which he calls a contingency: "it just worked out that way," and a hundred plausible histories "would have yielded other results (and moral dilemmas of enormous magnitude)." That parenthetical is the tell. Gould sees that if you mortgage your ethics to an empirical claim, every alternative history — or every new dataset — becomes a potential moral catastrophe. The dilemma only has "enormous magnitude" if you've made human dignity hostage to what the loci say.
The Coon anecdote shows the institutional consequence of taking on that mortgage. Once a moral position depends on an empirical claim, the claim must be defended by any means available — including a secret meeting to denounce a book almost nobody in the room had read. Condemnation without examination isn't an aberration here; it's the predictable behavior of people who believe that examining the question at all puts something sacred at risk. The anthropologists weren't protecting a finding, they were protecting a moral order they had welded to a finding.
The synthesis: the durable antiracist position is Gould's parenthetical made central — ground equal treatment in ethics directly, so that no correlation matrix, no cluster analysis, no future discovery can threaten it. Positions that instead follow Lewontin's route must eventually choose between honest revision and the AAA's raised hands.