Philip Kitcher’s account of scientific realism in 'The Advancement of Science' (AS) differs from his account in 'Science, Truth and Democracy' (STD). We demonstrate that (1) contrary to appearance, Kitcher in AS proposes a so-called Kantian realism that is accompanied not by a correspondence theory, but by a hybrid conception of truth. (2) Also, we point out that Kitcher does not pertain to the “promiscuous realism” proposed in STD stringently, but falls back on his Kantian realism of AS at points. (...) Here, we question Kitcher’s claim that his promiscuous-realist conception stems initially from commonsensical be-liefs. (shrink)