Functional thought experiments

The literature on thought experiments has been mainly concerned with thought experiments that are directed at a theory, be it in a constructive or a destructive manner. This has led some philosophers to argue that all thought experiments can be formulated as arguments. The aim of this paper is to draw attention to a type of thought experiment that is not directed at a theory, but fulfills a specific function within a theory. Such thought experiments are referred to as functional thought experiments, and they are routinely used in applied statistics. An example is given from frequentist statistics, where a thought experiment is required to establish the probability space. It is concluded that (a) not all thought experiments can be formulated as arguments, and (b) the role of thought experiments is more general and more important to scientific reasoning than has previously been recognized.