Published online by Cambridge University Press: 01 January 2022
This paper develops a stronger version of ‘inference-to-the-best explanation’ scientific realism. I argue against three standard assumptions of current realists: (1) realism is confirmed if it provides the best explanation of theories’ predictive success; (2) the realist claim that successful theories are always approximately true provides the best explanation of their success; and (3) realists are committed to giving the same sort of truth-based explanation of superseded theories’ success that they give to explain our best current theories’ success. On the positive side, I argue that (1) the confirmation of realism requires explaining theories’ explanatory success, not just their predictive success; (2) in turn this task requires a richer realist model of explanation that brings into the explanans both (a) successful theories’ epistemic virtues (e.g., unification and simplicity) and (b) the standards governing these virtues, as well as truth; (3) this richer realist model is further confirmed because it can better explain the success of theories in gaining wide acceptance among scientists; and (4) the model is further supported because it is superior to ‘preservative realism’ in providing a plausible rebuttal of the pessimistic meta-induction from the many past successful-but-false theories to the likelihood that our best current theories are likewise false.
I am happy to acknowledge the comments on this research by friends and colleagues: Eric Barnes, Paul Churchland, Craig Callender, Eric Watkins, Stathis Psillos, Brad Wray, Timothy Lyons, Peter Lewis, Ed Erwin, and Harvey Siegel.