Bohm and Aharonov's 1957 calculation ruled out Furry's classical-mixture hypothesis, not every classical account of the Wu-Shaknov correlations

physicsquantum-foundationseprbell-theoremhistory-of-sciencebohm-aharonovwu-shaknovfurry

David Bohm and Yakir Aharonov's 1957 Physical Review paper (108:1070) reanalyzed Chien-Shiung Wu and Irving Shaknov's 1950 photon-polarization-correlation experiment, originally run as a test of a pair-theory prediction with no stated connection to EPR. The paper's own words — quoted directly from the primary text by physicist Peter W. Milonni in a 2025 Physics Today letter, cited to page 1075 — state that the Wu-Shaknov result "is explained adequately by the current quantum theory which implies distant correlations, of the type leading to the paradox of [Einstein-Rosen-Podolsky], but not by any reasonable hypotheses implying a breakdown of the quantum theory that could avoid the paradox of ERP." Independently, a 2025 peer-reviewed historical-physics paper by Yu Shi states that Bohm and Aharonov "demonstrated that only the entangled state produce the correct theoretical value very close to the experimental result of Wu and Shaknov, while the products or their classical probabilistic mixture, as proposed by Wendell Furry, lead to very different calculation results."

Furry had proposed in 1936 that once two entangled particles separate, their joint state effectively reduces to a classical probabilistic mixture of individually determinate product states — the specific, named classical account the 1957 calculation shows gives a different coincidence rate than the one Wu and Shaknov actually measured. Both witnesses quoted above are reporting on the same single primary document, which this vault has not yet read directly; this is one root (the 1957 paper itself) corroborated by two independent readers of it, not two independent derivations.

This narrows the broader priority framing in claim-bohm-aharonov-1957-epr-test-predates-bell-1964: the 1957 result rules out one specific, named classical rival hypothesis, not every conceivable classical account of the correlations.