The grounding problem of modeling emotions in adaptive artifacts

A communality between research in artificial intelligence and synthetic emo tion is that is seems in both cases to be rather difficult to give an accep table definition of the naturally occurring counterpart. One could speculat e whether this is due to the multiplicity of the nature of both phenomena o r due to a categorical misconception. In this article, I try to briefly out line a number of different motivations for modeling emotions, and to relate those motivations to two different principal design approaches for computa tional models of emotion. From these two aspects, together with our current assumptions about mechanisms underlying human emotions, I conclude with so me speculations about adaptation in affective systems, and some implication s of the notion of grounding emotions in adaptive systems.