ion that renders so much social theory barely intelligible and certainly untestable. We sense, too, that we have rejoined the great co-operative effort to understand human society without illusions, which Aristotle launches with the opening sentence of his Politics: Observation shows us, first, that every polis is a species of association, and, secondly, that all associations are instituted for the purpose of attaining some good-for all men do all their acts with a view to achieving something which is, in their view, a good. Weber, no doubt, will refine and complexify our conception of what it is to act 'with a view to . . .', and will expand our knowledge of the things men have believed worth doing. But the Weberian programme is part of Aristotle's, a programme which Aquinas, in his reflective introduction to the Politics, defines as seeking knowledge of social wholes by careful consideration of their parts and their principia, their intelligible originating factors, particularly the (more or less) intelligent human doings (as distinct from makings) that actually constitute those social wholes, doings which are to be understood by understanding their motivating objectives. So: the empirical existence of that form of social relationship which we call authority (or 'domination') is understood when we distinguish the regularities of behaviour involved in authority-compliance relationships from those involved in 3 Aristotle, Politics 1:i i252ai. 4 Aquinas, In Octo Libros Politicorum Aristotelis Expositio, proem.