How can we make complexity work as part of a programme of engaged social science? This article attempts to answer that question by arguing that one way to do this is through a reconstruction of a central tool of a distinctively social science – the comparative method – understood as a procedure for elucidating the complex and multiple systems of causation that generate particular trajectories towards a desired future from the multiple sets of available futures. The article distinguishes between ‘simplistic (...) complexity’ and ‘complex complexity’. ‘Simplistic complexity’ seeks to explain emergence in complex systems as the product of simple rules and defines complex science as the process of establishing such rules. It can and does serve as the basis of technocratic social engineering in the interest of the powerful. In contrast ‘complex complexity’ recognizes the significance of social structure and willed social agency and does not reduce emergence to the mere working out of a restricted set of rules. Research programmes informed by this second approach must necessarily engage with social actors in context – they must be dialogical. This opens up the possibility of ‘complex complexity’ as a frame of reference for action-research directed towards the transformation of complex social systems. Comparative methods, and in particular Ragin’s qualitative comparative analysis approach, when deployed as part of such a programme, can provide meaningful information about the range of possible futures and the different configurations of causes which might generate particular desired social outcomes. (shrink)