One of the most influential arguments against the possibility of drawing a principled fact-convention distinction consists in the insight that because our beliefs are necessarily evaluated together, any statement can be retained or given up in the face of experience. The purpose of this paper is to establish that this argument does not undermine Poincaré s conventionalism in virtue of the fact that this doctrine does not simply amount to the claim that there are principles that are immune to revision. It will be argued that Poincaré’s most important insight into the structure of theoretical knowledge is that there exist principles that cannot be empirically tested because they make possible the application of theoretical terms to experience...