Progress in the Study of International Conflict: A Methodological Critique

This paper is a philosophy of science critique of behavioralist studies on international conflict. We find that the vast majority of studies on conflict fall into one of two categories: hypothesis testing and model building. A number of hypothesis-testing studies involve hypotheses which are believed to be the conse quences of some theory. We argue that these studies will not generate scientific progress because they fail to appraise more than one theory at a time. There are also many studies where the hypotheses tested are not purported to be the consequences of some theory. We argue that these studies too are unable to generate real progress because they are incapable of yielding novel information and are therefore in capable of being severely tested. The model-building activities of others do not offer much hope for sci entific progress either. This is primarily due to the fact that model builders focus on what is essentially a problem for the mathematician, not the scientist. In addition, we find that model builders, even when subjecting their models to empirical analyses, actually fail to test any theories or universal generaliza tions. In the hope of furthering our understanding about the causes of international conflict, we suggest that hypothesis testers reformulate their hypotheses so that they can be subjected to severe tests. We also suggest a change in focus for model builders which hopefully will enable them to progress to better, more comprehensive theories of international conflict.