Collective Agents and Collective Intentionality Using the EDA Model

The notion of collective intentionally is controversial. This has been object of research by philosophers (Tuomela and Miller, 1988; Searle, 1979; Tuomela, 1994), AI scientists (Hobbs, 1990) and communication scientists (Conte and Castelfranchi, 1995). Related notions, such as joint intentions, teamwork, and mutual beliefs underlie much work both in formal models and computer models of multi-agent systems research in DAI (Cohen and Levesque, 1990; OpsilaHare and Jennings, 1996). In this paper we will show that the EDA model (Epistemic-Deontic-Axiological) has the required expressiveness to represent normative multi-agent coordination in organisations, using notions such as commitment, responsibility, organisational role, obligation, authorisation and other notions related to goal-governed normative systems.