Experimental success and the revelation of reality

the miracle argument for scientific realism

Martin Carrier

pp. 137-161

in: Martin Carrier, Johannes Roggenhofer, Günter Küppers, Philippe Blanchard (eds), Knowledge and the world, Berlin, Springer, 2004

Abstract

The paper addresses the so-called miracle argument in favor of scientific realism and examines the viability of scientific realism as an explanation for the success of science. Scientific realism is committed to the claims that the theoretical terms in the mature sciences typically refer to real objects and that the theoretical laws in such sciences are typically approximately true. Instrumentalism or non-realism draws on the principles that factual claims need to be confirmed empirically and that experience fails to single out true assumptions. The miracle argument says that if a theory referred to fictitious objects, it would be miraculous that it is able to correctly predict observable effects. This argument transforms scientific realism into a hypothesis that is testable by the history of science. I perform such tests and conclude that the only type of realism that appears to be in agreement with the historical record is "realism of natural kinds." Theories that enjoy distinguished explanatory success truthfully establish equivalence relations among phenomena.