Normative sociology and normative behaviourism: Recent discussions of how empirical social science and normative political theory can inform each other

The relation between social science and political theory has traditionally been understood based on a classic fact/value distinction according to which the role of social science is to describe and explain and the role of political theory is to articulate and examine arguments for normative conclusions. Political theory should take account of relevant social science. But the contribution of social science to political theory has traditionally been understood as being purely descriptive. The question is whether social science is also relevant for the normative aspects of political theory and, if so, how. The paper examines two recent developments that both address this issue. One is Jonathan Floyd’s proposal for ‘normative behaviorism’, according to which the normative claims of political theory should be grounded on descriptive claims from social science. The other is Tariq Modood’s description of his own work as a form of ‘normative sociology’, which includes constructive normative reasoning as part of sociological investigations. Normative behaviorism seeks to base normative claims on descriptive social science, whereas Modood wants to include normative reasoning in sociology. But both approaches nevertheless question the classic division of normative and descriptive. Floyd challenges this distinction by arguing that normative claims about how we should live cannot be based on what he calls ‘mentalism’, i.e., the thoughts we have, but should rather be based on observations about how people actually behave. Modood argues that the standard division of labor between sociologists and political theorists should be rethought so that sociologists should acknowledge their normative commitments and argue explicitly for them. Floyd’s position may seem to be about a fundamental issue, and thus more radical, whereas Modood’s position can in principle be sustained without revising the classic division of descriptive and normative claims. The two approaches see different kinds of social science facts as relevant. Floyd argues that behavioral patterns involving insurrection and crime are especially relevant as grounds for normative principles, whereas Modood argues that discussions of principles of multicultural equality should pay special attention to the expressed views of members of the minorities in question. The paper considers the merits and problems facing each of these proposals – and how each approach casts critical light on the challenges facing the other. The paper focuses on how normative behaviorism and normative sociology see the relation between descriptive social science and normative political theory and how normative behaviorism and normative sociology relates to each other as views about this relation. More specifically, the paper examines what the two views imply for how to ground normative statements in political theory. Both Floyd and Modood have resisted criticisms that their approaches involve naturalistic fallacies by noting how behavior already can be seen as expressions of normative preferences, in Floyd’s case, or how practices already strongly associated with networks of norms, in Modood’s case. So one way of understanding both normative behaviorism and normative sociology is that they do not move from ‘is’ to ‘ought’ but offer specific ways of systematizing the normative content of specific social phenomena.