Sam Andrews

All Humeans hold, roughly, that laws are informative summaries of non-lawful matters of fact. Pragmatic Humeans go further: for them, what makes these summaries the laws is their usefulness to agents like us. By adding elements of our specific epistemic interests and constraints, pragmatists contend, we can arrive at satisfying explanations of otherwise surprising features of our actual laws and our actual scientific practice. But the pragmatic shift is not without problems. The more elements of our particular psychology we add to our nomic formula, the more susceptible we are to idealistic ratbaggery. Intuitively, what can or must happen does not depend on our particular cognitive architecture: we cannot change the laws by changing us. But if laws are distinguished from accidents by features of our psychology, then changes in our psychology might lead to differences in what the laws of nature allow. My aim here is to clarify the role of pragmatic constraints, and thereby respond to this challenge from creeping idealism. My strategy has two parts. First, I argue that pragmatic constraints determine the laws only indirectly, by generating a nomic formula that itself makes no reference to agents. Second, I discuss what sorts of agents are agents like us. I argue that pragmatists should appeal to a particular sort of idealized agent, one whose specific limitations and interests have been idealized away.