Abstract
We attack the SSK's rejection of the distinction between discovery and justification (the DJ distinction), famously introduced by Hans Reichenbach and here defended in a "lean" version. Some critics claim that the DJ distinction cannot be drawn precisely, or that it cannot be drawn prior to the actual analysis of scientific knowledge. Others, instead of trying to blur or to reject the distinction, claim that we need an even more fine-grained distinction (e.g. between discovery, invention, prior assessment, test and justification). Adherents of the SSK, however, maintain that the distinction is useless and perhaps nonexistent. We first argue against the assumption that the SSK's objection to the DJ distinction is just the same as Thomas Kuhn's. Second, we point out general weaknesses of the SSK's arguments against the DJ distinction. Finally, we argue that the distinction is useful not only in order to explicate what is meant by an evaluation but even for the empirical explanation of knowledge. We use two case studies from the history of cognitive science to support this point.