On Structure and Naturalness in Dialogical Models of Argumentation

It is recognized by researches from various disciplines (e.g.: informal logic, artificial intelligence & law, legal theory, computational dialectics) that argumentation is a process and that this process can adequately be modeled by means of a dialog. This paper evaluates three of these dialogical models: the best-known in general (MacKenzie’s DC), the best-known in AI & Law (The Pleadings Game), and a recently fully worked out AI & Law model (DiaLaw). The three dialogical models of this paper are discussed to provide a setting to elaborate on structure in dialogical models. The degree of structure is related to the often mentioned but yet never worked out notion of naturalness of the dialogs as they are represented in models of argumentation. The reason it is important to address the naturalness of dialogical models is because such models can and have to be introduced into legal practice. For this introduction to be successful it is necessary that the dialogs are natural, fitting in with the ideas and working method of lawyers.