On the very idea of a thought experiment
Abstract
Goffi and Roux are interested in what makes some thought experiments work, while others do not work. They do not attempt to draw an a priori line between two types of thought experiments, but rather ask the following question: inasmuch as thought experiments are arguments, and notwithstanding the fact that some of them might involve the contemplation of an imaginary scenario, how is it that some of them work, while others do not? Taking inspiration from a counterfactual thought experiment presented by Nicholas Rescher, they treat thought experiments as argumentative procedures resembling tests of consistency, which invite the experimenter to seek the weakest link in her body of beliefs. Equipped with this method, they examine two well-known successful thought experiments (Galileo’s two bodies strapped together, and Thomson’s violinist) and discuss Mach’s notion of thought experiments. Thus they reach the hypothesis that successful thought experiments respect the three following conditions: they do not deal with things, but with beliefs; they mobilise a set of beliefs shared by the interlocutors; and this set of beliefs has a hierarchical structure. Using once again examples written at different periods and taken from various disciplines (Descartes’ receding bodies, Aristotle’s weaving shuttles), Goffi and Roux argue that each of those conditions is individually necessary for a thought experiment to work. They finally conclude on the limits and consequences of their approach.