Evaluating theories

John Vasquez claims to follow Imre Lakatos but distorts his criteria for judging theories and evaluating research programs. Vasquez claims that facts observed can falsify a theory by showing that its predictions are wrong. He fails to consider the puzzles posed by the interdependence of theory and fact. He places all realists in a single paradigm despite the divergent assumptions of traditional and structural realists. In contrast to Vasquez, I argue that explanation not prediction, is the ultimate criterion of good theory, that a theory can be validated only by working back and forth between its implications and an uncertain state of affairs that we take to be the reality against which theory is tested, and that the results of tests are always problematic.