In Chapter 3 we presented a case for saying that principles of scientific method are generally of the following sorts: deductive and non-deductive principles of inference; values by which theory choices can be made; and methodological M-principles, which govern theory choice and can include values. On what grounds do we think that any of these are correct? There is some controversy about the grounds on which the rules of deductive logic are to be justified; some argue that they involve circularity since they are employed in their own justificatory arguments (see §6.1). Even more controversial are the grounds for accepting inductive inferences, values and M-principles. It seems appropriate to call the investigation into principles of method and their justification metamethodology. This chapter explores what this might be, and what grounds it can offer for accepting principles of scientific method.

In §4.1 we set out a “three levels” picture of the sciences, methods and metamethods. Objections have been made to metamethods on the grounds that their justification involves either circularity or an infinite regress. In §4.2 and §4.3 we outline some of the a priori and empirical approaches to metamethodology found in the philosophy of science that either circumvent these objections or show us how to live with them. In §4.4 we evaluate one common metamethodology: reflective equilibrium. This has been advocated by some methodologists as a way of bringing particular judgements about good theory choices into accordance with principles of a theory of scientific method.