When You Ground Equality in Data, You Make It Falsifiable
2026-08-24 · AI-generated insight
Derbyshire's taxonomy sorts thinkers by where they anchor human equality: Religionists in God-given attributes, Culturists in the psychic unity of a blank slate, Biologians merely in shared species membership — explicitly allowing "regional differences." Read against this grid, Lewontin is doing something stranger than either note admits.
Lewontin is a biologist making a Biologian-style argument (genetic variance statistics) in service of a Culturist conclusion (racial classification has no significance, equality is empirically vindicated). His final move gives the game away: the claim that classification "is positively destructive of social and human relations" is a moral judgment, not a finding — but it's presented as flowing from the variance data. He has converted equality from an axiom into an empirical hypothesis.
That conversion carries a hidden cost that Derbyshire's framework makes visible. If equality rests on the statistical claim that between-group variation is small, then equality becomes hostage to the data. The tag on Lewontin's note — correlation across many variables — points at exactly the weak joint: Edwards' famous rebuttal showed that while any single locus varies little between groups, the correlation structure across many loci classifies individuals by ancestry with near-perfect accuracy. Lewontin's statistics were right; his inference was wrong. And because he had staked the moral conclusion on the statistical premise, the moral conclusion looked refuted along with it — a disaster entirely of his own architectural making.
Derbyshire, meanwhile, isn't neutral either. Calling Biologism "just ordinary scientific method applied to humanity" while giving the other two views psychological and historical genealogies is rhetorical sleight of hand — his own view is the only one that apparently needs no explanation.
The real lesson sits between the two notes: the three "theories of human nature" are less rival descriptions of the world than rival strategies for insulating moral equality from facts. The Religionist puts it beyond biology; the Culturist puts it beyond genetics; Lewontin tried to put it inside genetics and thereby exposed it. The most defensible position may be Derbyshire's Biologian one, oddly enough — accept whatever regional variation exists, and refuse to let moral standing depend on it in the first place.