In a “prepare, then measure” experiment, it is assumed that the ontic state screens off the preparation procedure from the measurement outcome. The Pusey–Barrett–Rudolph (PBR) theorem shows that, in any HS ontological model satisfying a minimal assumption, the quantum state of a system is uniquely determined by its ontic state, and is therefore an objective fact about the system’s ontology according to that model.

This work argues that the HS ontological models framework makes metaphysical commitments that do not follow from realism alone, and that the conclusion of the PBR theorem depends on these additional commitments. A recent interpretation of quantum theory, Causal Relational Quantum Theory (CRQT), provides a concrete counterexample. CRQT is based on a dynamics-first ontology. Ontic states do emerge within this ontology under certain conditions, but these conditions cannot be met in the PBR scenario. In particular, the full ontology of the systems considered in the PBR experiment does not screen off preparations from measurement outcomes. It is argued that, in CRQT, this failure of screening-off does not imply any form of temporal nonlocality, although it would in a state-first ontology. The quantum state in CRQT is not a fact about the ontology of the system, but nor is it “epistemic” in the sense of the Harrigan and Spekkens framework. From the perspective of CRQT, the moral of the PBR theorem is not the reality of the quantum state, but rather that fundamental reality is dynamical, and that ontic states have an emergent status.