Why does Fourier analysis appear everywhere in quantum mechanics?

A chord, not a crowd

Strike three keys on a piano. The air in the room carries one pressure wave. There are not three waves taking turns, and there is no gap between them; at any instant the air at your ear has one pressure, and that pressure rises and falls in a single complicated pattern.

Yet you hear three notes. Your ear does what Fourier showed could always be done: it takes the one wave apart into sinusoids, pure tones of definite frequency, and finds three families of them, one for each key. The notes are not an illusion. A tuning fork in the room, cut for one of those frequencies, will start to hum on its own, and ignore the other two. The notes are really in the wave. But they are not three things in the room. They are the structure of one thing.

Hold that picture. Much of what puzzles people about quantum mechanics is a version of it.

Why sines?

Fourier’s theorem says that any reasonable wave can be written as a sum of sinusoids. That is mathematics, and it is true of any wave at all. The physics is in the next step: for a wave in a medium that is the same everywhere, the sinusoids are not just a convenient bookkeeping; they are the wave’s natural motions.

Here is why. Slide a pure sinusoid along and you get the same sinusoid back, shifted in phase: it is the one shape that cannot tell where it is. If the medium is the same everywhere, and the wave equation is linear, nothing can make it change shape. It can only grow, shrink, or slide. And because the equation is linear, two solutions add to a third, so each sinusoid can be followed on its own. A sinusoid of wavelength $\lambda$ oscillates at its own frequency $\omega$, and that is all it does:

In words: $A$ is the wave’s height, and each mode keeps its wavelength, turns at its own rate, and never changes its shape. Decompose a wave into other pieces, into bumps say, and in general each piece changes shape as it travels, spreading until the pieces overlap and stop being the pieces you started with. Decompose it into sinusoids and no piece ever changes shape; only its height and its phase move. That is why Fourier analysis appears wherever there is a linear wave in a uniform medium: sound, light, water, and the quantum field.

It is also why position and momentum come as a Fourier pair in quantum mechanics. De Broglie’s relation ties a wave’s momentum to its wavelength, so a state’s momentum content is its Fourier spectrum, and a wave confined to a small region must contain a wide spread of wavelengths. The position–momentum uncertainty relation, seen this way, is a theorem about waves, as true of a short pulse of sound or radio as of an electron; quantum mechanics adds only that the wave’s spread is the spread of what a measurement will find.

The quantum field is a Fourier composition

Quantum field theory, the framework that underlies particle physics, describes the world as fields. The electron field, the electromagnetic field, and the rest are waves in a space that is the same everywhere, and their free equations are linear. So their natural motions are Fourier modes, and the theory is built on them. The field is written as a sum over modes of every wavelength. For light, each mode is a little oscillator with its own frequency.

Now the word “particle.” What quantum field theory calls a particle is not a small body that carries a wave around with it. It is one quantum of excitation of one mode: the smallest amount by which a mode can be excited. One electron is one quantum in the electron field, and a mode of the electron field can hold one quantum or none. The modes come first, and the countable things are what the modes can hold. ACT keeps the word “particle” for something else, the event, and the end of this article says why.

A single photon “in two places at once” shows what the mode picture buys. Write the photon’s state as “here” plus “there” in equal parts, and it looks like a paradox. Now notice that “here plus there” is itself a mode, the one the two paths of the apparatus offer: one two-lobed shape, built from many sinusoids, with one hump on each side. The photon is one excitation of that one mode. Nothing here is two of anything. It is one note, with a shape that happens to have two humps, the way a guitar string’s second harmonic has two. This article stays with one quantum; with two or more, the composition is of the whole field’s state, which is where entanglement lives, and that is a subject for another article.

Superposition is composition

The textbook word is superposition, and the word misleads. It suggests two things stacked on top of each other, two possibilities held in suspense. The mathematics says something plainer: one wave, with components, as a chord has notes. The components are real in the way the notes are real. They carry amplitude. Each evolves on its own. A tuner can pick one out. And while nothing has recorded which component is which, they can still interfere, as two notes beat against each other in the air.

What the components are not is a crowd of things waiting to see which one wins. They are the structure of one wave in which, so far, nothing has happened. Anchored Causality Theory keeps two senses of “real” apart for exactly this reason. A component is real as structure: it is part of what the wave is. An alternative becomes real as a fact only when a record has made it actual. The chord is real before anyone plays a note of it alone.

Here the chord stops helping. Put three forks in the room and all three hum; send one quantum in a two-lobed mode toward two detectors and only one fires. That difference is what the event law is for, and the rest of this article leads up to it.

The apparatus is a tuner

An FM radio does not create the station it plays. The space around you carries one electromagnetic field, the sum of every broadcast at once. The tuner, a resonant circuit, picks out one frequency and leaves every other station on the air. This is the right picture of what a measuring apparatus does to a quantum field, and each piece of the picture has been seen in the laboratory.

A cavity is a tuner for an atom’s light. An atom can emit only into a mode the field around it offers. Put it in a cavity tuned to the colour of light it emits and it emits faster: Goy, Raimond, Gross and Haroche saw this in 1983. Put it between plates too close together to carry light of its wavelength in the polarization it emits, and its emission is suppressed: Hulet, Hilfer and Kleppner, 1985. What leaves the atom is an excitation of a mode the surroundings make available, not a projectile that would fly whatever the surroundings.

A Stern–Gerlach magnet sorts the same way. Since 1922 it has split a beam of silver atoms into two beams along the magnet’s axis; by 1927 it had done the same with hydrogen, the simplest atom. In both, the magnetism is the spin of a single electron. Turn the magnet and it sorts along the new axis, as a new setting of the dial brings in a new station. And the magnet has found nothing out. In 2021 Margalit and colleagues built a full-loop Stern–Gerlach interferometer on an atom chip: the two spin-split wave packets of a cold atom cloud were brought back together, and the spin superposition came back whole, with high visibility. Sorting a wave into components can be undone, so long as nothing has kept a record of the split. It is not a measurement; it is tuning.

The apparatus picks out components. Only a record picks the outcome.

Where Fourier stops

A physicist reading this far will have an objection ready, and it is a good one. A state can be decomposed in many ways. The same wave is a sum of sinusoids, or of localized bumps, or of any other complete set of shapes you like. Mathematics does not prefer one. So why should the Fourier decomposition be “what the state is” rather than any other?

The answer is that the choice is not ours, and it is not Fourier’s either. The dynamics and the symmetry of the medium decide. For a free field in a space that is the same everywhere, the natural decomposition is Fourier, for the reason given above: those are the pieces that keep their shape. For a field meeting an apparatus, the coupling decides, and it selects the components it can hold stable. A spin sorted by a magnet is the reminder that the stable decomposition is not always literally Fourier. The surroundings decide too. In 2000 Turchette and colleagues built two kinds of noisy surroundings, reservoirs made to order, for one trapped ion. One jostled it with random forces, and destroyed superpositions of two swings at a rate that grew with their separation. The other jittered the trap’s stiffness, and destroyed superpositions of different numbers of quanta at a rate that grew with the difference. The theory those measurements confirmed says which states each kind of bath leaves alone: a steady swing under one, a definite number of quanta under the other. Change the coupling, change which components survive. Zurek gave this fact its name in 1981, the pointer basis, and the experiments bear it out: what an apparatus can record is an output of how it couples, not a choice of basis made by anyone once the coupling is fixed.

And there is a second limit, which matters more. Even once the dynamics have chosen a decomposition, the decomposition does not say which alternative becomes the fact. Decompose a chord into its three notes, and you have not thereby played one of them alone. The components coexist in the wave. Taking the wave apart into them explains their structure; it does not make any alternative actual. That step is not in the Fourier mathematics at all.

So the chain has three links, and Fourier is the first:

1. The wave has mode structure. Free fields fall into Fourier modes by themselves; this is where Fourier analysis lives.
2. The coupling selects which decomposition can be recorded. A magnet sorts by spin along its axis, a bath keeps whatever it can hold stable, and a cavity offers only the modes it has. This is the tuner.
3. A released record makes one alternative actual. This is the event, and it is a separate step with its own law.

Ontology recapitulates mathematics

Physics has profited before from reading its own mathematics as a description of what there is. Planck’s energy element $h\nu$ was, for Planck, a device for the oscillators in a hot cavity’s walls; Einstein took it to mean that light itself comes in quanta, and it took the field two decades to agree. Dirac’s negative-energy solutions were an embarrassment until, reread, they turned out to describe antimatter. In each case the formalism had been describing something real for years before anyone agreed to believe it. ACT’s wager differs in one respect, and the box below says so: the quanta and the positron earned belief with a new number; ACT changes none, and earns its place by the account it gives.

Anchored Causality Theory (ACT) makes that wager about the quantum field. The mathematics says: one field, composed of modes; particles, in field theory’s sense, as quanta of modes; an apparatus as a coupling that selects components. ACT takes that at face value and adds one commitment the mathematics leaves open: a released record makes one alternative actual. So the electron is a stable wave-mode structure of one field, never a particle on a path. A detector’s bath, the surroundings that soak up its heat and noise, keeps a record that tells the alternatives apart, and when the record has passed beyond recall, one alternative becomes an event: a mark in the one actual history. After an event, the alternatives the record did not select keep the first kind of reality and never acquire the second. They are not destroyed, and they are not things. They are structure of one wave in which no event has occurred. The wave is what exists; the particle is what happens.

The Fourier connection is the first link of that chain, and it earns its place because it is the one everyone can hear. Three notes in one wave. Three stations in one field. One photon in one two-lobed mode. Once the chord is the picture, superposition stops looking like two things held in suspense and becomes composition, and the question that remains is the one ACT is about: what makes one note sound alone.

Where to go next

1. J. Fourier, Théorie analytique de la chaleur (Firmin Didot, Paris, 1822).
2. P. Goy, J. M. Raimond, M. Gross & S. Haroche, Phys. Rev. Lett. 50, 1903 (1983).
3. R. G. Hulet, E. S. Hilfer & D. Kleppner, Phys. Rev. Lett. 55, 2137 (1985).
4. W. Gerlach & O. Stern, Z. Phys. 9, 349 (1922); T. E. Phipps & J. B. Taylor, Phys. Rev. 29, 309 (1927).
5. Y. Margalit et al., Sci. Adv. 7, eabg2879 (2021).
6. Q. A. Turchette et al., Phys. Rev. A 62, 053807 (2000).
7. W. H. Zurek, Phys. Rev. D 24, 1516 (1981); Rev. Mod. Phys. 75, 715 (2003).
8. K. Sonderegger, Anchored Causality Theory: A Record-History Ontology for Open Quantum Systems, preprint, version 3.6 (2026), doi:10.5281/zenodo.23811463.