Socially Friendly and Group Protecting Coalition Logics

We consider extensions of Coalition Logic (CL) which can express statements about inter-related powers of coalitions to achieve their respective goals. In particular, we introduce and study two new extensions of CL. One of them is the “Socially Friendly Coalition Logic” SFCL, which is also a multi-agent extension of the recently introduced “Instantial Neighborhood Logic” INL. SFCL can express the claim that a coalition has a collective strategy to guarantee achieving its explicitly stated goal while acting in a ‘socially friendly way’, by enabling the remaining agents to achieve other (again, explicitly stated) goals of their choice. The other new extension is the “Group Protecting Coalition Logic” GPCL which enables reasoning about entire coalitional goal assignments, in which every group of agents has its own specified goal. GPCL can express claims to the effect that there is an action profile of the grand coalition such that, by playing it, every sub-coalition of agents can guarantee satisfaction of its own private goal (and thus, protect its own interests) while acting towards achievement of the common goal of the grand coalition. For each of these logics, we discuss its expressiveness, introduce the respective notion of bisimulation and prove bisimulation invariance and Hennessy-Milner property. We then also present sound and complete axiomatic systems and prove decidability for both logics