The present study aims to develop a method for the analysis of argumentative patterns in spontaneous dialogue (Chapters 1-3), and to apply this method to samples of face-to-face interaction (Chapters 4-7). The purpose of the applied study is twofold: to evaluate the proposed method and to provide a contrastive description of the communicative style of Swedish and Spanish speakers in intra- and intercultural same-gender and mixed-gender dialogues. The study emphasises the co-operative and dynamic features inherent in argumentation in interaction as well as the dialogical nature of all conversational contributions. Argumentation is defined as a dynamic process whereby opinions are negotiated in interaction. The argumentative process is analysed into separable argumentative moves that are performed explicitly or implicitly, and whose domain is the idea unit (sense unit). The primary aim of argumentation is identified as that of making one standpoint, as reflected in an overtly or covertly expressed opinion, prevail over other possible or expressed standpoints/opinions. This general objective encompasses the specific interactive aims of creating adherence in the audience to a standpoint, and that of making the audience prepared to act. The model proposed comprises a set of argumentative and non-argumentative moves, defined by four sets of binary distinctive features: [+/– new topic], [+/– agreement], [+/– new information], and [+/– standpoint]. In addition, the model discriminates between self-directed and other-directed moves. The empirical study focuses on six of the argumentative moves (acceptance, rejection, other-support, self-support, other-objection, and self-objection). Linguistic strategies used for implementing these moves are identified. The strategies and the frequencies with which the various moves occur are correlated with two prototypical communicative styles: ‘collaborative’ vs. ‘self-assertive’. On the basis of the analysis, argumentative profiles are traced for each of the four participant categories in each of the four contexts analysed. The analysis, apart from revealing systematic differences between the groups, lends empirical support to the proposed method.