Thinking About Thought Experiments in Physics Comment on “Experiments and Thought Experiments in Natural Science”

The main claim of Atkinson’s paper is that “… thought experiments in physics are of value only when they are related to or inspire real scientific experiments”. The paper illustrates its thesis by analyzing three examples: Galileo’s thought experiment intended to show that, contrary to Aristotle’s view, bodies of different weight do not fall with different speed, the EPR experiment designed to prove incompleteness of quantum mechanics and, as a third example, string theory. The claim’s tacit assumption is that there is a well defined and sharp difference between thought experiments and “real” scientific experiments. I call this assumption tacit because no attempt is made in the paper to spell out systematically the conceptual difference between thought as opposed to “real” experiments. Lack of an epistemic specification of this difference takes much of the bite out of the claim, and as a result the claim remains somewhat weak and vague. Rather than trying to make Atkinson’s claim sharper, in what follows I comment on the three examples Atkinson analyzes. In Atkinson’s evaluation Galileo’s thought experiment does not disprove either the weak (qualitative) or the strong (quantitative) Aristotelian dogma concerning the rate of fall of heavy bodies. Why? Because in Atkinson’s interpretation Galileo’s thought experiment is an attempt to give an a priori argument against the Aristotelian dogma, which Atkinson takes to be a statement that can only be verified (or falsified) empirically. On such an interpretation of Aristotle’s dogma and of Galileo’s thought experiment Galileo’s attempt is futile indeed, and there can hardly be any disagreement about this. But the details of Atkinson’s analysis of the Aristotelian dogma and of Galileo’s thought experiment raise a few questions. Atkinson argues that the Aristotelian dogma cannot be proved to be logically inconsistent (as Galileo thought to have succeeded in showing) because under certain physically realizable