The Limits of Behavioural Explanation in Politics

Abstract

The unique contribution of American scholarship to the study of politics is undoubtedly the development of behaviouralism. As “a successful protest” against the limits of traditional political science, the behavioural approach has certainly broadened our perspective by calling attention to hitherto neglected areas of political research. It has also made us methodologically more sophisticated by its insistence on rigorous empiricism and on theory-building as a major task. And yet it is difficult to be convinced that “all segments of political science can be treated behaviorally,” as one student has recently claimed. Behaviouralism is certainly an effective way of studying politics. But it is quite a different thing to say that there exists no aspect of politics which cannot be treated behaviourally. It is the aim of this paper to enquire whether there may not exist segments of politics which by their very nature resist the behavioural treatment.Let us first define what is to be understood by behaviouralism. In view of its inherent “ambiguity,” it may be best to start by listing three major assertions which I think set apart behavioural explanations from other kinds of statements in political studies. These assertions are: (1) individuals rather than groups constitute proper units for analyses; (2) facts must be separated from values; and (3) legitimate explanations always run in terms of laws or generalizations, but never in terms of descriptive statements of particular occurrences.