How should critical realism affect the practice of social science? This paper responds to this and related questions by suggesting some methodological implications of the realist theory of emergence. Given that critical realism understands causation as the interaction of emergent causal powers, and that the theory of emergence describes the type of structural relations that underpins such powers, we can practise social ontology by seeking to identify these structural relations in the social domain. Such methods, however, cannot stand or fall (...) purely on philosophical grounds; their validity also depends on whether they work. Hence this paper briefly illustrates the application of the method to social ontology, using examples from the theory of social institutions. (shrink)