Testing finite-speed hidden influences with a four-qubit cluster-state witness: an experimental proposal
Quantum correlations between distant measurements might, in principle, be produced by a hidden influence that travels at a finite speed V > c in some preferred frame. Two-party Bell experiments can only push the lower bound on V upward. Bancal et al. (2012) showed that any such model that forbids signalling must satisfy a four-party Bell-type inequality built only from three-party marginals, and Li, Hu, Deng and Scarani (2026) found a robust four-qubit linear-cluster-state witness of this kind, S4 <= 6, with quantum value 6.83 and white-noise visibility threshold 0.879. No experiment has tested it. This proposal turns that witness into an experimental plan for a bound on finite-speed hidden influences. Stage 1 is a single-site photonic test with a four-photon cluster source, a per-slot Sagnac-Pockels switch and a pre-registered test-supermartingale analysis that remains valid under memory and optional stopping and includes a fixed allowance for random-number predictability. With a published source of fidelity 0.95, about 1.6e4 four-fold events (roughly 10-11 hours) give 90% power at p < 2.87e-7. A pass would exclude the class of no-signalling hidden-influence models considered, under stated assumptions that include strengthened fair sampling; it is not loophole-free. Later spacelike stages need networked matter qubits that do not yet exist. The record contains the proposal, a short feasibility note for experimental groups, and all calculation scripts. AI software tools assisted with the calculations and drafting; the author reviewed all content.
Authors
- Richard Joe Phillips (ORCID: https://orcid.org/0009-0008-0497-7712)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24426204
- Primary Topic
- Quantum Mechanics and Applications
- Type
- preprint