ABSTRACT
Let’s say that a theory is fictionalist concerning some region of thought if it claims that thoughts in that region are, can be, or should be regarded as possessing this similarity with fiction: they don’t have to be true to be good.1 This already allows for three versions, depending on which similarity with fiction is selected: (i) the thoughts are in fact regarded as like fiction; (ii) the thoughts can be regarded as like fiction; (iii) the thoughts should be regarded as like fiction. The first version (i), that the thoughts are so regarded, claims that ordinary thinkers of these thoughts do (for the most part) regard them as not needing to be true to be good, or at least with modest Socratic questioning could easily be persuaded to appreciate that this is how they had been, perhaps implicitly, treating these thoughts. The third version (iii), that the similarity with fiction should be regarded as holding, makes an independent claim: given every relevant consideration, the best thing to believe is that this similarity holds. This might be so even if the ordinary thinkers of the thoughts in question did not regard them as like fiction, and it might fail to be so even if they did. In these two cases, ordinary thinkers will be in error. The second version (ii), that the relevant thoughts can be regarded as similar to fiction, requires elaboration: what constrains how the thoughts can be regarded? The initial vague answer is
that the thoughts can be so regarded without detracting from the value (or the most important value) that they have. For example, in Field’s fictionalism about mathematics, the “can” amounts to: can without jeopardizing the applicability of science to the world.
