Some naturalistic conceptions of philosophical methodologies interpret the doctrine that philosophy is continuous with science to mean that philosophical investigations must implement empirical methods and must not depart from the experimental results that the scientific application of those methods reveal. In this paper, I argue that while our answers to philosophical questions are certainly constrained by empirical considerations, this does not imply that the methods by which these questions are correctly settled are wholly captured by empirical methods. Many historical cases (...) of successful answers to philosophical questions involve strategies and methods that allow the divergence from empirically established results. I develop this idea and then illustrate it with Frege's methodology in his Foundations of Arithmetic. (shrink)