Degrees of Freedom in the Social World: Towards a Systems Analysis of Decision*

THOMAS SCHELLING taught us that in ordinary human affairs, con ̄ict and common interest are ubiquitously intertwined.1 For when it comes to variety, the occasion of pure con ̄ict (known to some of its friends as the zerosum game) is as under-represented in human affairs as the occasion of undiluted common interest (known as the pure coordination game). The undiluted extremes are the exceptions, when it comes to counting kinds, while the mixed-motive kind of occasion is the rule. Things look a bit different, however, when one looks at sheer numbers of true-life occasions, as I will explain. Schelling also taught us that in the diverse space of mixed-motive affairs, intermediate between pure con ̄ict and pure coordination, there is more true-life collaborative behaviorÐmore trustings and promise-keepingsÐthan our prescriptive decision theories can accommodate, never mind explain. This collaborative behavior, as long ago another ThomasÐThomas HobbesÐwas painfully aware, requires some explaining precisely because of the presence of ineliminable con ̄ict of interest. The problem confronting those who aspire to explain such collaboration is to identify weighty motivations in its favor, in the face of the weighty, but countervailing motivations. What shall concern me here is not the collaboration that transpires in the face of con ̄ict of interest, but the success which meets us more than halfway in those limiting cases of pure coordination, so as (for example) not to collide in roadways and corridors. For in sheer quantity, the occasions for pure coordination outnumber the occasions for anything else perhaps a hundredfold. We The Journal of Political Philosophy: Volume 7, Number 4, 1999, pp. 453±477