← quantum · contents

§ Phase xiv · Fields as Particles · Chapter 01

Fourier modes as quanta

By the late 1920s ordinary quantum mechanics had quantized atoms and electrons but not light. In 1927, working at a small writing desk in Cambridge, Paul Dirac asked the obvious next question. Take the electromagnetic field, decompose it into Fourier modes, and treat every mode as a harmonic oscillator. The rungs of those infinite ladders, he saw, are photons. The same trick, applied to every other field in nature, turned out to be the rest of physics.

The book so far has been about particles. An electron lives in three dimensions of space and one of time, its position uncertain by the Heisenberg rule, its energy quantized into discrete orbital ladders. Light, in the same story, has played a peculiar double role. We have used Einstein’s photons as bookkeeping for emission and absorption, and we have used Maxwell’s continuous fields when we wanted to draw pretty pictures of polarization. We have never reconciled the two. The picture of “a particle in a wave function” works admirably for one electron in a box, but it does not tell you what light is, and it does not tell you what you should do when a high-energy photon converts into an electron-positron pair on a Geiger counter at three in the morning.

The fix, when it came, was a reversal so simple it is hard to believe how long it took. Ordinary quantum mechanics takes a classical particle, with its position and momentum, and promotes those numbers to operators. The new prescription, born in Cambridge in 1927, takes a classical field, with its value at every point of space, and promotes that whole field to an operator. Particles, in this rewritten theory, are no longer the primary objects. They are the rungs on the energy ladders of the field’s normal modes. The electromagnetic field is the fundamental thing in nature. A single photon is a small, integer excitement of one of its harmonic modes.

The young man who first wrote this down was 25 years old. He had finished his PhD two years earlier on a fellowship from St John’s. He worked alone, in pen, almost in silence, and he had already in 1925 invented his own version of matrix mechanics during a single intense weekend of reading. Heisenberg later said of him that he gave the impression of being “in a class by himself.” His office was austere. His meals were taken alone. The 1927 paper he produced (which we will read several quotations of below) is fifty-five pages of cool mathematical prose that quietly demolished the old division between matter and radiation. From here on, in this book, particles and fields are no longer two different kinds of stuff. They are the same stuff, seen at two different magnifications.

To see why a field looks like a collection of oscillators, start with a familiar mechanical analogue. Stretch a string between two fixed points. Pluck it. The string vibrates, and its motion is a sum of standing-wave modes: a fundamental tone, a second harmonic at twice the frequency, a third at three times, and so on without end. Each of those modes is mathematically a one-dimensional oscillator. The total motion of the string is the sum, mode by mode, of independent oscillators. There is nothing quantum about this; it is just Fourier’s 1807 theorem applied to a classical wave equation. Anyone who has tuned a guitar has felt it directly. The same trick works in three dimensions for a vibrating drum, for the air in a flute, for the surface of a pond. The mode decomposition is universal, because the wave equation is linear and translation symmetric.

Now imagine the same decomposition applied to a field that fills all of space. Take a scalar field φ(x, t), some number defined at every point. The classical equation of motion (Klein and Gordon would soon write it down in relativistic form) is again a linear wave equation. Fourier expand it. Every plane wave with wave vector k contributes an independent piece, each piece oscillating at its own characteristic frequency ω_k. The whole field, like the plucked string, is mathematically a vast collection of independent oscillators, one per allowed momentum. The only difference from a string is that there are infinitely many modes, packed at every wavelength shorter than the size of the universe and longer than the smallest length physics knows how to discuss.

So far, none of this is quantum. The leap is to treat each oscillator quantum mechanically. We already know what that means: a quantum oscillator at frequency ω has discrete energies E_n = (n + ½) ℏω for n = 0, 1, 2, … It has a creation operator a† that raises n by one, an annihilation operator a that lowers n by one, and a vacuum state |0⟩ at the bottom of the ladder. Do this for every Fourier mode of the field. Each mode k now carries its own pair of operators a_k and a†_k, its own ladder of states, and its own integer occupation number n_k. The total state of the field is the list of all those integers.

What is the physical content of all those integers? The bookkeeping is that n_k particles of momentum k are present in the field. A field in state |n_k = 1, everything else zero⟩ is a single particle, with momentum ℏk, energy ℏω_k. A field in state |n_k = 2⟩ is two particles, both with momentum ℏk, indistinguishable in every measurable sense. A field in state |n_k = 1, n_q = 1⟩ is one particle of momentum ℏk plus one of momentum ℏq. The infinite tower of states of one harmonic oscillator, which we used in earlier chapters to describe the rungs of a single trapped vibration, has been silently reinterpreted: the n-th rung of mode k now means “n particles of momentum k.” A single field has become an inexhaustible factory of particles.

The vacuum is the state in which every mode sits at n = 0. It is the lowest-energy state of the system, the one no operator can lower further. Notice what it is not. It is not empty space. The harmonic oscillator at n = 0 is not stationary; it has the irreducible ½ ℏω of zero-point energy required by the Heisenberg uncertainty principle, because position and momentum cannot both be zero at once. Every Fourier mode of every field contributes its own ½ ℏω_k to the vacuum energy. The empty universe, in this language, is a seething superposition of zero-point oscillations across every wavelength of every field that exists. A great deal of the next chapter is about what that actually means and which of its predictions can be measured. For now, hold on to the slogan: the vacuum has structure. Nothing in field theory is ever truly nothing.

The natural mathematical home for all this is the ladder of operators. For each mode k of a free scalar field, define a creation operator a†_k and an annihilation operator a_k. They satisfy what is called a commutation relation, [a_k, a†_q] = δ(k - q), an algebraic identity that is the field-theoretic analogue of [x, p] = iℏ from chapter one. The vacuum |0⟩ is defined by a_k |0⟩ = 0 for every k. A one-particle state of momentum k is built by acting once with a creation operator:

|k⟩ = a†_k |0⟩

A two-particle state of momenta k and q is built by acting twice:

|k, q⟩ = a†_k a†_q |0⟩

The commutation relation determines whether a†_k and a†_q commute, in which case swapping the two creations gives the same state and the particles are bosons (photons, gluons, the Higgs). For half-integer-spin fields one writes a different algebra, the anticommutator {b†_k, b†_q} = 0, which forces the state to flip sign under exchange. That sign flip is the Pauli exclusion principle of chapter ten, rewritten one layer deeper. Bose statistics and Fermi statistics are not extra postulates of nature. They are properties of the algebra of creation operators.

A(x, t) = Σ_{k, λ} √(ℏ / 2 ω_k V) [ a_{k,λ} ε_{k,λ} e^{i(k·x - ω_k t)} + h.c. ]

H = Σ_{k, λ} ℏ ω_k ( a†_{k,λ} a_{k,λ} + ½ )

This reorganization of physics dissolves one of the deepest puzzles of pre-1927 quantum theory: why all electrons are identical. In the particle picture you would have to postulate it as an axiom. Why should two electrons in two different atoms have exactly the same mass, charge, and spin, to the precision of any experiment ever performed? In the field picture there is nothing to explain. There is only one electron field in the universe. Every electron we ever measure is one excitation of that single underlying field. Two electrons in two different atoms are not two different objects with mysteriously matching properties. They are two ripples on the same global pond. The “identicalness” of identical particles, which Pauli used so heavily in chapter ten when writing down the exclusion principle, is built into the architecture of the theory. It is not extra. It is the only thing the theory could mean.

The same observation reorganizes the question of particle creation and destruction. In ordinary quantum mechanics the number of particles is fixed. The Schrödinger equation evolves a fixed-N wave function. There is no operator in that machinery that can take you from one electron to two, or from a photon to nothing. Yet experiment is full of such processes: a beta decay produces an electron that was nowhere a moment before, a high-energy photon converts into an electron-positron pair, a positronium atom annihilates into two photons. Each of these is forbidden in single-particle quantum mechanics. Each is routine in field theory. The creation operator a†_k literally creates a particle that was not there a moment before, by adding one to the integer n_k. The annihilation operator a_k literally destroys it. Interactions in the field Hamiltonian (terms that couple different fields together) shift quanta between modes and between species. Pair production is just an interaction vertex that creates a quantum of the electron field, a quantum of the positron field, and annihilates a quantum of the photon field, all at the same spacetime point. The Geiger counter clicks because the field rearranged its integer ladders.

It is easy, reading a chapter like this, to slip past how strange this picture is. The world contains, on this account, no fundamental particles at all. It contains fields, perhaps a few dozen of them in the modern Standard Model: an electron field, a muon field, an up-quark field, a down-quark field, a photon field, a gluon field, a Higgs field, and so on. Every observed particle is an excitation in one of those fields. The fields fill space whether or not anything has been excited; the vacuum is the state in which every ladder sits on its bottom rung. Everything else (atoms, chairs, planets, the chemistry of life, the light from distant galaxies) is a long enumerative description of which rungs of which ladders are currently up, and at which positions. Quantum field theory in 1927 was a piece of mathematics. By 2026 it has become the cleanest answer we have to the question “what is matter made of,” and the answer is that matter is not made of anything at the bottom. It is a pattern of integers in a field of oscillators.