Alain Aspect · Testing entanglement
Can an experiment decide whether Einstein was right?
That time, that place
Bell's inequality was a blueprint, but a hard one to build. You need many entangled photon pairs and you must count them without losing too many.
Clauser and Freedman made the first attempt in 1972. The result favoured quantum mechanics. But loopholes remained.
Why this question
The awkward one ran like this.
If the detector orientations are fixed before the run, you cannot rule out that the source somehow 'knew' the settings and emitted photons to suit. Then a violated inequality still leaves hidden variables alive.
Contrived, but logically a hole that had to be plugged.
Aspect's group decided to change the settings after the photons had left. Mid-flight, a switch operating on nanosecond timescales redirected each photon towards a different analyser. The source cannot know in advance — to know, a signal would have to outrun light.
What was found
The inequality was violated, decisively. The gap was far beyond statistical doubt.
The world does not satisfy both of Einstein's conditions at once — values fixed in advance, and no influence between distant things. At least one has to go.
The remaining loopholes closed in turn. In 1998 Zeilinger's group pushed the detectors much further apart; in 2015 several laboratories closed the main loopholes simultaneously. The 2022 Nobel Prize in Physics went to Clauser, Aspect and Zeilinger.
And entanglement stopped being a controversy and became a material. Quantum cryptography and quantum computing run on this very property — the one Einstein invoked as absurd, now the operating principle of equipment.
Nature is local, and values are fixed before measurement
The inequality violated even when the settings were switched after the photons were already in flight
Entanglement became a material for technology, and led to the 2022 Nobel Prize