Scientific progress is normally portrayed as a series of stable, rational choices between competing theories, research programs, or research traditions within a discipline. Progress of that sort cannot occur in political science because any comparisons among research traditions can only proceed in the context of a set of empirical problems which are socially determined. Political scientists have a say in this determination—but so does the larger society in which political science is embedded. The rationality of any choice among research traditions is therefore historically contingent. The progress of political science may be reconceptualized as an increasing capacity to cope with contingency in its empirical problems.