Communication and the possibility of deceit: a game theoretical analysis (Amsterdam)

Where one can communicate, one can manipulate and deceive. Norms are used to limit deceit. The aim of this IP is to enhance a better understanding of deceptive versus reliable communication, and the evolution of norms that preserve linguistic conventions. The starting point is to view communication as a kind of a game between speakers and hearers. A hearer has the task to interpret the utterance in the way intended by the speaker, and a speaker has the task to give enough clues by her use of the utterance to enable the hearer to do so.  David Lewis explained in this way an conventional meaning of expressions in terms of stability of signaling games. Messages are not endowed with pre-existing meanings, but the meanings depend on the game-theoretical equilibria.

Lewis assumed that conventions perpetuate themselves because they serve a common interest: to solve a coordination problem. Gricean pragmatics, too, is deeply rooted in a Cooperative Principle: the idea that a (normal) conversation has a common purpose or direction, shared by all participants. A common language is a useful tool to coordinate one’s actions in case the preferences of the agents are aligned. But this common language opens the door for deception once preferences are not perfectly aligned. Indeed, agents misuse communication systems ranging from explicit lying to keeping secrets in a court of law. But this behavior threatens to undermine (the trust in, and use of) these same linguistic conventions and conversational norms. This gives rise to the main problem of this subproject: How it is possible that linguistic conventions and conversational norms remain stable under this pressure of non-cooperative behavior?

For further information, please contact Robert van Rooij (project leader): R.A.M.vanRooij [at] uva.nl