The Comparative Method and the Study of British Politics

Abstract

My title suggests paradox. How can you make comparisons when you are dealing with only one system? The question has poini and I am going to hinge these brief introductory remarks on methodology on a distinction between what I shall call method of comparison and method of hypothesis. When we first began discussing these questions of comparative politics quite some years ago, we talked a good deal about what we called the logic of In trying to imagine what procedure this logic involved, we would think of an interesting phenomenon such as two-party systems and then try to work out an orderly procedure for determining necessary and sufficient conditions for its existence. Frankly, in our projections of what we thought this method might achieve, prospects were somehow not nearly so bright as we felt they ought to be. Was not comparison social scientist's equivalent of natural scientist's laboratory? It was only when we added to our operations method of hypothesis that prospects began to brighten. We moved our stress from induction to deduction. We approached comparison through theory rather than approaching theory through comparison. From theory-from Aristotle or Hobbes, from Marx or Weber-we drew out hypotheses about important and interesting events and then sought to test them in particular contexts. In this way interesting questions about political development and about such things as role of elites, classes, and intellectual innovation were injected into comparative and historical studies. Of course, I am aware that two methods are complementary. When you have tested a hypothesis in one particular context you will probably go on to test it in another. If you have found hypothesized correlation you will want a control to assure you that connection is real, not accidental. Moreover, as you move on from case to case it is likely and desirable that your initial hypotheses or hunches will be developed and revised. Nevertheless, difference in emphasis can mean a difference in research strategy. It is perhaps mainly a difference in what you focus on at start: theory or fact. The former-the method of hypothesis-is, I think, particularly appropriate for study of individual systems. Sensitized in countless ways by your knowledge of theory, you examine some system or aspect of a system, draw out hypotheses that seem relevant, and test them against facts.