SummaryThe key terms in Kant's argument for the synthetic apriority of mathematics are analyzed. The result is a somewhat “idealized” interpretation of these terms, which, however, is appropriate in respect of Kant's main argument. Taking this interpretation as a framework, a model for giving evidence for numerical statements is presented, which is in good agreement with Kant's argument, and according to which numerical statements are indeed synthetic and also, in a sense,a priori.Thus they formally render counter‐instances to Hume's thesis that no general synthetic statements can bea priori.Nevertheless they do not play the part given to them in Kant's metaphysics, since they turn out not to be “genuinely” general, so that no Copernican revolution is required to explain their epistem