We will argue that social competence is an emergent mental phenomenon, and as such, there is no requirement to build discrete "social" modules into an agent. In fact, we argue that there are definite advantages to be gained from the emergent approach to social competence in complex, open, multi-agent environments. In order to capitalise on these advantages we need to design socially competent agents with the ability to reason on different levels (reactive, deliberative, meta) within complex social situations. By analysing the sociological theory of Pierre Bourdieu, we describe the design of a socially competent agent through the instantiation of a generic layered agent architecture. Our instantiation provides a methodology for specifying heuristics and parameters for different layers of such architectures. Furthermore, Bourdieu's habitus-field theory is hybrid in the sense that it tries to explain the effect of individual behaviour on societal structures and vice versa. This is where the great strength of the theory lies, and where we expect a useful cross-fertilisation of ideas into AI to occur. For as much as space permits, we will illustrate our argument with a scenario from the domain of shipping companies. This scenario is defined by its openness, diversity of agents as well as tasks and time restrictions. Our work leads us to the conclusion that building social agent architectures has definite engineering advantages, underlining the importance of this concept for both MAS and DAI research.
展开▼
