WorksPhysics and Astronomy
An Incomplete-Preference Quantum Anscombe–Aumann Representation
Abstract
This work develops a generalised Anscombe–Aumann framework in which acts map a finite set of states of the world into the effect interval of a finite-dimensional Hilbert space, while preferences need not be complete. The construction keeps three logically distinct questions separate. First, an incomplete-preference multi-utility theorem supplies affine utility functionals. Second, finite-dimensional operator geometry determines the form of each affine functional. Third, admissibility identifies which normalised probability–state pairs are compatible with the full preference. We introduce behavioural state-independence as the natural Anscombe–Aumann interpretation of the idea that the identity of the states of the world should not affect how quantum consequences are compared. Because an incomplete preference makes this aggregate behavioural condition weaker than a utility-by-utility condition, the factorisation theorem requires a stronger, explicitly operational representation-level condition. Under that condition and Loewner monotonicity, every affine utility functional is positively affinely equivalent to a normalised utility of the form . Consequently, a suitable incomplete preference relation is represented by a family of admissible pairs in . We then identify a canonical maximal compatible family and show that the scalar-effect subspace carries an ordinary classical multi-utility representation, with scalar events appearing as the vertices of the scalar-act cube.