Abstract

An intuitive argument for scientific realism suggests that our successes in predicting and intervening would be inexplicable if the theories that generate them were not approximate y true. This argument faces many objections, some of which are briefly addressed in this paper, and one of which is treated in more detail. The focal criticism alleges that appeals to success cannot deliver conclusions that parts of science are true in the sense of truth-as-correspondence that realists prefer. The paper responds to that criticism, in versions proposed by Michael Williams, Michael Levin, and, especiaIly, Paul Horwich, by arguing that critics typically stop at a shallow level of psychological explanation. If we probe more deeply we discover a genuine explanatory role for correspondence truth