Abstract

The standard underdetermination argument relies on the assumption that empirical evidence is the only epistemic constraint on theory-choice. One prominent response to this has been the invocation of theoretical virtues, properties of our scientific theories that scientific realists take to be epistemic in nature and that are such that, if they are had by our theories, make it more likely for those theories to be true. It thus becomes a main goal for scientific realists to establish a link between the theoretical virtues and truth. However, despite the fact that the virtues are frequently invoked in response to underdetermination arguments, how to do so has not been addressed in any detail. In this chapter, I will spell out the most promising approach in making this link: an account according to which the question of whether the virtues are epistemic is answered empirically, through an examination of cases of epistemically successful (and unsuccessful) theories. I will illustrate how this approach works concretely by discussing a case-study from mid-nineteenth century Britain concerning the debate about the transmissibility of puerperal fever. After showing how some specific virtues are put to work in this case, I will explain what is required in order to make a more general argument for the epistemic potential of the theoretical virtues. I then argue that putting the virtue-question on empirical grounds is enough to undermine the anti-realist position on underdetermination – regardless of whether, in fact, the virtues turn out to be epistemically potent or not.