Metaphysics constitutes sets of statements or principles; methodology constitutes sets of rules. Metaphysics is constitutive; methodology is regulative. That, metaphysics gives birth to rules of research — scientific or pseudoscientific — was obvious for long; Wittgenstein and Popper, following a lead of Bertrand Russell, went further and tried to do away with the constitutive while retaining the regulative which might be necessary for scientific research. In Wittgenstein's case things obviously don't work. I shall venture to argue that Popper's case is also objectionable. (In the case of Wittgenstein, as well as in that of Popper, incidentally, I refer exclusively to the author's first and classic book.)