Bell Tests: Proving Entanglement Is Real
A clever inequality lets an experiment decide whether entangled particles secretly agreed in advance — and reality breaks the limit, ruling that out.
This whole module rests on entanglement being genuinely quantum — a real, non-classical link. But a skeptic has an obvious objection: maybe entangled particles aren’t mysterious at all. Maybe, when they were created together, they just agreed on their answers in advance — like two people carrying sealed envelopes with matching notes. That would explain the correlations with boring classical “hidden variables,” no spookiness required. For decades this was philosophy — untestable. Then came a way to decide it with an experiment. Hold the question: how could any measurement tell “agreed in advance” apart from “genuinely quantum”?
The skeptic’s escape: hidden variables
The innocent explanation is called local hidden variables. “Hidden variables” = the particles carry pre-set answers we just can’t see; “local” = each particle’s outcome depends only on those local pre-sets and its own measurement, not on what’s done to the faraway partner. If that were the whole story, entanglement would be no more mysterious than two envelopes mailed from the same room. The problem: if you only ever measure the same thing on both particles, pre-agreement and quantum entanglement predict identical results — you can’t tell them apart. The breakthrough was realizing you must measure them along different, varied settings and look at the pattern of correlations across those settings.
Bell’s inequality: a number the skeptic can’t exceed
John Bell’s insight: any local-hidden-variable theory — no matter how the pre-agreed answers are arranged — must obey a mathematical limit on how strongly the results can correlate across different measurement settings. Call it a ceiling: a Bell inequality puts a hard number that pre-agreement can never beat. It’s a bound derived from just two assumptions (locality + pre-set values), so it holds for every possible “envelope” scheme at once. Quantum mechanics, by contrast, predicts correlations that exceed that ceiling for the right settings. So the two pictures make different, numerical predictions — and now it’s an experiment, not a debate.
Why varied settings are the whole trick: • Measure both particles the same way every time → pre-agreement and quantum give identical perfect matches. No test. • Measure along several different angles, randomly chosen, and tally how often results agree for each combination → pre-agreement can’t push the overall correlation past Bell’s ceiling, but quantum entanglement can. The varied settings are what expose the difference.
The verdict: reality breaks the limit
Starting in the 1970s–80s and sealed by increasingly airtight experiments since, real measurements on entangled particles violate Bell’s inequality — they show correlations stronger than any local pre-agreement could ever produce. The skeptic’s comfortable envelope story is experimentally ruled out. Nature is not both local and pre-determined: entanglement is a genuine, irreducibly quantum correlation, exactly as this module assumed. (Note the careful wording from the teleportation lesson: this does not enable faster-than-light signaling — the correlations are real but can’t carry a message on their own.) Bell tests are why we can build quantum key distribution, trust teleportation, and treat entanglement as a real resource rather than a bookkeeping trick.
Imagine twins who always give the same answer to any question, and you suspect they simply memorized an answer sheet beforehand. If you only ever ask them the same question, you can’t catch them — a memorized sheet explains it. So instead you put them in separate rooms and ask each a randomly chosen question from a big list, over and over, and study how often their answers line up across all the combinations. It turns out there’s a hard limit on how well any pre-memorized answer sheet can make their responses correlate across varied questions. The twins beat that limit — consistently — so no answer sheet could have done it. Something deeper than pre-agreement is going on. That’s a Bell test.
How a Bell test settles the argument, step by step: 1. Make many entangled pairs; send one particle to Alice, one to Bob, in separate labs. 2. For each pair, each randomly picks one of a few measurement settings and records the outcome. 3. Compute the combined correlation across all setting-combinations and compare to Bell’s ceiling. 4. Result: the number lands above the ceiling — impossible for any local-hidden-variable (pre-agreed) scheme, but exactly what quantum mechanics predicts. Verdict: the correlations are genuinely quantum, not pre-arranged.
This is the reading. The interactive version — active-recall quiz, a hands-on experiment you run in your own AI, and an earned mastery check — is free in the app.
Start this lesson free →