Communication Without Mindreading

I'm presenting joint work with Raquel Fernandez and joint work with Nicholas Asher and Antoine Venant on the same topic: how can two agents successfully communicate through only displaying and observing certain signals (i.e. without reading each others minds). I will first argue that this is even a problem. Then I will present a model that allows one to defeasibly compute what has been agreed upon in a conversation. Finally, I will outline how one can, without mindreading, derive the agents intentions from what has been agreed upon (again, defeasibly).