On poor and not so poor thought experiments. A reply to Daniel Cohnitz

Abstract

We have never entirely agreed with Daniel Cohnitz on the status and r?le of thought experi ments. Several years ago, enjoying a splendid lunch together in the city of Ghent, we cheer fully agreed to disagree on the matter; and now that Cohnitz has published his considered opinion of our views, we are glad that we have the opportunity to write a rejoinder and to explicate some of our disagreements. We choose not to deal here with all the issues that Cohnitz raises, but rather to restrict ourselves to three specific points. The first point concerns the way in which Cohnitz represents our views. He summa rizes them in five theses, (PA1)-(PA5), and while we have no dispute with the wording of (PA1)-(PA4), we consider (PA5) to be somewhat misleading. Putting into our mouths the declamation that In philosophy we cannot turn to crucial experiments (whereas in science we can) wrongly suggests that we take crucial experiments to constitute the pivotal difference between philosophy and science. At no place do we claim that disputes over thought experi ments in physics are always settled by carrying out a crucial experiment. Cohnitz writes: As Peijnenburg and Atkinson seem to admit, this [the settling by a crucial experiment] is not the case for all physical thought experiments. Seem to admit? Nay, we roundly asseverate the fact! After that sentence, Cohnitz continues as follows: