Abstract. There are two main traditions in defining a semantics for agent communication languages, based either on mental attitudes or on social commitments. In this paper, we translate both traditions in a different approach in which the dialogue state is represented by the beliefs and goals publicly attributed to the roles played by the dialogue participants. On the one hand, this approach avoids the problems of mentalistic semantics, such as the unverifiability of private mental states. On the other hand, it allows use to reuse the logics and implementations developed for FIPA compliant approaches.