One of the most perplexing questions of game theory is how to analyze potential deviations from an equilibrium by a coalition of players. This project will take a "neologistic" approach to the problem, that is, to model negotiating statements explicitly, and to study their credibility. Based on ideas of others, the approach is so powerful that even in two player games, it forces one in some cases to retreat from equilibrium analysis to a rationalizable mode. But in such cases, additional "natural language" assumptions must be formulated, or else neologisms are always impotent (unless one player is artificially silenced). The potential applications to this research are unlimited: theoretical progress on coalition formation is awaited by researchers in many applied fields, including industrial organization, law and economics, finance, and international trade. The project also promises important insight in the understanding of the roles of communication, signalling and renegotiation in social settings.
