Monotonic vs non-monotonic changes:
As for now, only monotonic changes are implemented: G(t1) < G(t2) if t1<t2 in the case of growing a SemRep. (but note that nodes could be removed due to them being pruned!)
Add edges with stump nodes (unbound variables?) (or without any node at all) so that one can set up expectations, etc
Update nodes! (non-monotonic)
In one option, graph topology remains the same and the node concept is updated.,
In another option, graph topology does not remain the same (for example if I an "update" relation is introduce). In which case the update might at first be expressed through a monotonic change which might, later on, be merged into the main graph t1: A => t2: A<--update-- B => t3: B
Main question: