There are immense differences not only as regards what counts as an explanation of social phenomena, but as regards the object of such an explanation. This last is often obscured by the uncritical assumption that the social sciences, including psychology, have as their object 'the explanation of behavior.' But if one looks even cursorily at the literature, this seems to cover a very wide range of objects including: explanation of the acts of kinds of individuals (or kinds of behavior), e.g., criminals or criminal behavior, and includes both the explanation of the behavior of 'groups', including gangs, mobs or 'social movements,' and of institutions, e.g., the behavior of governments or firms. Still others seem interested in explaining the properties of institutions, e.g., bureaucracies, or explaining 'system behavior,' including, e.g., social reproduction and transformation, conflict and crisis. Still others aim at explaining the powers and competences of persons and others have an entirely different object: call it the explanation of 'events', e.g., a war or revolution, or (as above) kinds of social events, including wars, revolutions and the like. One could go on here and include, e.g., explaining how selves are constituted, or how cultural objects function in maintaining solidarity, etc.As regards the character of social scientific explanation, there remains the lingering confusion between explanation and description, and an uneasy bifurcation between causal explanation and interpretation (Little, 1991;Skocpol, 1984). Moreover, there is debate over the character of both, e.g., between those who hold that a causal explanation requires appeal to covering laws, likely of a statistical sort (Coleman, 1990: 1) and those who, for very different reasons, deny this. For the former, who still strangely dominate mainstream thinking on such matters, any explanation which does not also produce predictability is at least an inferior explanation, if an explanation at all. There are still questions regarding what should be included in a proper causal account, including unclarity as regards the role of the past of and of structure and human agency in explaining outcomes. Some offer explanations in terms 'reasons,' or more narrowly in terms of 'rational choice' models of action. Still others offer what may be called 'structural' accounts, and some of these explain by appeal to