Conspiracy theories and the hidden evidence penalty

Abstract Many regard conspiracy theories as intrinsically epistemically suspicious. Generalists defend this view, holding that there is something epistemically wrong with conspiracy theories as a class. Particularists deny it, contending that conspiracy theories must be evaluated on a case-by-case basis. This paper invokes a norm of theory choice from the philosophy of science and metaphysics to defend a weak form of generalism. According to the norm, we should prima facie disbelieve theories that posit undetectable entities. I develop a graded notion of undetectability, such that the more undetectable the posited structure, the greater the epistemic penalty a theory incurs. Conspiracy theorists, then, face a dilemma: either (1) posit a highly detectable (but undetected) conspiracy that should be rejected on evidential grounds or (2) posit a highly undetectable conspiracy that should be rejected on anti-undetectability grounds. Most conspiracy theories posit a high degree of secrecy, running into the second horn. This is illustrated through a range of examples, from “lizard people” to 9/11 “truthers.”