Abstract

It is an essential part of Kant's conception of regulative principles and ideas that those principles and ideas are in a certain sense indeterminate. The relevant sense of indeterminacy is cashed out in a section in the Antinomies where Kant says that the regress of conditions of experience forms not a “regressus in infinitum” but a “regressus in indefinitum.” The mathematics that Kant appears to rely on in making this distinction turns out to be problematic, as Jonathan Bennett showed long ago. But I suggest that despite this, there is another mathematically legitimate way to make Kant's point, one enunciated by, among others, Michael Dummett. This reading is corroborated, I suggest, by Kant's conception of reason as a radically open-ended endeavor.