"Putting Theoretical Virtues to Work"
Abstract: The theoretical virtues (fruitfulness, explanatory power, consilience, etc.) are standardly invoked by scientific realists as a response to underdetermination: virtuous theories are more likely to be true, so the virtues can break epistemic ties between empirically equivalent rivals. Against realists, I argue this picture is mistaken on several grounds: any theory can be engineered to possess virtues, principled comparisons between theories that all have virtues are impossible to produce in advance, and false theories exhibit virtues in much the same way approximately true ones do. However, historical case studies also show that the virtues function epistemically at least sometimes and so, against anti-realists, I argue that it is a mistake to think of the virtues as merely pragmatic. As a better option to both, I develop the Epistemic Labour View, on which the virtues play a genuinely epistemic role not by being truth-conducive, but by doing work in promoting live scientific aims. The two main ingredients of this view are pluralism about epistemic goals and a distinction between a theory’s merely possessing a virtue and that virtue doing real epistemic labour. I illustrate the view with the historical debate over continental drift and draw some broader conclusions about how the virtue debate — and the realism debate more generally — should be reframed.