A Semantics and Methodology for Ceteris Paribus Hypotheses

ABSTRACT. Taking seriously the arguments of Earman, Roberts and Smith that ceteris paribus laws have no semantics and cannot be tested, I suggest that ceteris paribus claims have a kind of formal pragmatics, and that at least some of them can be verified or refuted in the limit. 1. Many, perhaps most, of the results of scientific inquiry are generaliza-tions towards which we have a peculiar attitude: although their apparent logical form is general, and we assert them and use them in prediction, in designing experiments, in deriving new relationships, and in reject-ing contrary hypotheses, they have actual or possible (feasibly possible) counterexamples, we know they have counterexamples, and we cannot say precisely under what conditions counterexamples will not arise in the future. The claims hold in “normal ” conditions, whatever those are, but not otherwise. Earman, Roberts and Smith (this issue; hereafter ERS) call such claims “putative ceteris paribus laws”. Putative ceteris paribus laws occur