Spin Networks — How Space May Be Made of Quantum Atoms

Trillions of times every second, you might be blinking out of existence and blinking back. Your body. Your thoughts. The unbroken you reading this sentence. That is the claim, at least, and it circulates widely in popular accounts of one of the most serious attempts physics has ever made to understand what space is actually made of.

The claim is wrong. It is wrong by a factor of about ten thousand billion billion billion. And, more importantly, it is wrong in a way the theory itself never proposed.

The theory is loop quantum gravity. Its central objects are mathematical structures called spin networks. And its core claim, stated carefully, is that space at the smallest scale imaginable may not be a smooth stage that goes down forever but may instead be built from discrete grains of geometry. Not a backdrop. A fabric.

This is what the theory actually says, what the equations actually require, and why the popular cartoon and the careful physics are not the same thing.

The problem nobody could shake

For nearly a century, physics has carried two magnificent theories that refuse to get along. General relativity, published by Albert Einstein between 1905 and 1915, describes gravity as the curvature of a smooth four-dimensional spacetime, a continuous fabric bent by mass and energy. Quantum mechanics, developed across the 1920s and 1930s, describes matter and the other forces as discrete, grainy, probabilistic — built from indivisible quanta and irreducible uncertainty.

Each theory is among the most precisely tested in the history of science. The GPS in your phone only works because engineers correct, every second, for the time-dilation effects of general relativity. The transistors that run that GPS were designed using quantum mechanics. Both theories are right where we can test them.

But when you try to apply quantum rules to gravity itself, when you ask what spacetime is doing at the smallest scales, the mathematics breaks down. The smooth fabric and the grainy quantum refuse to be the same thing. The two frameworks are mutually incompatible at the foundations, and reconciling them is the central unsolved problem of theoretical physics.

Loop quantum gravity is one of several serious attempts. The Italian physicist Carlo Rovelli, the American physicist Lee Smolin, and the Indian-American physicist Abhay Ashtekar developed the program starting in the late 1980s and through the 1990s, building on a reformulation of general relativity Ashtekar had introduced in 1986. In a 2021 review, Ashtekar and his collaborator Eugenio Bianchi described the central lesson of the program in a single phrase. Gravity is spacetime geometry, and so geometry itself must be quantized.

Spin networks: a Penrose idea from 1971

The mathematical key was invented decades before the theory that now depends on it.

In 1971, the British mathematical physicist Roger Penrose, who would later share the 2020 Nobel Prize in Physics for his work on black hole singularity theorems, introduced a curious combinatorial object he called a spin network. At the time it was a piece of pure mathematics, an attempt to build space out of nothing but discrete relationships and whole numbers, with no continuous geometry assumed underneath. Penrose’s motivation was philosophical as much as physical. He wanted to see if a notion of three-dimensional direction could be recovered from purely combinatorial data, without any background space sitting behind it.

It was elegant, and for years it was largely a curiosity. The machinery that would let it describe real gravity did not yet exist.

Two decades later, Rovelli and Smolin discovered that spin networks were exactly the right basis for the quantum states of geometry in loop quantum gravity. The match was not approximate. Spin networks turned out to be the eigenstates of the operators that measure area and volume in the quantum theory of geometry. A piece of pure mathematics invented for one reason had been waiting for its physics application.

A spin network, in the loop quantum gravity framework, is a graph. The edges of the graph carry numbers called spins, which are half-integers and which set the quantum unit of area each edge contributes when it crosses a surface. The nodes of the graph carry numbers called intertwiners, which set the quantum unit of volume each node contributes. The whole graph encodes a quantum state of three-dimensional geometry.

Crucially, the spin network is not embedded in space the way a city map is laid on top of a region. It does not describe a fabric draped over a hidden stage. It is the fabric. There is no underlying continuous space against which the network is defined. There is only the network, with its discrete relational structure.

The area operator and the discrete spectrum

The most striking result of loop quantum gravity, the result that gave the field its first sustained excitement, was a calculation Rovelli and Smolin published in 1995. The operator that measures the area of any two-dimensional surface in space turns out to have a discrete spectrum.

The formula is precise. If a single edge of a spin network punctures a two-dimensional surface transversely, it contributes to the area an amount proportional to:

A = ℓₚ² × √(j(j+1))

where ℓₚ is the Planck length, approximately 1.6 × 10⁻³⁵ meters, and j is the spin quantum number on that edge. If the surface is punctured by multiple edges, the contributions add. The total area cannot take just any value. It can only take values built from sums of these discrete contributions.

The smallest non-zero area is roughly the Planck length squared, which is approximately 2.6 × 10⁻⁷⁰ square meters. A volume operator was constructed soon afterward, and it too has a discrete spectrum. The geometric quantities most basic to ordinary intuition, area and volume, are not continuous in this framework. They come in chunks.

This is the technical statement behind the popular phrase atoms of space. The 2004 Scientific American article by Smolin, illustrated with diagrams of spin network states corresponding to polyhedra, presented the picture to a general readership and gave the framework its public face. The phrase has stuck.

The cartoon to avoid

When people hear that space is discrete, they picture a grid. A three-dimensional graph-paper lattice. Tiny cubes of space stacked like bricks. A literal pixelation of reality with hard edges at the Planck scale.

That image is wrong, and the reason it is wrong is a principle at the very heart of general relativity called diffeomorphism invariance. It means, roughly, that the location of the graph in space carries no physical meaning. If you take a spin network and slide it, stretch it, deform its embedding, move the nodes around without cutting any links, you have not produced a different physical state. You have produced the same physical state described differently.

In the theory, the physical states are not the embedded graphs themselves but equivalence classes of them. Whole families of deformed pictures, all counted as one. There is no fixed lattice pinned to fixed points of space, because points of space in that absolute sense do not exist.

What is discrete is the relational structure: how many links, what spins, how they connect, and the spectra of the operators built from it. What is not discrete is a literal pixel grid you could, in principle, photograph. The diagram to keep in mind is not graph paper. It is a knot. What matters is how it is tied, not where on the table it sits.

Spin foams and the evolution of geometry

A spin network captures the quantum state of space at a single moment. To do dynamics, the theory needs an account of how one spin network evolves into another. The construction physicists use is called a spin foam.

A spin foam is what you get when you take a spin network and trace its history. The edges become surfaces. The nodes become edges. The whole structure becomes a two-dimensional object embedded in something analogous to four-dimensional spacetime. Each spin foam represents a possible history of geometry, a possible way one quantum state of space transitions into another.

The covariant formulation of loop quantum gravity, developed primarily in the 2000s, uses spin foams as the analog of Feynman diagrams in particle physics. Just as a Feynman diagram represents one possible history of how particles can interact, a spin foam represents one possible history of how geometry can evolve. The full physical amplitude for a transition is a sum over all such histories.

This is not a film of space evolving. There is no master clock ticking the spin foams from one frame to the next. The spin foam is itself the object the theory considers fundamental. Time, as a parameter external to the geometry, is not present.

The Wheeler-DeWitt equation: where time disappears

The absence of time at the deepest level of loop quantum gravity is inherited from an older equation. In 1967, the American physicist Bryce DeWitt, building on work by John Wheeler, published the canonical quantum gravity analog of Schrödinger’s equation for the cosmos.

Written compactly, it is Ĥψ = 0. The Hamiltonian operator, acting on the wavefunction of the universe, gives zero. This is the Wheeler-DeWitt equation, and it contains no time variable.

Compare it to ordinary quantum mechanics. The Schrödinger equation has a partial derivative with respect to time. It describes a wavefunction that evolves. The Wheeler-DeWitt equation has no such term. It describes a wavefunction that simply is.

The technical name for this situation is the frozen formalism. The cosmos, in this formulation, does not run. It stands.

Now hear carefully what that does and does not mean, because it is the single most abused idea in the whole subject. Time is absent from the fundamental equation is not the same as nothing ever happens. Things plainly happen. You are reading this. The seconds are passing. Change is the most obvious fact of experience. What the frozen formalism says is narrower and stranger. The familiar external time parameter has dropped out of the deepest description, and the experience of time and change must be recovered in some other way.

The standard recovery is called relational time. You do not ask how the universe evolves in time. You ask how everything else in the universe looks when a chosen field, serving as a clock, has a particular value. Then the next. Then the next. Out of that relational bookkeeping, the familiar river of time re-emerges. Not as a fundamental given, but as a pattern in how the parts of the world keep pace with one another.

In Ashtekar’s words, writing about quantum cosmology, there is no external time to phrase questions about evolution. Time, in this picture, is not a container. It is a relationship.

The flickering misreading

The headline claim that opened this article — that you flicker out of existence trillions of times per second — comes from naively turning the Planck time into a universal frame rate. The Planck time is the fundamental tick scale of the theory, approximately 5.39 × 10⁻⁴⁴ seconds. The reciprocal, the rate at which a hypothetical cosmic strobe would update, is roughly 1.85 × 10⁴³ per second.

Set that against trillions per second, which is 10¹². The naive number is larger than a trillion by a factor of 10³¹. Trillions per second undersells the headline figure by thirty-one orders of magnitude. It is not even in the right universe.

But the deeper problem is not that the number is too small. It is that the number should not be computed at all, because the theory does not license it. Loop quantum gravity does not assert a universal Planck-rate clock ticking everywhere in unison. It does not say that macroscopic persons blink out of existence and back at any rate. Not trillions per second. Not 10⁴³ per second. Not any rate.

Remember the problem of time. There is no master clock to tick. And remember what the discreteness actually is. A discreteness in the spectrum of geometric operators. In the possible answers to how much area is here. Not a literal strobe light flashing the cosmos on and off.

The flicker, as a literal cosmic strobe, is not there.

Loop quantum cosmology and the Big Bounce

Loop quantum gravity does make one prediction vivid enough to picture, though even here the precise version of the theory matters. Apply the loop techniques not to the whole universe but to a simplified, symmetric model of an expanding cosmos, and you get a subfield called loop quantum cosmology.

This is a symmetry-reduced approximation, not the full theory. Results proven there do not automatically hold in general. But within the model, something striking happens to the beginning of the universe.

In standard general relativity, if you run cosmic expansion backward, everything converges to a single point of infinite density. The Big Bang singularity, where the equations break down and physics ends. Loop quantum cosmology replaces that catastrophe with something else.

Its effective equation modifies the standard cosmological law with a single correction term. The Hubble parameter squared, H², equals the usual 8πG times the density divided by 3, but now multiplied by the factor (1 − ρ/ρ_critical), where ρ_critical is a maximum density set by the Planck scale. As the density climbs toward the critical value, the term in parentheses falls toward zero, and the expansion rate is forced to vanish. The universe stops collapsing and rebounds.

The singularity is replaced by a bounce. The Big Bang becomes, in this model, a Big Bounce, a transition from a previous contracting universe to our expanding one. The foundational paper establishing this, by Ashtekar, Tomasz Pawlowski, and Parampreet Singh, appeared in 2006.

The critical density at which the bounce occurs is about 0.41 times the Planck density. An enormous density, but a density, not a size. Ashtekar’s review provides a concrete example. A model universe that grows to a maximum radius of a megaparsec, millions of light-years across, undergoes its quantum bounce at a minimum volume of about 5.7 × 10¹⁶ cubic centimeters. That is not a pinpoint. It is a sizable volume. The trigger for quantum gravity is not that the universe is small but that it is dense.

Can any of this be tested?

The honest answer is that direct tests are hopeless. The Planck scale is roughly twenty orders of magnitude beyond any conceivable instrument. Physicists must look for indirect echoes.

The most discussed echo concerns a deep question. If space has a smallest scale, does that smallest scale break Lorentz invariance, the symmetry that guarantees the laws look the same to all observers moving at constant velocity and that nothing outruns light? Some models of Planck-scale discreteness predict that very-high-energy light should travel at a slightly different speed than low-energy light. The granularity of space should, over enormous distances, smear a flash of light by energy.

In 2009, the Fermi Gamma-ray Space Telescope observed a gamma-ray burst called GRB 090510 and found no energy-dependent delay. High-energy and low-energy photons arrived together, setting a bound that pushed the simplest such effects beyond the Planck energy itself. A later analysis tightened the limit further. In 2011, observations of another burst, GRB 041219A, by the INTEGRAL satellite, used the polarization of its light to place tight constraints on a related effect called vacuum birefringence.

These results are real and important. But here too, precision matters. They constrain specific Lorentz-violating effective models. They do not test loop quantum gravity as a whole, because the discreteness in loop quantum gravity is relational and does not obviously single out a preferred frame at all. The experiments rule out some ways space might be grainy. They leave the careful version untouched, and unconfirmed.

What the honest landscape looks like

Loop quantum gravity is not alone in the field. It is one of several serious, competing programs, each making different bets about what lies beneath spacetime. String theory proposes that the fundamental objects are not points but tiny vibrating strings, and that gravity emerges, unified with the other forces, in a framework that typically requires extra dimensions. Causal set theory, developed by Rafael Sorkin and others, proposes that spacetime is, at bottom, a discrete set of events ordered only by cause and effect. Causal dynamical triangulations, developed by Renate Loll, Jan Ambjorn, and Jerzy Jurkiewicz, builds spacetime from a path integral over tiny triangulated building blocks, glued together with a strict causal rule.

None of these has achieved decisive experimental confirmation. None can yet claim victory. The fair way to describe the landscape is not loop quantum gravity versus the truth but several serious attempts at a problem that remains genuinely unsolved.

What is known with relative confidence inside the loop quantum gravity framework is real and not small. The kinematical structure, the space of quantum states, can be built rigorously. The spin-network techniques do make the geometric spectra discrete. The cosmological model produces its bounce robustly in simple cases.

What remains genuinely open is also real. The full dynamics, governed by the Hamiltonian constraint, is not settled. The recovery of the smooth classical world we actually see has not been fully derived. The exact relationship between the full theory and its cosmological shadow is debated. The coarse graining of the spin-foam sum is an unsolved problem. And the verdict that matters most is the simplest one. No experiment yet selects loop quantum gravity over its rivals.

Carlo Rovelli, reviewing his own life’s work, wrote that direct or indirect experimental support is lacking and that the theory could very well turn out to be physically wrong. That is the voice of the science.

What the picture cannot answer

Loop quantum gravity, taken at its best, is a structure of breathtaking mathematical rigidity. Geometry quantized into precise discrete spectra. Black-hole entropy chained, exactly, to area in fixed units of the Planck length squared. The collapse of a universe turned, at a sharply defined critical density, into a rebound. A covariance so constrained that the location of the fundamental graph itself must be divided out as physically meaningless.

Whatever this is, it is not the residue of chaos. It is law upon law upon law, calibrated with a precision the older arguments between order and disorder never dreamed of. The fine-tuning the program inherits, the constants it presupposes, the free parameter it must set by hand to match the entropy of a black hole, every one of these is a place where the universe turns out to be not loose and accidental but tight, specific, and mathematically exact.

This is the question the picture cannot finally settle from inside itself. Whether such precision is the expected output of an unauthored process, or whether the order of this depth and consistency points beyond itself to a source, is not a question loop quantum gravity is equipped to ask. It is the question the structure of the theory silently raises.

This is the question the companion documentary on the Sleepy Joe Space YouTube channel takes up at length.

What stays

What stays from loop quantum gravity is a picture of space far stranger than ordinary intuition allows, and far more disciplined than the popular cartoons suggest. Space, at the Planck scale, may not be a smooth stage that goes down forever. It may be built from discrete quanta of geometry, encoded in spin networks, with area and volume coming in countable chunks. The continuity of the world you experience may be an excellent approximation emerging from a fundamentally granular foundation.

What stays equally is the discipline. The discreteness is spectral and relational, not a literal lattice. The timelessness of the Wheeler-DeWitt equation does not abolish the time you live in; it relocates time as a relational, emergent feature of large, smooth systems. The cosmic flicker is a misreading. The bounce is real within its model but does not, by itself, explain why the underlying laws and constants take the values they do. The theory is mathematically rigorous and experimentally unconfirmed.

If space is granular all the way down, the smoothness you experience is an appearance. An appearance requires someone to whom it appears, and that someone, whatever else it turns out to be, is the one continuous thing the theory describes around but cannot produce.

The river is real even if it rises from a still and timeless spring.