Abstract

I elaborate and defend an epistemic reading of the indispensability argument, which allows us to fully embrace the epistemic contribution of mathematics to the empirical success of science without holding that such contribution amounts to a reason for the reality of mathematical entities. I argue that the indispensability argument is not an instance of inference to the best explanation. After highlighting the distinction between physical and mathematical posits, I contend that even though inference to the best explanation may work in postulating unobservable physical posits, we lack reasons to believe that it performs the same role in view of mathematical posits.