Email (record): We favor formal models of heuristics rather than lists of loose dichotomies: a reply to Evans and Over