The aim of this paper is to develop the notions of particularism and generalism in argumentation theory. Generalism is the claim that the very possibility of argumentation depends on the existence of general rules that establish which data support which conclusions, while particularism is the negation of that claim. The problem is that it is often not clear what these rules consist of and in what sense argumentation depends on them. To clarify these issues, I will first show that the generalist can defend his position by arguing: (1) that most argumentative practices are generalist; (2) that it is better to argue by appealing to general rules; and (3) that it is not possible to distinguish between good and bad arguments without general rules. The particularist can accept (1) and (2), but not (3). The challenge for the generalist is to propose a conception of rules that supports (3). In the second place, I will distinguish various ways of understanding general rules using the criteria of logical function and scope. This will allow me to explore various interpretations of generalism and, from there, to propose the most promising version of the particularist thesis.