Creating and testing states of knowledge in groups.

Tentative abstract:  
 
A popular way of describing states of knowledge in groups is via Kripke structures.  How do we create a particular state of knowledge described by a finite Kripke structure in a (finite) group of  agents?  And given such a group of agents with beliefs about some propositions, how can we test their states of knowledge via observing their behavior?   
 
Note that if the agents are babies, toddlers or animals, or people likely to lie, observing behavior is the best way we can find out their beliefs. 
 
We use Shakespeare's Much Ado about Nothing and the Wimmer and Perner work on Maxi as illustrations.