In the late 1980s, Friedrich Kratochwil's and Nicholas Onuf's assertion that norms matter in international relations (Kratochwil, 1989; Onuf, 1989) sparked a major controversy. Since the mid-1990s, however, the question has no longer been whether norms matter but how they matter. Constructivist research, following Kratochwil and Onuf, postulates that norms constitute identity. Agents come to adopt and internalise – in short select – norms that define their identity. Rationalist research, by contrast, conceives of norms in a different way. Actors do not select norms. They may adopt them for strategic purposes but do not internalise them. Creating norms and complying with them is assumed to be the outcome of attempts by strategic agents to maximise their benefits.

At first glance, this debate is a contestation between adherents of the logic of appropriateness and proponents of the logic of consequences (March and Olsen, 1989). The one side interprets agency as the following of identity-constituting norms whereas the other understands agency as cost–benefit calculation. Yet a closer look provides a somewhat more nuanced picture. Constructivist research, trying to understand how norms get selected, has gone beyond a more or less static conception of appropriateness. While Constructivists have proposed a number of different processes through which agents come to select norms, the burgeoning literature on advocacy networks has featured most prominently. This literature is underpinned by the logic of argumentation.