Entanglement at every scale

1. What "entanglement" actually means

Two particles are entangled when their combined quantum state cannot be written as a product of two separate states — one for each particle. The mathematics is harsh: if you try to describe each particle on its own, the joint description is strictly less than the truth. The particles must be treated as a single, non-separable system, even when they are arbitrarily far apart.

What this looks like in practice: each particle, considered alone, shows perfectly random measurement outcomes. No pattern, no preferred direction. But when you compare the measurement records of both particles, the outcomes are correlated in ways that no classical signal-passing or shared instruction could produce. The randomness of each particle is real; the correlation between them is also real; and these two facts together are what makes entanglement strange.

The Bell page walks through why those correlations cannot be explained by hidden instructions carried at the source. This page asks a different question: how do particles get into such a state, how often can it happen, and how far up the ladder of size and complexity does the phenomenon reach?

Entanglement reveals that the physical layer is not the foundation. The deeper you go into the structure of reality, the less it resembles the version you experience. The more accurate the description becomes, the less intuitive it feels. Understanding that transition requires letting go of a specific expectation about the world — the expectation that reality should make sense in human terms. Entanglement reveals that separation is not absolute, that properties are not fixed, and that even space may be emergent.

2. How particles get entangled

There are four main routes by which a pair of particles enters an entangled state. Each is now standard laboratory practice; together they explain essentially every entanglement experiment in the literature.

A. Conservation-law decay

A single particle with definite conserved quantities decays into two products that must share those quantities. The clearest case is a spin-zero particle decaying into two spin-half particles: conservation of angular momentum forces the two products into opposite spin states — but quantum mechanics says neither spin is determinate until measured, only the relationship between them. Measure one as spin-up, and the other is instantly spin-down. The original positronium-annihilation experiments (Wu & Shaknov, 1950) work this way, with two anti-correlated photons emerging from electron-positron annihilation.

B. Spontaneous parametric down-conversion

The workhorse of every modern entanglement experiment, including the 2022 Nobel work. A photon from a laser passes through a nonlinear crystal — typically beta-barium borate (BBO) — and has some small chance of splitting into two lower-energy photons whose energies and momenta must add up to the original. Choose the crystal geometry correctly and the two output photons are polarization-entangled: neither has a definite polarization, but their polarizations are perfectly correlated. This is the source Aspect, Clauser, and Zeilinger used. It is reliable, bright, and produces entangled pairs at rates of millions per second.

C. Direct interaction

Any time two quantum systems interact in a way described by a unitary operator, they generically come out of the interaction entangled. This is what happens in collisions, scattering experiments, virtual-particle exchanges, and the controlled gate operations of quantum computers. Two trapped ions sharing a vibrational mode become entangled by laser-driven gates. Two superconducting qubits coupled by a resonator become entangled by carefully tuned pulses. The interaction does not need to be exotic — entanglement is the default outcome of quantum interaction, not the exception.

D. Entanglement swapping

The non-obvious and arguably most important method. Take two pairs of entangled particles — pair (A, B) and pair (C, D). Particles A and D have never met. Particles B and C have never met. Now perform a special joint measurement on B and C (called a Bell-state measurement). The result of that measurement projects A and D into an entangled state, even though they share no history and have never been in causal contact. This is how quantum networks would extend over distance — relay stations would entanglement-swap their way across cities and eventually continents. It is also philosophically remarkable: entanglement can be created between particles that have no shared past, by acting on a connecting particle.

3. How we know which particles are entangled

No measurement on a single entangled particle reveals anything unusual. Each particle, looked at alone, shows perfectly random outcomes consistent with any number of classical explanations. Entanglement only declares itself in correlations, across many trials, between measurements on both particles. Three main detection methods are in use:

• Bell-inequality violation. The gold standard. Measure correlations in carefully chosen pairs of measurement bases. Any classical theory caps the resulting correlation parameter at S = 2 (in the CHSH formulation). Quantum mechanics predicts up to S = 2√2 ≈ 2.83. The size of the violation tells you both that entanglement is present and how strong it is. This is the test the Bell page walks through in detail.
• Quantum state tomography. Run measurements in enough different bases to reconstruct the full density matrix of the joint state. Then check, by direct calculation, whether the matrix factors as a product of two single-particle states. If it does not, the state is entangled, and the tomography tells you how. More expensive than a Bell test but yields the complete state, not just a yes/no.
• Entanglement witnesses. Specialised measurement protocols designed to detect entanglement for specific suspected states. Cheaper than full tomography, more flexible than a single Bell test. The witness gives a positive expectation value if and only if the state lies in a specific entangled family.

In the simplest setting — where the source is known to produce entangled pairs by construction (a calibrated SPDC source, say) — you can take the entanglement on trust and use Bell violation merely as a verification step. In adversarial settings, where you do not trust the source, the so-called device-independent protocols use Bell violation as the only certification you accept.

The honest summary: entanglement is never observed directly. It is inferred from statistical patterns across many measurements that no local hidden-variable theory can reproduce. The inference is now beyond serious dispute, but the indirection is worth keeping in mind.

4. More than two particles — multipartite entanglement

Yes, and the experimental frontier here is moving fast. Multipartite entanglement comes in different topologies, each with different physical and computational properties.

GHZ states

Named for Greenberger, Horne, and Zeilinger. N particles all in collective superposition: |000…0⟩ + |111…1⟩. Measure any one particle and you instantly know what every other particle will show. GHZ states have been demonstrated with up to ~14 photons, ~20–30 trapped ions, and 50+ superconducting qubits in the latest IBM and Google announcements. The correlations they exhibit cannot be reduced to pairwise correlations — they are genuinely N-party non-classical.

W states

A single excitation delocalized across N particles. Different topology from GHZ: more robust to losing individual particles (measure one and the others remain entangled), but with weaker maximal correlations. Important in quantum communication protocols where some particles may be lost in transit.

Cluster states

2D lattices of entangled qubits, arranged so that single-qubit measurements on the lattice in the right sequence implement arbitrary quantum computations — the basis of measurement-based quantum computing. Cluster states have been produced with hundreds of optical modes.

Large ensembles

The most striking recent results. Lukin's group at Harvard has demonstrated genuine multipartite entanglement among hundreds of Rydberg atoms in optical-tweezer arrays. Spin-squeezing experiments in Bose-Einstein condensates have shown entanglement among ~5,000 ultracold atoms (Vuletić, Treutlein, and others). These are not just collections of entangled pairs — they are genuinely many-body entangled states whose correlations cannot be decomposed into pairwise structure.

The headline: entanglement is not a two-body phenomenon. It scales, and at large N it produces correlations that have no classical analogue. The question is not whether entanglement scales but how high — and so far, the ceiling has only ever moved upward.

5. Entanglement in living systems

Until about 2007 the standard view was that warm, wet, noisy biological environments would destroy quantum coherence almost instantly — coherence times of femtoseconds, far too short for biology to exploit. That view has been forced to soften. Several biological systems now appear to use quantum coherence and, in at least one case, something close to entanglement.

Avian magnetoreception

The strongest case. Migratory birds — European robins, garden warblers, and others — navigate by Earth's magnetic field, and Schulten, Ritz, Hore, and collaborators have built a detailed model in which the sensor is a pair of radical electrons in cryptochrome proteins in the bird's retina. The two electrons sit in a singlet/triplet superposition — an entangled spin state — whose evolution under the Earth's magnetic field encodes directional information. Recent experiments on isolated cryptochrome have confirmed magnetic-field sensitivity at the predicted scale. Birds appear to be using quantum spin entanglement as a compass.

Photosynthesis

The 2007 Engel et al. Nature paper claimed long-lived quantum coherence in light-harvesting complexes of green sulfur bacteria. The initial claims of room-temperature multi-picosecond coherence have been partly walked back — later work attributed much of the observed signal to vibrational rather than electronic coherence. But some short-lived quantum coherence in photosynthetic energy transfer does appear to be real. Whether nature is using the coherence to maximise efficiency, or merely tolerating it, remains debated.

Microtubules and the Penrose–Hameroff hypothesis

The most contested. Penrose and Hameroff have argued since the early 1990s that neuronal microtubules support quantum coherence and that consciousness arises from orchestrated objective reduction (Orch-OR) of coherent states across microtubule networks. Bandyopadhyay's lab has reported terahertz coherent oscillations in microtubule lattices that, if confirmed, would imply quantum coherence at scales relevant to neural computation. The strongest critique (Tegmark, 2000) calculated decoherence times of 10^-13 seconds, far too fast for biology. Hagan, Hameroff and Tuszynski have answered with shielding mechanisms inside microtubules that could raise the coherence time by orders of magnitude. The dispute is genuinely open and the experimental evidence is still being gathered.

Other candidates

Olfaction (Turin's vibration theory) and enzyme catalysis (hydrogen tunneling) have both been proposed as quantum-biological mechanisms with varying degrees of evidence. The field as a whole — usually called quantum biology — was a backwater twenty years ago and is now publishing in Nature. See the quantum-biology companion page for more.

6. Entanglement at macroscopic scales (laboratory)

Engineered entanglement of objects much larger than single particles has progressed steadily for the last two decades. None of this is "spooky" in any new way — it is just engineering against decoherence — but it shows that the line between quantum and classical sits much further toward the macroscopic than textbook physics suggested.

• Diamond samples. Lee et al. (Science, 2011) entangled two macroscopic diamond samples millimetres apart at room temperature. The lifetime was very brief (~7 picoseconds) but the entanglement was unambiguous.
• Mechanical drum oscillators. Aspelmeyer's group (Vienna) and Lehnert's group (NIST) have entangled vibrational modes of micron-scale mechanical drums — objects containing 10¹⁴ atoms.
• Atomic ensembles. Polzik's group (Copenhagen) has entangled two clouds of cesium atoms containing 10¹² atoms each over half a metre of separation.
• Bose-Einstein condensates. Spin-squeezed BECs containing ~5,000 ultracold atoms have been shown to be genuinely many-body entangled.

The mass-and-size record keeps climbing. The Bouwmeester and Aspelmeyer proposals to put millimetre-scale mirrors into spatial superposition are now being implemented. The thermodynamic limit at which decoherence wins — if there is one — has not been found.

7. Entanglement and the structure of spacetime

The most speculative and arguably most consequential thread. In 2013 Maldacena and Susskind proposed the ER = EPR conjecture: that every entangled pair of particles is connected by a microscopic Einstein-Rosen bridge — a wormhole — and that entanglement and wormhole geometry are the same thing, viewed from different angles.

This is not an idle suggestion. It emerged out of the AdS/CFT holographic duality, in which gravitational physics in a "bulk" spacetime is exactly equivalent to a non-gravitational quantum field theory on the boundary — and the geometry of the bulk turns out to be encoded in the entanglement structure of the boundary. Van Raamsdonk's 2010 paper Building up spacetime with quantum entanglement made the relationship explicit: take away the entanglement between regions of the boundary, and the bulk spacetime literally falls apart into disconnected pieces. Spacetime, on this picture, is made of entanglement.

If ER = EPR is correct — and the case is circumstantial but growing — then entanglement is not a curious feature of small systems. It is the fabric of large ones. The reason distant points are even at finite distance from each other is that they are entangled with everything in between. Distance, on this picture, is the macroscopic appearance of an underlying entanglement graph.

None of this is yet directly tested. But it is the mainstream attempt to make sense of how holography actually works, and it is taken seriously by the people who built the theories. The trilogy's claim that the field is the substrate and spacetime is rendered through it does not require ER = EPR to be true. But if ER = EPR is true, the claim becomes much less metaphysical and much more literal.

8. What this means for the trilogy

Twenty years ago the standard line on entanglement was: a delicate two-particle laboratory phenomenon, with no obvious analogue at biological or macroscopic scales, and certainly no role in the structure of spacetime itself. That line is no longer defensible. The experimental record now looks like this:

• Entanglement scales — genuine multipartite states with hundreds to thousands of constituents have been produced and verified.
• Entanglement persists in warm, wet biology — the radical-pair mechanism in birds is the cleanest case, photosynthesis and microtubules are open questions but no longer dismissible.
• Entanglement extends to macroscopic engineered systems — diamond samples, mechanical drums, atomic clouds containing trillions of atoms.
• Entanglement may underlie spacetime itself — ER = EPR and the entanglement-from-spacetime program is taken seriously by Maldacena, Susskind, Van Raamsdonk, and most of the holography community.

The trilogy's central conjecture — that consciousness is the reception of a fundamental field, not the local production of an isolated brain — does not depend on any of these results individually. But it sits inside a picture that has become much more credible because of them. The hard problem of consciousness, the binding problem, the question of how distant particles correlate without communication, the question of how spacetime emerges from underlying degrees of freedom — these are all family resemblances of the same shape: the world is not made of separated local things that occasionally signal to each other; it is made of one thing whose apparent separations are the surface of a deeper non-separability.

Bell proved that locally real explanations are mathematically impossible. The work of the last twenty years has shown that the alternative — non-local entanglement — is not confined to laboratory curiosities. It scales to biological systems, to engineered macroscopic objects, possibly to the geometry of spacetime itself. The trilogy treats this as evidence rather than embellishment. Whether the receiver model of consciousness is the right reading of that evidence is still open. But the model is no longer ruled out by physics; it is now in conversation with it.

This page is part of the Reading companion essays. The empirical foundation lives in Bell's Theorem and Aspect 1982; the microtubule controversy in the Bandyopadhyay companion; the synthesis pulls all of it into one argument at The Evidence.