%0 Journal Article %J Philosophica %D 2003 %T The dual nature of thought experiments %A De Mey, Tim %X

It is often assumed that thought experiments are either experiments or arguments. In this paper, I argue that this disjunction is not an exclusive one and that we can only account for the demonstrative force or evidential significance of thought experiments by conceiving of them as both experiments and arguments. First, I distinguish between three related, but nevertheless distinct problems that thought experiments in physics pose. In this context, I also show that the choice between an experiment view and an argument view on the nature of thought experiments pops up in connection with two of those problems: (1) the problem of the source of thought-experimental knowledge and (2) the problem of the evidential significance of thought experiments. Subsequently, I argue that as far as the issue of evidential significance is concerned, we should at least recognise that thought experiments have a dual structure: an experiment-like set up and an. argument-like winding up. Then I introduce Hans Radder's conceptual framework for the analysis of "real" experiments and I apply it to thought experimet:J.ts. Finally, I argue that Radder's distinction between the aspects of 'theoretical description or interpretation" and "material realization" allows us to move on from the recognition of the dual structure of thought experiments to a full-blown dual nature view of them.

%B Philosophica %V 72 %P 61–78 %G eng