ABSTRACT
I am happy to support Mauricio Suárez in his programme to develop and defend experimental realism. In particular I am happy to endorse as a sound rule of thumb his doctrine that good causal evidence can be expected to trump derivation from even very well-confirmed theory. Suárez here provides serious, careful answers to objections to my closely related claims in support of entity realism over theoretical realism, but he does not explain in any detail why causal evidence, or directly relevant evidence from a good experiment, will generally weigh more than deductions from a good theory. Here I shall review some of the reasons for expecting this to be true. In any particular case, though, I would suppose that the issue will hinge on how good the experiment is versus how good the theory and how secure the deduction. Importantly, it matters how well-confirmed the theory is for cases very like the one in question, where I mean “like” in well-understood ways; what matters about the confirmation of the theory is not variety of evidence but rather the weight of evidence for the very particular specification of the theory used in the derivation.
