ABSTRACT
Aristotle, as far as we know, has no conception of thought experiments. He does not discuss them in his works, and his writings do not show signs that he has identified a distinctive mode of pointing out something, philosophically or otherwise, that could plausibly be described in terms of what we call thought experiments. Aristotle does not even have a word for “thought experiment.” And it is not difficult to see why this is so. To start with, for our conception of “thought experiment” to make the specific sense it has (however vague it might be), it requires a context in which the concept of “experimentation” has some currency among a relevant community of thinkers. But in Aristotle’s time, there is no scientific community to which the conception of “thought experiment” could, as it were, speak: the notion relies on a specifically post-Aristotelian conception of experimentation, according to which experimenters contrive repeatable physical events, by means of which they demonstrate some correlation between physical states or events that are usually not available to immediate experience. An important aspect of this is that the experimenter produces her demonstration through exerting control over the relevant parameters (typically, with the aid of technical instruments) so as to isolate the occurrence of the phenomenon about which she wishes to demonstrate her point. The basic idea is that she, by wilfully manipulating the parameters, makes available to repeatable experience a regularity in nature that was previously not evident to her scientific community. According to this conception, experimentation is an artificial means to make available to repeatable experience hitherto unobservable facts about nature or, shorter still, a means to provide such data in otherwise unobservable natural terrain. Our notion of thought experiment depends on some such ideas of empirical experimentation. The difference is of course that thought experiments are not confined to physical facts, which is also why they do not require artificial production of physical events; what makes them experiments of thought is that repeatable imagined scenarios are used instead. 1 But, notwithstanding this difference, our notion of thought experiments seems to preserve what is basic for the early modern idea of empirical experimentation: they are imagined scenarios, wilfully contrived through manipulations of relevant parameters and with the 58purpose of isolating, and thus disclosing to experience, some previously non-evident fact; only that thought experiments, instead of providing us with experiences in an unqualified sense, provide us with what Ernst Mach calls a Gedankenerfahrung, an experience in thought (Mach 1905, 186).
