Often, empirical observations on politics in sub-Saharan Africa and political science theories do not match. A discipline which to a large extent is based on top-down concepts derived from dynamics in Western European or North America is indeed bound to face difficulties when it tries to make sense of developments in the global South. Clearly this goes for concepts such as ‘the state’ or ‘civil society’. It is against this background that academic disciplines which work with a different set of methods and related research questions increasingly make inroads into territory in which political science has traditionally claimed competence. The volume Twilight Institutions, edited by Christian Lund, who is a professor at the Department of International Development Studies, Roskilde, Denmark, is a good example of this welcome trend.

The collection of eleven articles analyses, from a bottom-up perspective, how public authority actually works in an environment of weak statehood. The title of the volume refers to the paradoxical way in which many institutions in Africa are ‘not the state but exercise some form of public authority’ (p. 1). The articles were first published in 2006 in the journal Development and Change, and some originate from a workshop organized for Ph.D. researchers in 2002. Other contributions come from anthropologists already well established such as David Pratten, Lars Buur, Carola Lentz or Pierre-Yves le Meur. The case studies are drawn from Benin, Burkina Faso, Ghana, Mozambique, Senegal, Nigeria, South Africa and Zambia and cover a variety of state–society situations. The policy fields under review include the provision (or non-provision) of security, control of land, the politics of decentralization and the rule of law. The actors operating in these fields are local politicians, traditional authorities, civic associations, young men and external agencies.

This volume offers well-researched and original insights into an important area of politics. Compared to contributions from anthropology on politics in Africa in the 1990s, this volume takes a more fruitful direction: the political-scientist bashing that one usually comes across is getting more ritualistic. Rather the editor and some of his contributors have read political science research and take a keen interest in conceptual and theoretical debate. Although the editor does not aim at generalizing theoretical statements, one methodological conclusion from this volume can be singled out: since there are no general patterns emerging from the contributions, it is important to map how certain policy fields are organized. And, for this exercise, the post-disciplinary dialogue about concepts has just started.