Derive a constrained law for encounter-order effects
Derive a constraint governing semantic-state transformations when two individuals encounter the same works in different orders, and determine whether the resulting order effects fall into a constrained family that distinguishes the quantum-probability account from signed classical models.
References
The constraint has not yet been derived; the theory is incomplete without it, which is why I am posing the problem. The constraint cannot be borrowed from the QQ equality. That equality covers two questions put to a reader who is already who she is. Encounter order is a different object: encounters transform the state itself, and no existing family of constraints covers the order in which such transformations are applied. The three accounts are still separate. Bayesian updating on interchangeable evidence predicts no order effect at all. Unconstrained classical models allow order effects of any shape. My account predicts order effects that fall into a constrained family yet to be identified.
I cannot yet say whether quantum probability does better. What would settle it is a result I have not proved: when two people meet the same works in different orders, my theory says they end up with different constellations, and it does not yet say how different. If the differences fall into a constrained family, the quantum version rules out possibilities the classical one allows. If they can take any shape, the two accounts are equivalent and mine is a notation.