Home > Positioning > Persons > Bell

John Bell (1928–1990)

Bell proved that quantum mechanics is incompatible with local realism. His theorem (1964) — and the experimentally testable inequalities it generates — showed that no theory in which physical properties exist independently of measurement (realism) and in which influences cannot travel faster than light (locality) can reproduce all the predictions of quantum mechanics. The result refuted Einstein’s preferred resolution of the EPR paradox — local hidden variables — without thereby settling the interpretation question that the Einstein-Bohr dispute had opened. Nature violates Bell’s inequality, as Alain Aspect’s experiments (1982) and subsequent tests have confirmed. The theorem does not dictate which element of local realism fails — locality, realism, or some assumption about the independence of measurement settings — and the choice among the options maps onto the interpretations of quantum mechanics that remain in contention.

Life

Born 28 June 1928 in Belfast, Northern Ireland, into a working-class family. His father was a horse dealer. Educated at the Belfast Technical High School, Queen’s University Belfast (BSc in experimental physics, 1948; BSc in mathematical physics, 1949), and the University of Birmingham (PhD, 1956), under Rudolf Peierls, on the CPT theorem in quantum field theory.

Worked at the Atomic Energy Research Establishment, Harwell (1949–60), on accelerator physics and nuclear theory. Joined CERN in Geneva (1960), where he spent the rest of his career as a theoretical physicist working primarily on accelerator design and particle physics. The foundational work for which he is remembered — Bell’s theorem, the Bell inequalities, and his papers on the measurement problem — was done alongside his main employment in particle physics, often described by Bell as a “hobby.” The 1964 paper was written during a sabbatical at SLAC (Stanford Linear Accelerator Center).

Bell was sharply critical of the Copenhagen interpretation of quantum mechanics and sympathetic to Bohm’s hidden-variable approach — not because he thought Bohm’s theory was necessarily correct but because its existence refuted the claim (widespread since von Neumann’s 1932 proof, which Bell showed was flawed) that hidden-variable theories are impossible. Died 1 October 1990 in Geneva, of a cerebral haemorrhage, at the age of sixty-two.

Bell’s theorem

The EPR argument (Einstein, Podolsky, Rosen, 1935) had argued that quantum mechanics is incomplete: if two entangled particles are separated and measured, quantum mechanics predicts perfect correlations between certain measurements, but provides no local mechanism to explain them. Either the particles carry predetermined values (hidden variables) that determine the outcomes, or the measurement on one particle instantaneously affects the other (non-locality). Einstein preferred the first option; Bohr rejected the framework of the argument but did not provide a precise counter-model.

Bell’s contribution was to make the dispute empirically testable. He derived an inequality — a mathematical constraint on the correlations between measurements on entangled particles — that must be satisfied by any local hidden-variable theory. Quantum mechanics predicts correlations that violate the inequality. The violation is not marginal; it is substantial and unambiguous.

The inequality. In the simplest form (the CHSH inequality, a generalisation by Clauser, Horne, Shimony, and Holt, 1969): the correlation between measurements on entangled particles, summed over four combinations of measurement settings, must be bounded by 2 in any local realistic theory. Quantum mechanics predicts a maximum of 2√2 ≈ 2.83. Experiments consistently observe values near the quantum-mechanical prediction, violating the local-realist bound.

What the theorem rules out. Bell’s theorem rules out the conjunction of locality and realism (and the assumption that measurement settings can be freely chosen independently of the hidden variables — “measurement independence” or “free will”). It does not dictate which assumption to abandon:

The theorem is silent on which option is correct. Its power is negative: it eliminates an entire class of theories (local hidden variables) and forces the remaining options into the open.

The measurement problem

Bell was equally concerned with the measurement problem — the question of how and when quantum superpositions (a particle in a superposition of being here and there) become definite outcomes (the particle is found here). The Copenhagen interpretation invokes “collapse” — the wave function collapses upon measurement — but does not specify what constitutes a measurement or when collapse occurs. Bell regarded this as a fundamental incoherence, not a minor gap.

His 1990 paper “Against ‘Measurement’” — one of his last — argued that the word “measurement” has no precise meaning in quantum mechanics and should not appear in the fundamental formulation of the theory. A fundamental theory should specify what happens in all circumstances, not only when a “measurement” is performed. The paper advocated for theories that avoid the measurement problem entirely: Bohm’s mechanics (which has definite particle positions at all times), spontaneous collapse theories (GRW), or other approaches that replace the vague notion of measurement with precise physical processes.

Where Bell stops

Bell’s theorem is a no-go result — it tells you what cannot be the case (local realism) but does not tell you what is the case. The theorem is compatible with every major interpretation of quantum mechanics: Copenhagen, Bohm, many-worlds, relational QM, GRW. The choice among these interpretations is not resolved by Bell’s work; it depends on further philosophical and physical considerations (the role of locality, the reality of the wave function, the acceptability of non-locality or branching). Bell personally favoured Bohm’s approach and was sympathetic to GRW, but he recognised that his theorem did not settle the interpretation question.

The experimental tests of Bell’s inequality have been progressively refined — closing the “locality loophole” (Aspect, 1982), the “detection loophole” (Giustina et al., Hensen et al., 2015), and finally both simultaneously in “loophole-free” tests. The superdeterminism loophole — the possibility that the choice of measurement settings is correlated with the hidden variables — remains in principle unclosable by experiment, since any experimental test must assume that the settings are freely chosen. Whether superdeterminism is a live option or a philosophical dead end is debated.

Key works

See also: Einstein · Bohm · Bohr · Born · Relational quantum mechanics