Intuition pumps and the proper use of thought experiments

I begin with an explication of 'thought experiment'. I then clarify the role that intuitions play in thought experiments by addressing two important issues: (1) the informativeness of thought experiments and (2) the legitimacy of the method of thought experiments in philosophy and the natural sciences. I defend a naturalistic account of intuitions that provides a plausible explanation of the informativeness of thought experiments, which, in turn, allows thought experiments to be reconstructed as arguments. I also specify criteria for distinguishing bad 'intuition pumps' from legitimate thought experiments. These criteria help us to avoid being seduced by the dangerous suggestive power of misleading intuitions.