Michael Stuart (London School of Economics)
Imagination is part of the scientist's toolbox, but it is unlike the other tools. In experiment, mathematical reasoning, computer simulation, etc., we mostly try to filter out random fluctuations and unpredictable or uncontrolled events. But we would not recommend the same for imagination. Indeed, part of its power comes from its particular lack of regimentation; it is only in imagination we are encouraged to ignore what we know. In this paper I argue that cognitive science-based epistemologies of scientific imagination cannot account for this kind of epistemic luck. As a result, these accounts must introduce an element of epistemological anarchy.
