Modeling conversational exchange using games

Conversations often involve an element of planning and calculation of how best one can achieve one's interests. In a setting where the interests of the dialogue agents do not coincide with those of their interlocutors, strategic reasoning takes centrestage - it is not only important `what' one says but `when' and `how' one says it. Examples of such conversations abound in day-to-day life be it political debates, courtroom exchanges, negotiations, bargaining, marital disputes and so on. To predict and justify outcomes of such exchanges, it is natural to model them as games. Such a model should take into account both the strategic and the linguistic aspects of conversations. All previous attempts, including signaling games and its extensions, have totally ignored the linguistic considerations of conversations, among other drawbacks.
In this interdisciplinary course, we describe some recent work on the modeling of strategic conversations using infinite games, that uses elements from linguistics, game theory, topology and computer science. We show how such a model addresses the drawbacks of the earlier approaches and can be used to carry out a more comprehensive analysis of such conversations.