II The Error - Statistical Philosophy

Despite the challenges to and changes in traditional philosophy of science, one of its primary tasks continues to be to explain if not also to justify, scientific methodologies for learning about the world. To logical empiricist philosophers (Carnap, Reichenbach) the task was to show that science proceeds by objective rules for appraising hypotheses. To that end many attempted to set out formal rules termed inductive logics and confirmation theories. Alongside these stood Popper’s methodology of appraisal based on falsification: evidence was to be used to falsify claims deductively rather than to build up inductive support. Both inductivist and falsificationist approaches were plagued with numerous, often identical, philosophical problems and paradoxes. Moreover, the entire view that science follows impartial algorithms or logics was challenged by Kuhn (1962) and others. What methodological rules there are often conflict and are sufficiently vague as to “justify” rival hypotheses. Actual scientific debates often last for several decades and appear to require, for their adjudication, a variety of other factors left out of philosophers’ accounts. The challenge, if one is not to abandon the view that science is characterized by rational methods of hypothesis appraisal, is either to develop more adequate models of inductive inference or else to find some new account of scientific rationality. (Mayo, 1996, p. ix)