Abstract

The paper joins Horwich’s criticism of stipulationist accounts of a priori knowledge, and raises some problems for his own account of the a priori. It first questions the assumed separability of scientific investigation and non-scieentific assertoric practices in regard to norms of adequacy. It also questioned Horwich’s Restriction Assumption according to which only the former are answerable to the standards of empirical adequacy and overall simplicity (which threaten apriority in the case of science). Finally, it criticises his argument that inability to think otherwise might guarantee apriority, pointing to science-driven reflective revisability of possibly innate beliefs.