I have to look at pragmatics to define possible coding schema for extracting co-referenced, co-intentionality, and perspective taking index. All this parameters reflect on how people feel the process of the communication, their meta-representation of the ongoing communicative process. Three problems need to be faced:

1- how to assess in real-time the representation of the problem of the person without influencing his/her representation of the whole process (Heisemberg Indetermination’s Principle).
2- how to relate a single representation to the global representation of the ongoing communication ( A -> B = A <- B ?). 3- How people can work spatially, communicate and commit to other's statements? Can we invent a task that make this more operational?