Volume 27 · Part Fifteen · Contemporary Borrowings · Chapter 48 of 53

The Continuous Fabric: Bell's Theorem, Teleportation, and What Non-Locality Does Not Buy

The most thoroughly confirmed surprise in physics, separated line by line from the popular reading it usually arrives wrapped in.

What Bell actually proved

Bell's 1964 result is not a claim about quantum mechanics being strange. It is an inequality about any theory in which each particle carries its outcome-determining properties with it and no influence travels faster than light. Such a theory constrains the correlations between distant measurements. Quantum mechanics predicts correlations that exceed the constraint, and the CHSH form of the bound makes the excess a single measurable number: local theories cap the combination at 2, quantum mechanics reaches 2 root 2.

The experiments closed the escape routes one at a time. Freedman and Clauser saw the violation in 1972, Aspect sharpened it in 1982, Weihs added fast random switching in 1998, and in 2015 three groups — Hensen with entangled electron spins in separated diamonds, and the Shalm and Giustina photon experiments — closed the locality and detection loopholes simultaneously. The freedom-of-choice loophole was pushed to cosmological distance in the 2018 Cosmic Bell tests, which drew the measurement settings from photons emitted by high-redshift quasars. The 2022 Nobel Prize to Clauser, Aspect and Zeilinger was the discipline recording that the question is closed.

So the conclusion is exact and narrow: no local hidden-variable theory reproduces the observed statistics. At least one of locality or definite pre-existing values has to go. Which one you drop is an interpretive choice, and physics has not made it for us — Bohmian mechanics keeps the values and abandons locality, Everettian readings keep locality and abandon single outcomes, and the plain textbook practice declines to answer.

The half the postcards leave out

The no-communication theorem is the constraint that makes entanglement safe. Take any entangled pair. Whatever measurement one party performs, and whatever result they get, the other party's local statistics are exactly what they would have been with no measurement at all. This is not an engineering limitation awaiting a clever protocol; it follows from the trace structure of the formalism, and it is why entanglement coexists with relativity without contradiction.

That kills the phrase that does most of the damage in popular write-ups: a change here does not register there. The correlation is only visible after the two records are brought together and compared, and bringing them together is an ordinary classical transmission at or below light speed. What is non-local is the explanation of the correlation. What is strictly local is every signal, every influence you could use, every fact either party can observe on their own.

The distinction is worth stating in the form that survives contact with an engineer: entanglement is a correlation resource, not a channel. It can strengthen what a channel does. It cannot replace one.

Teleportation, priced honestly

The Bennett protocol of 1993, first demonstrated by the Zeilinger and Rome groups in 1997, is the clearest illustration of the accounting. To move an unknown qubit state from Alice to Bob you need a shared entangled pair established in advance, a joint Bell-basis measurement on Alice's side, and two classical bits sent to Bob so he knows which of four corrections to apply. Only then does his qubit hold the state.

Every clause is a cost. The original state is destroyed by Alice's measurement, which is the no-cloning theorem being enforced rather than evaded. The entangled pair is consumed — one pair per qubit teleported. The two classical bits travel at the speed of light, so the whole procedure is subluminal end to end. Nothing material moves, nothing arrives early, and the name is the worst in physics: it transfers a state, having first shared correlation and then paid classical postage.

The engineering has become serious anyway. Teleportation has been run over 143 kilometres between Canary Islands, from the Micius satellite to ground at 1400 kilometres, across metropolitan fibre in Calgary and Hefei, and between separated quantum-memory nodes in the Delft three-node network. Entanglement swapping links segments, and it is the mechanism behind quantum repeaters and the device-independent cryptography whose security rests directly on a measured Bell violation. This is the practical answer to whether non-locality is real: people are building infrastructure on the exact size of it.

The one-fabric reading, in its own column

There is something genuine underneath the postcard, and it is a structural fact rather than a mystical one. In quantum mechanics the state of a composite system is not in general a product of states of its parts, so 'two systems, each with its own properties' is a special case rather than the default. Entanglement entropy is a real measured quantity; area-law scaling in ground states and the entanglement structure of quantum field theory make separability look like a coarse approximation that happens to hold for everything we normally handle.

That is the defensible version: separability is emergent, not fundamental, and it emerges through decoherence — interaction with an environment that redistributes correlations so quickly that macroscopic objects behave as if they had independent properties. The classical world of separate things is the limit, not the ground floor.

What does not follow: that everything is entangled with everything, that observation tugs a cosmic web, or that distance is an illusion. Entanglement is fragile, specific, and monogamous in a precise technical sense — maximal entanglement with one system forbids it with any other. It has to be prepared, and it degrades. A universe whose deep description is non-separable is still a universe in which two rocks on a table share essentially no entanglement, and in which the metric structure of spacetime survives every Bell test ever run. The holographic proposals that make geometry itself emerge from entanglement — Ryu-Takayanagi, ER equals EPR, the it-from-qubit programme — are the serious form of the intuition, and they are conjectures inside a duality that has no experimental confirmation and no established connection to the world we live in.

Equations borrowed

Bell's inequality (1964) and the CHSH form: local realism bounded at 2, quantum bound 2 root 2 (Tsirelson)
Loophole-free Bell tests: Hensen et al. (2015), Shalm et al. (2015), Giustina et al. (2015); Cosmic Bell tests (2018)
No-communication theorem: local marginal statistics are unchanged by a distant measurement
No-cloning theorem (Wootters-Zurek, Dieks, 1982)
Teleportation protocol (Bennett et al., 1993) and entanglement swapping; Micius satellite and Delft multinode results
Entanglement entropy, area laws, and environment-induced decoherence as the route to apparent separability
Ryu-Takayanagi and ER=EPR as conjectural entanglement-geometry correspondences within AdS/CFT

Validity band

Bell's theorem and the no-communication theorem are proved statements within the standard formalism, and the experimental violations are among the most robust results in physics. The teleportation protocol and its resource costs are established and demonstrated. Decoherence accounts for apparent separability with quantitative timescales. The claim that separability is emergent rather than fundamental is a reading of the formalism that most physicists would accept as a description of the mathematics. The entanglement-geometry programme is conjecture inside a duality with no experimental test, and nothing in this chapter treats it as evidence about our universe. The popular 'single cosmic web' formulation has no technical content in the form usually stated and is not adopted here.

Falsifier

The chapter fails at once if any experiment demonstrates superluminal signalling using entanglement, or shows a distant measurement altering local marginal statistics — either would break the no-communication theorem and with it this chapter's central distinction. It also fails if a loophole-free Bell test is shown to have a systematic error large enough to restore a local hidden-variable account, or if teleportation is demonstrated without classical communication. The emergent-separability reading would be undercut by a demonstration that entanglement entropy in field theory is an artefact of regularisation with no operational meaning.

Where this chapter is weakest

The chapter has to fight on two fronts and may lose ground on both. Against the popular reading it insists on no-signalling, which risks sounding deflationary about a result that genuinely is astonishing — Bell violations really do rule out a picture of the world that almost everyone finds natural, and understating that is its own distortion. Against the technical reading it stops short of an interpretation, which means it leaves the reader with a proof that something must give and no account of what. The 'separability is emergent' formulation is the chapter's most quotable line and its least secured: it is a statement about the mathematics whose ontological weight depends entirely on which interpretation the reader was carrying when they arrived.

The volume-wide audit of these weak points is collected in Where This Volume Is Weak.