Realism and the Failure of Science: Reversing the No Miracle Argument

Scientific realism is generally believed to be supported by an argument from the success of science, also known as the no miracle argument (NMA). As Putnam put it, "Realism is the only philosophy that doesn't make the success of science a miracle (1975, p.73).The idea is simple and at first sight indisputable. Theories invoking entities such as electrons or genes are successful in explaining and predicting numerous phenomena.If electrons and genes exist and what our theories tell us about them is true (or approximately true), then the success of these theories makes sense. But if these entities do not exist, or what science tells us about them is false, then this success is totally incomprehensible-a miracle. I would like to reverse this argument by arguing first, that realism is not unique in providing an explanation of the success of science, and second, that the occasional failures of science provide a much better argument for realism than its success. NMA also figures in the debate over the optimistic versus the pessimistic meta-induction. The optimist proposes an inductive inference from the numerous successes of science in the past to the likelihood of its success in the future. And similarly for the pessimist, except that here past failures of science, rather than successes, are projected onto the future. These inferences are mostly discussed in the context of the controversy about scientific realism. Both realists and their opponents employ an inference to the best explanation: the truth of scientific theories is claimed (by champions of the optimistic meta-induction) to constitute the best explanation of their success, whereas their falsity is claimed (by upholders of the pessimist meta-induction) to constitute the best explanation of their frequent failures. Armed with NMA, and with historical data about the successes or failures of science, optimists use their meta-induction to support realism while pessimists use theirs to undermine realism.