Or, Part CCLXXXIV of my long-running argument that knowledge requires much less by way of epistemic goodies than most epistemologists think.

Brad DeLong is teaching his 10 year old daughter chemistry.

bq. The Ten-Year-Old has thoughtfully gone off and is drawing pictures of how the eight electrons in the valence shell might “orbit” the nucleus, and wondering why eight electrons in the valence shell is a particularly stable configuration.

bq. She doesn’t know Coulomb’s force law. She knows no orbital mechanics. She definitely does not know that the solar-system model of the classical atom is self-contradictory. She knows no spherical harmonics. She knows no quantum mechanics. Yet, still, she now knows more about electrons and their impact on chemistry than anybody in the world knew a century ago, back before Niels Bohr.

All very cute and all, but shouldn’t some epistemologist be coming along around now saying that if she has that many false beliefs at the base of her chemical beliefs, then none of the superstructure really constitutes knowledge. After all, inference from inconsistent premises is hardly a reliable form of belief formation now, is it?