The goal of non-cooperative foundations is to provide credible non-cooperat
ive models of negotiation and coalition formation whose equilibrium outcome
s agree with a given cooperative solution. Here we argue that this goal is
best achieved by explicitly modeling the physical environment and individua
l preferences, and by constructing game forms independent of preferences to
implement the cooperative solution. We propose a general model of the phys
ical environment; we characterize the coalitional functions arising from en
vironments via alpha- and beta-effectively; and we take up an issue of "pay
off consistency" with special reference to the core. (C) 1999 Academic Pres
s.