Generalized anomaly constraints on phases
Derive a mixed one-form background variation, cancel it by five-dimensional inflow, and state the resulting limits on symmetric infrared phases.
On this page
The question for today
What you’ll understand
- Distinguish a background anomaly of global symmetry from inconsistent dynamical gauge redundancy.
- Check the signs and form degrees in Maxwell electric–magnetic inflow.
- Use anomaly matching as an obstruction rather than a unique phase prediction.
An anomaly is an obstruction to realizing all desired background gauge redundancies simultaneously in an isolated theory with its declared symmetries. It can be present even when the global symmetries are exact at zero background. In four-dimensional source-free compact Maxwell theory, the electric and magnetic one-form symmetries have a mixed anomaly. Tracking its variation shows how a five-dimensional inflow action cancels a boundary phase. Matching this anomaly constrains the infrared theory, but does not select a unique spectrum, topological order, or phase diagram.
A counterterm can move an anomaly between symmetries
Use oriented D equal to four, no theta term, unit electric and magnetic probe normalization, and path-integral phases exp(iS). Let Be and Bm be compact two-form backgrounds. The electric background transformation shifts a by a one-form lambda e and Be by its derivative. The invariant field strength is F minus Be. Add a magnetic source term using Bm wedged with that invariant difference, while retaining the corresponding invariant Maxwell kinetic term.
This scheme preserves the electric background redundancy, but a magnetic shift of Bm can produce a variation when Be has curvature. Adding the local term Bm wedged with Be trades this magnetic variation for an electric one, rather than removing the mixed obstruction. The anomaly is the nontrivial class after allowed local counterterms have been considered. A variation of one particular formula alone is not enough to establish a physical anomaly without that counterterm analysis.
The boundary variation is canceled one dimension higher
Under a magnetic background transformation, integrate the derivative of its one-form parameter by parts on a closed four-manifold. The Bianchi identity removes the curvature of F in the source-free sector, leaving a term proportional to lambda m wedged with the electric background curvature. Its minus sign follows from the degree of the one-form parameter. This is a background variation of a global-symmetry probe, not a claim that the original dynamical U(1) gauge symmetry has become inconsistent.
Now regard the four-manifold as the oriented boundary of a five-manifold and add the displayed bulk term. Stokes' theorem gives a boundary variation with the opposite sign. The combined bulk–boundary phase is invariant for the local variation. The extra dimension is part of anomaly inflow, not an additional measured spatial dimension of the original system. Large transformations and nontrivial bundles require the globally quantized inflow definition rather than unrestricted multiplication of globally nonexistent potential forms.
Matching excludes a trivial outcome without fixing the answer
A nontrivial anomaly of an exact symmetry is preserved under renormalization-group flow when the symmetry and its background interpretation are maintained. A unique symmetry-preserving gapped infrared state with no intrinsic topological order cannot supply the required nontrivial response by itself. Gapless degrees of freedom, symmetry breaking, or an appropriate nontrivial topological realization can instead participate in matching, subject to the detailed anomaly and dimensional constraints.
These are allowed routes, not promises that every proposed topological theory or broken phase matches the anomaly. The gapless Maxwell photon is one consistent realization of the electric–magnetic anomaly. Noninvertible symmetries can impose related obstructions, but their full junction and category data may be needed. Conversely, canceling one anomaly by stacking an opposite sector removes only that obstruction; it does not prove a symmetric trivial phase exists or eliminate other anomalies and dynamical constraints.
Work it through
Cancel an explicit background phase by inflow
For smooth background data in the declared local representative, suppose ∫M₄ lambda m ∧ dBe = 2π². Find the four-dimensional variation, its path-integral phase, and the bulk inflow variation and phase.
The boundary variation is , giving phase −1.
The oriented five-dimensional variation is , also giving phase −1.
The action variations sum to zero and the phases multiply to one. Their cancellation relies on the matched orientation, coefficient, and common background transformation.
The isolated boundary phase is −1, canceled by the opposite bulk action variation; the combined phase is one.
The living notebook
Explore the model
Interactive lab · waves
Matched contributions can cancel while each remains present
Reset and set BOTH Phase difference to 0 degrees and Amplitude ratio to 1; predict reinforcement. Then Reset and set BOTH Phase difference to 180 degrees and Amplitude ratio to 1; predict exact cancellation of the sum. This equal-frequency sine-wave illustration has no background fields, anomalous measure, bulk boundary, gauge redundancy, or renormalization flow. It illustrates cancellation only and does not compute products of path-integral phases.
- Sum at x = 0
- 0.866
- Resultant amplitude
- 1.732
- Phase difference
- 60°
Two waves share the same frequency and wave number. Playback advances t in sin(2x − t) + a sin(2x − t + phase). Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.
Look forCancellation of a total response requires the correct paired contributions; it does not mean each contribution vanishes independently.
Pause & reason
Check your understanding
Your turn
Try it yourself
Stack two independent boundary sectors coupled to the same backgrounds, with mixed-anomaly coefficients +1 and −1 in the displayed convention. For ∫lambda m ∧ dBe = π², find their separate and total variations. Compare with stacking two +1 sectors, and state the limit of the inference.
Take a moment to make a prediction or work it out on paper. Open a hint when you need one.
Give me a hint
Anomaly variations add in the action. The compact U(1) coefficient here is an integer level, not a Z₂ label.
Compare with the solution
The +1 sector varies by −π/2 and the −1 sector by +π/2, so the total variation is zero for this anomaly. Two +1 sectors instead vary by −π, with phase −1 for the stated profile. Opposite coefficients cancel this mixed anomaly; the integer coefficient two is not automatically trivial. Cancellation alone does not construct a symmetric trivial gapped phase or rule out other obstructions.
Keep the idea
Take it with you
- A global-symmetry anomaly concerns simultaneous background coupling, modulo allowed local counterterms.
- Inflow cancellation requires matched form degrees, compact normalization, signs, and boundary orientation.
- Anomaly matching constrains possible infrared phases without determining a unique dynamical outcome.
Read further
Original sources and deeper explanations to follow at your own pace.
Optional progress, saved in this browser when available. Revisit any lesson at any time.
Lesson not marked complete.