Abstract
There is an influential, ongoing debate between traditionalists and experimentalists about how to carry out conceptual analysis by means of the method of possible cases. The debate concerns whose intuitions are evidentially relevant to philosophical theories, and which methods are most appropriate for collecting such evidence. The aim of this paper is not to take sides in this debate, but to question the monopoly that the method of possible cases has in contemporary discussions of philosophical methodology. Since early analytic philosophy is replete with methods of conceptual analysis that make no essential use of intuitions, these discussions are inevitably myopic. In particular, it will be argued that Frege’s logical analysis of arithmetic, Hilbert’s axiomatic analysis of geometry, and Poincaré’s transcendental analysis of physics are non-psychologistic methods that yield profound philosophical conclusions. None of these methods are carried out from the philosopher’s armchair or the scientist’s laboratory; instead they properly belong to mathematics and physics. Thus, by ignoring these methods, both traditionalists and experimentalists have neglected the significant role that the exact sciences can play in conceptual analysis.