The most critical question which a philosophical analysis of social ontology must answer is the linkage between the micro and the macro. Whether one conceives of these levels as ‘the individual and society’ or as ‘agency and structure’ or even as ‘choice and order’, an answer to the question of their mutual relations basically determines the kinds of social entities whose existence one is committed to believe. One family of solutions to the micro–macro problem relies on a reductionist strategy, either reducing the macro to the micro (microreductionism) or vice versa, reducing the micro to the macro (macroreductionism). The first strategy is often illustrated with classical or neo-classical microeconomics in which the key social entities are rational decision-makers making optimising choices constrained only by their budgets and ranked preferences. But the branches of microsociology born in the 1960s (ethnomethodology and social constructivism) are also microreductionist even if their conception of agency is quite different, based on phenomenology and stressing routine behaviour rather than rational choice (Garfinkel 2002; Berger and Luckmann 1967). Microreductionism does not imply disbelief in the existence of society as a whole, only a conception of it that makes it into an epiphenomenon: society is simply an aggregate or sum of either many rational agents or many phenomenological experiences shaped by daily routine. In other words, this macro entity does not have emergent properties of its own.