Methodological naturalists regard scientific method as the only effective way of acquiring knowledge. Quite the contrary, traditional analytic philosophers reject employing scientific method in philosophy as illegitimate unless it is justified by the traditional methods. One of their attacks on methodological naturalism is the objection that it is either incoherent or viciously circular: any argument that may be offered for methodological naturalism either employs a priori methods or involves a vicious circle that ensues from employing the very method that the (...) argument is aimed to show its credentials. The charge of circularity has also been brought against the naturalistic arguments for specific scientific methods; like the inductive argument for induction and the abductive argument for the inference to the best explanation. In this paper, I respond to the charge of circularity using a meta-methodological rule that I call ‘reflexivity requirement.’ Giving two examples of philosophical works, I illustrate how the requirement has already been considered to be necessary for self-referential theories. At the end, I put forward a meta-philosophical explanation of the naturalism-traditionalism debate over the legitimate method of philosophy. (shrink)