Theory Change in Science

Abstract

Nowadays theories about science change even more rapidly than science itself. Not too long ago nearly everyone would have agreed that the primary method of criticizing scientific theories was falsification; (the disagreements were over what positive things could be said about theories which survived testing). Then Kuhn (1962) discovered normal science and Agassi (1966) argued that Boyle’s Rule was dogmatic (why should one always keep the experimental result and drop the theory?) and so it was suggested that the method of science was the removal of inconsistencies, preferably not by ad hoc stratagems. However Feyerabend (1962) soon showed that the demand for consistency was much too stringent since most progressive new scientific theories were inconsistent with the best theories and observational evidence available at the time. Lakatos (1970) then proposed a sort of Pollyanna theory of science — scientific research programs should be evaluated by counting only their successes and ignoring their failures. But Feyerabend (1970) argued that even this standard was too restrictive and claimed that the methods of science were counter-induction, proliferation, and “anything goes.”