Fields That ClickBonus ② / Where Does the Field Come From (Gauge) — series finale

Until now we've read fields as "something given." The final question — why does the field exist in the first place?

Where Does the Field Come From Phase carries an ambiguity of "you can take any reference you like." The moment you demand that it be
freely chosen at each point — a field is born, as a logical necessity, as the partner that balances the books.

Tools you need: all 6 episodes + Bonus ①, the phase (angle) of a complex number This episode's core: ∂ → D = ∂ − iA

This series has consistently kept saying one thing — units-carrying quantities are stage machinery chosen by humans; what decides the physics is only the ratios without units. Meters, seconds, newtons are our conventions. So what happens if we push that idea all the way to the end? The quantum world too has one great "freedom of convention" — the reference of a wave function's phase. Where you set zero is nothing but how we keep our books. Here we ask boldly — does that ledger have to be aligned identically everywhere in the universe? Couldn't we choose it separately at each point? Pursue the answer and you arrive at one of the deepest strikes in twentieth-century physics — the moment you allow the phase to be freely chosen at each point, a field is born to patch up the discrepancy. The field is not "something that exists from the start" but the inevitable partner for allowing local freedom of convention.

01The Ambiguity of Phase — Rotating It Globally Does Nothing

In quantum mechanics, a particle is represented by a complex wave function \(\psi=|\psi|e^{i\theta}\), with a magnitude \(|\psi|\) and a phase (the angle \(\theta\)). But only the probability \(|\psi|^2\) is observable. So even if you rotate the phase everywhere at once by \(\theta\) ——

Global phase rotation (a global transformation) — no effect on observation
$$\psi\to e^{i\theta}\psi\quad\Rightarrow\quad|\psi|^2\to|e^{i\theta}\psi|^2=|\psi|^2$$

The probability doesn't change one bit. Where you set the zero of phase is how you keep your books, and the universe doesn't care. Exactly the same structure as when we said "units are conventions" in Episode 4 — the reference of phase is also a unitless convention (gauge).

02Rotate Freely at Each Point — and "the Difference From the Neighbor" Breaks

Now rotate the phase by a different \(\theta(x)\) at each place (a local transformation). If rotating globally is harmless, choosing it at each point ought to be allowed too. But physics contains a derivative = the operation "measure the difference from the neighboring point," and that causes trouble.

Try it — differentiate a local phase and an extra term appears $$\partial_x\!\left(e^{i\theta(x)}\psi\right)=e^{i\theta(x)}\Big(\partial_x\psi+i(\partial_x\theta)\,\psi\Big)$$

On top of the original \(\partial_x\psi\), an extra term \(i(\partial_x\theta)\psi\) appears. If \(\theta\) is constant (global rotation) then \(\partial_x\theta=0\) and it vanishes, but with a location-dependent \(\theta(x)\) it doesn't. Because of this term, choosing the phase freely at each point changes the form of the equation — the physics comes to depend on "how I keep my books." That's a problem.

03The "Partner" That Balances the Books — That Is the Field

What to do? There is one strategy — introduce a partner field \(A\) that cancels exactly that extra term, and replace the derivative with a "derivative accompanied by a partner (the covariant derivative)."

The covariant derivative — a derivative taking the partner A along
$$\partial_x\ \longrightarrow\ D_x=\partial_x-iA_x$$

And impose the rule that when you rotate the phase by \(\theta(x)\), at the same time the partner moves in step, \(A_x\to A_x+\partial_x\theta\). Then the extra term \(+i(\partial_x\theta)\) from the local phase and the partner's change \(-i(\partial_x\theta)\) cancel exactly, and the equation becomes invariant under relabeling of the local phase. This is gauge invariance.

And — the true identity of the partner \(A\) we just introduced "to balance the books" is precisely the electromagnetic field (the electromagnetic potential). Both the electric and magnetic fields come out of this \(A\). In other words, the order is reversed from the usual. It's not "the electromagnetic force exists, so the electron moves" — rather, because we want to choose the phase freely at each point, the electromagnetic field is inevitably required. The ambiguity (freedom of convention) comes first, the field after. The very freedom of convention this series has kept calling "stage machinery" was the mother that summons the field.

04Play With It — the Partner A Protects the Physics

The figure below. The upper row is the phase of each wave function at a row of points (clock hands). With the slider "local gauge transformation," you rotate the phase at each point differently from place to place (= relabeling the books arbitrarily at each point). The lower row is the substance of the physics = the consistency with the neighbor (the covariant derivative \(D\psi\)).

With partner A: on, no matter how much you apply the local transformation, the lower row (the physics) stays perfectly constant — however we choose the phase, the physics is invariant (gauge invariant). With partner A: off, every local transformation makes the lower row scatter into disorder — the physics comes to depend on our arbitrary bookkeeping. The partner \(A\) that soaks up this disorder and protects the physics is exactly the electromagnetic field.

Figure: top = the phase of each wave function (clock hands). The "local gauge transformation" rotates each point differently. Bottom = the substance of the physics (consistency with the neighbor Dψ). With partner A on, the physics stays constant (invariant); off, it scatters. The partner A that soaks up the disorder is the electromagnetic field
wave-function phase ψ (top row) substance of the physics Dψ (bottom row)

05And Everything Connects

With this one strike, the whole series is tied into one strand. How strongly the partner \(A\) bonds with the wave function — that strength is Episode 4's \(\alpha\approx1/137\). The speed at which the partner \(A\) travels is Episode 2's \(c\) (because its carrier, the photon, has zero mass, hence \(c\)). Give that photon a mass via the Higgs of Episode 6, and you get Episode 3's short range (the weak force). And when the ambiguity of phase becomes a more complex "ambiguity of an internal direction," the same logic gives birth to the weak force and the strong force — all three forces are the result of "allowing some freedom of convention at each point." A field is the partner that keeps a local ledger consistent. That is what appeared to us in the guise of unitless ratios: lag (Episode 1), speed (Episode 2), range (Episode 3), strength (Episode 4), thinning-out (Episode 5), mass (Episode 6). The units-carrying numbers were stage machinery to the very end.

◇ ◇ ◇
The honest line — is the gauge principle a "proof" or a "principle"?

That demanding invariance under local \(U(1)\) phase transformations inevitably produces the covariant derivative \(D_\mu=\partial_\mu-iqA_\mu\) and the electromagnetic potential \(A_\mu\); that its coupling constant is \(\alpha\); and that localizing the non-abelian groups \(SU(2)\times SU(3)\) gives the weak and strong forces (the Standard Model) — these are established physics. The connections to Episode 6's mass (a massive gauge boson is short-ranged), Episode 2's \(c\), and Episode 4's \(\alpha\) are genuine too.

But "demanding local invariance" is not a proven necessity — it is a guiding principle — and why nature obeys this principle remains as a deeper mystery (though the Standard Model built from this principle matches experiment to extraordinary precision, so it is not mere taste). Also, this piece is an intuitive version of the skeleton of the gauge principle; actual quantum field theory layers on top of it renormalization, spontaneous symmetry breaking (Episode 6), non-abelian structure, and more. The aim of this bonus is to grasp the shift of viewpoint that "a field can appear as the partner for allowing local freedom of convention." For deeper development, see the sister series "Cosmology That Clicks" Episode 8 and "Force That Clicks" Episode 9.

Practice problems (solvable with this episode's ideas)
  1. Under a global phase rotation \(\psi\to e^{i\theta}\psi\) (\(\theta\) constant), how does the probability \(|\psi|^2\) change?
    Show the answer
    \(|e^{i\theta}\psi|^2=|\psi|^2\). It doesn't change at all. So the global reference of phase is a convention (gauge) that has no effect on observation.
  2. What is the "extra term" that appears when you differentiate a local phase \(e^{i\theta(x)}\)? Why does it vanish for a constant phase?
    Show the answer
    The extra term is \(i(\partial_x\theta)\psi\). If \(\theta\) is constant then \(\partial_x\theta=0\) and it vanishes. Only when \(\theta(x)\) varies from place to place does it remain, and to cancel it a partner \(A\) is needed.
  3. What is the true identity of the partner \(A\) introduced via the covariant derivative \(D_x=\partial_x-iA_x\)? State the reversal of order in one line.
    Show the answer
    The partner \(A\) is the electromagnetic field (the electromagnetic potential). It's not "a force exists so we can allow the ambiguity," but "we want to allow the ambiguity at each point, so a force = field inevitably appears" — the ambiguity first, the field after.
  4. The partner \(A\)'s coupling strength, its speed of travel, and the range once it is given a mass — with which episodes of this series does each connect?
    Show the answer
    Coupling strength = Episode 4's \(\alpha\). Speed = Episode 2's \(c\) (the photon has zero mass). Giving it a mass = the Higgs of Episode 6 → the short range of Episode 3, \(\lambda=\hbar/mc\). Gauge is the core of it all.

Bonus ② Summary / Fields That Click, completeThe Field Was the Partner That Keeps Conventions Consistent

The reference of a wave function's phase is, like units, a convention (gauge). Rotating it globally has no effect on observation. But localize it — allow it to be freely chosen at each point — and the derivative produces the extra term \(i(\partial_x\theta)\), breaking the equation. Introduce the partner \(A\) that cancels it via the covariant derivative \(D_x=\partial_x-iA_x\), and that \(A\) is precisely the electromagnetic field. The ambiguity (freedom of convention) comes first, the field after. The field does not exist from the start; it was the inevitable partner for keeping a local ledger consistent.

With this the series is tied into one strand. The partner \(A\)'s coupling is Episode 4's \(\alpha\), its speed is Episode 2's \(c\), given a mass it becomes Episode 3's short range, and with a complex ambiguity, the weak and strong forces. The field appeared in the guise of unitless ratios — lag (Episode 1), speed (Episode 2), range (Episode 3), strength (Episode 4), thinning-out (Episode 5), mass (Episode 6). Think from there and it never gets complicated — the units-carrying numbers were stage machinery to the very end, and what decided the physics were the ratios alone. This is everything the six episodes and two side roads of "Fields That Click" wanted to say.

This document is Bonus ② (finale) of the "Fields That Click" series, reading for physics-loving high-schoolers and undergraduates. That demanding invariance under a local \(U(1)\) gauge transformation \(\psi\to e^{i\theta(x)}\psi\) inevitably leads to the covariant derivative \(D_\mu=\partial_\mu-iqA_\mu\) and a gauge field (the electromagnetic potential \(A_\mu\)); that its coupling constant is the fine-structure constant \(\alpha\); that localizing the non-abelian gauge groups \(SU(2)\times SU(3)\) gives the weak and strong interactions and constitutes the Standard Model; and that a massive gauge field (spontaneous symmetry breaking, Episode 6) brings about a short range (Episode 3) — all of these are established standard physics. That "demanding local invariance" is not a proven necessity but a guiding principle whose own grounding is an open problem (though the Standard Model agrees with experiment to high precision), and that this piece is an intuitive version of the skeleton, omitting renormalization, non-abelian structure, and the like — these are stated explicitly in the "honest line" in the body. The figure is a schematic of the invariance of the covariant derivative (Dψ) under local phase transformations, not a numerical lattice-gauge-theory calculation. — To print, use your browser's "Print" and "Save as PDF" (in the print version the slider and answers are static and hidden). Related: Episode 1, Lag / Table of contents / sister series Cosmology That Clicks · Force That Clicks.

Print / save as PDF: ⌘+P (Ctrl+P on Windows). On screen, the "local gauge transformation" slider and "Partner A: on/off" toggle whether the physics (bottom row) stays invariant or scatters. "Show the answer" opens each solution.