Is the global form of the Standard Model gauge group a fact or a convention?

Object. The gauge group of the Standard Model is (SU(3)×SU(2)×U(1))/Γ for one of the four subgroups Γ ⊂ ℤ₆. All four give identical local physics (every correlation function of local operators). They differ in which line operators (Wilson and 't Hooft lines) exist, i.e. in the spectrum of allowed non-dynamical probes and monopoles, and in twisted sectors on non-simply-connected spacetimes.

Presenting problem. Two camps, both with computations behind them. - It's a choice. Aharony–Seiberg–Tachikawa, Tong: the Lagrangian does not determine Γ; specifying a theory requires specifying its line-operator content, and the four choices are all consistent. - It's forced. In Furey's division-algebra construction, the colour-electric group comes out as U(3) = SU(3)×U(1)/ℤ₃ because the particles are the algebra acting on itself and every one of the 64 operators in the Clifford algebra satisfies 3Q ≡ triality; no operator carrying the bare fundamental with the wrong charge exists anywhere. In the exceptional-Jordan construction (Dubois-Violette–Todorov) the group is S(U(2)×U(3)), which is the full ℤ₆ quotient. Every grand-unified embedding also yields ℤ₆.

Diagnosis offered. The two camps are answering different questions. The global form is decided by what exists (closure of the "room": if matter is the algebra and there are no probes outside it, the group is the faithful quotient) and measured by linking (line operators). The Lagrangian is silent because it doesn't specify the room. In logical terms Γ is an independent statement relative to the local axioms, and "the room is closed" is an added axiom that decides it, in the way V = L decides the continuum hypothesis.

Differential. 1. Nature's room is closed (no probes beyond matter): Γ = ℤ₆, and the monopole spectrum is that of the quotiented group. 2. Nature admits probes the algebraic frameworks lack: Γ is smaller, and any "matter = algebra" framework is falsified by a line operator. 3. The question is empirically inaccessible at any foreseeable energy and remains a convention.

Discriminating tests. Any observation of a 't Hooft line / magnetic charge in a representation forbidden by ℤ₆; any consistent algebraic completion that produces a smaller Γ (a finite check: enumerate its operators for a bare fundamental with the wrong charge).

Stakes. Whether "which group" is physics.

---

Update (2026-09-13): the analogous question is settled by instrument in an existing system.

A superconductor provides a worked example of a global-form question being answered with a magnetometer rather than argued.

The condensate is a charge-2e field, so the U(1) is broken by a charge-2 order parameter and the subgroup left unbroken is ℤ₂ rather than trivial. That subgroup determines the flux quantum: vortices carry Φ₀ = h/2e = 2.067834×10⁻¹⁵ Wb and not h/e. The factor of two is read off directly, and it is the standard experimental demonstration that the carriers are pairs.

So which quotient survives is not a convention there. It fixes an observable, it is measured routinely, and different global forms would give different instrument readings.

Effect on the differential. Branch 3 — that the question is empirically inaccessible at any foreseeable energy and remains a convention — is weakened, though not refuted. The superconducting case shows the type of question is answerable by measurement, which removes the general argument that global-form questions are meaningless. What it does not show is that the specific question here is answerable, since the obstruction there was never conceptual but a matter of energy scale: the line operators and magnetic charges that would discriminate are not accessible.

So the honest revision is that branch 3 must now be argued on energy-scale grounds alone, and can no longer lean on the stronger claim that global form is a matter of convention. Branches 1 and 2 are unaffected.

A test this suggests. Look for other condensed-matter systems in which the unbroken subgroup differs and the corresponding quantum changes — charge-4e superconductivity, if it exists, would give ℤ₄ and a quarter flux quantum. A second instance with a different subgroup would make the general point empirical rather than illustrative.