In this paper, David Zarefsky suggests some constraints that political arguers face when trying to persuade an audience, and discusses some of the devices with which they respond to these constraints. In his treatment of these devices Zarefsky makes use of the concept of strategic manoeuvring as proposed by van Eemeren and Houtlosser. By taking into account the three manifestations of strategic manoeuvring-topical potential, audience adaptation and an effective presentation (van Eemeren and Houtlosser 2002, p. 139)-he identifies and discusses several possible ways of dealing with these situational constraints. Regarding the 'activity type' (van Eemeren and Houtlosser 2005) of political argumentation, Zarefsky focuses on large and open-ended public debates that engage entire societies. He rightfully indicates that it seems strange to consider these kinds of political argumentation as a specific kind of institutionalised discourse: political argumentation is in principle unregulated, free-form and requires no technical expertise of its participants in the discussion. In order to be able to discuss strategic manoeuvring within this kind of political context, characteristics of political argument need first to be specified. Zarefsky mentions four characteristics that can be of help to define the genre and to establish its conventions. In these comments, I will focus on the first part of the paper, which is about these characteristics of political argumentation: as a supplement to Zarefsky's paper, I will give a tentative analysis of how the four characteristics mentioned constrain the possibilities to manoeuvre strategically. 76-77). The activity type, therefore, may