Gauge Fixing, BRST Constraints, and the Gribov Problem

Gauge-fixed functional equations determine gauge-dependent Green functions subject to BRST or Slavnov–Taylor relations and a choice of gauge-fixing prescription. Perturbative Faddeev–Popov gauge fixing is local in configuration space; it does not remove Gribov copies globally. Consequently an infrared propagator branch, a positivity-violating gauge-fixed two-point function, and a gauge-invariant confinement observable are different statements.

Required background. BRST cohomology and physical observables supplies the physical-state distinction, Gribov copies and limits of local gauge fixing supplies the global obstruction, and closure, symmetry, and branch selection supplies the finite functional system. Helpful background. Non-equivalent confinement definitions prevents a gauge-fixed diagnostic from being promoted to a gauge-invariant definition.

Local BRST and Slavnov–Taylor control

In covariant gauge, the gauge-fixed Yang–Mills action contains

$$S_{\mathrm{gf+gh}} =\int\mathrm d^d x\, \left[ \frac{1}{2\xi}(\partial_\mu A_\mu^a)^2 +\bar c^a \left( -\partial_\mu D_\mu^{ab} \right)c^b \right].$$

The Faddeev–Popov operator is

$$\mathcal M^{ab} =-\partial_\mu D_\mu^{ab}.$$

Locally, BRST invariance packages gauge, ghost, and auxiliary-field variations into a nilpotent differential. Its renormalized construction and physical-state implications are developed in Becchi, Rouet, and Stora 1976, pp. 294–304. The quantum effective action obeys a Slavnov–Taylor functional identity

$$\mathcal S(\Gamma)=0$$

when the regulator, measure, counterterms, and boundary conditions preserve the construction or when the associated breaking terms have been restored.

A functional truncation rarely retains the complete identity. It should therefore report a dimensionless residual,

$$r_{\mathrm{ST}}(p_i) =\frac{ \lVert\mathcal S_i(\Gamma_{\mathrm{trunc}})\rVert }{ \epsilon_i+ \lVert\mathcal S_i\rVert_{\mathrm{reference}} },$$

for specified tensor projections and momenta. Transversality of the gluon self-energy,

$$p_\mu\Pi_{\mu\nu}(p)=0,$$

is necessary but does not test the full ghost and vertex content of the Slavnov–Taylor identities.

Landau-gauge propagator–vertex test

At $\xi=0$, write the Euclidean gluon and ghost propagators as

$$D_{\mu\nu}^{ab}(p) =\delta^{ab} \left( \delta_{\mu\nu} -\frac{p_\mu p_\nu}{p^2} \right) \frac{Z(p^2)}{p^2},$$

and

$$D_{\mathrm{gh}}^{ab}(p) =-\delta^{ab}\frac{F(p^2)}{p^2}.$$

A coupled closure for $Z$, $F$, and the ghost–gluon and three-gluon vertices must use one renormalization scheme and compatible tensor projections. In Landau gauge, Taylor’s kinematic theorem constrains the ghost–gluon vertex at vanishing incoming ghost momentum and, under its regularity assumptions, yields a finite vertex renormalization with $\widetilde Z_1=1$ in the associated scheme Taylor 1971, pp. 436–444.

This does not make the vertex bare at arbitrary momenta. A closure that replaces the full vertex by its Taylor-limit value must vary the omitted transverse and non-Taylor momentum structures and test their propagation into $Z$ and $F$.

Renormalization conditions such as

$$Z(\mu^2)=1, \qquad F(\mu^2)=1$$

select normalization, not the infrared branch. Scaling and decoupling-type solutions can arise from different infrared boundary conditions or gauge prescriptions. Solver attraction to one of them is not proof that it is the unique gauge-fixed realization.

Gribov copies and global gauge fixing

The condition

$$\partial_\mu A_\mu=0$$

can intersect one gauge orbit more than once. Infinitesimal copies occur when $\mathcal M$ has a zero mode. Restricting to the first Gribov region,

$$\Omega =\left\{ A:\partial_\mu A_\mu=0, \mathcal M[A]\ge0 \right\},$$

removes configurations beyond the first horizon but does not generally select one representative per orbit. The fundamental modular region is more restrictive and remains globally nontrivial. Gribov identified the local failure Gribov 1978, §§ 2–4; Singer proved the corresponding global topological obstruction under broad conditions Singer 1978, pp. 7–12.

The standard nilpotent BRST construction is tied to the unrestricted local Faddeev–Popov formulation. A restriction to a Gribov region or a nonperturbative copy-weighting prescription can modify BRST realization. One must state which functional measure is being solved before imposing a Slavnov–Taylor identity derived from another measure.

Gauge-fixed versus gauge-invariant claims

$Z(p^2)$, $F(p^2)$, a gauge-fixed running coupling, and their spectral representations depend on the gauge and copy prescription. They can:

• test ultraviolet perturbation theory;
• compare continuum and lattice calculations in the same gauge prescription;
• enter consistently constructed gauge-invariant amplitudes;
• constrain a proposed infrared closure.

They do not alone define confinement. Reflection-positivity violation of a transverse gluon propagator says that this gauge-fixed field does not create a positive-metric asymptotic particle in that reconstruction. Confinement statements require gauge-invariant spectra, line operators, static-source behavior, or another declared gauge-invariant diagnostic with matter and global form specified.

The functional-equation closure and validation map places the gauge prescription before closure. The functional-method validation comparison requires Slavnov–Taylor residuals, branch and copy sensitivity, a same-gauge regulator comparison, and a gauge-invariant external observable.

Common pitfalls

Treating Landau gauge as a unique global representative. The differential condition still admits Gribov copies.

Using transversality as the full Slavnov–Taylor test. Vertex and ghost identities contain information not visible in the two-point transverse projector.

Calling positivity violation a confinement proof. The violated positivity belongs to a gauge-fixed field. The physical claim must be phrased in gauge-invariant observables.

Exercises

1. Show that a zero mode of $\mathcal M[A]$ generates an infinitesimal copy of a Landau-gauge field.

Solution

For an infinitesimal gauge transformation $\delta A_\mu^a=D_\mu^{ab}\omega^b$,

$$\delta(\partial_\mu A_\mu^a) =\partial_\mu D_\mu^{ab}\omega^b =-\mathcal M^{ab}\omega^b.$$

If $\mathcal M\omega=0$, both $A$ and $A+D\omega$ satisfy Landau gauge to first order. They lie on the same gauge orbit, so the condition is not locally unique at the horizon.

2. A gluon propagator violates reflection positivity but a gauge-invariant bound-state correlator has a positive spectral pole. Are the results inconsistent?

Solution

No. The gauge-fixed gluon field acts in an indefinite or constrained state space and need not have a positive spectral measure. The gauge-invariant composite belongs to the physical observable algebra and can have a positive pole. The two statements concern different operators.

Continue

Functional-Method Validation and Error Control supplies the separate symmetry, branch, truncation, parameter, continuation, and numerical errors. Use Non-Equivalent Confinement Definitions before attaching a gauge-invariant interpretation.

References

• Becchi, Carlo, Alain Rouet, and Raymond Stora. “Renormalization of Gauge Theories.” Annals of Physics 98 (1976): 287–321. DOI.
• Gribov, V. N. “Quantization of Non-Abelian Gauge Theories.” Nuclear Physics B 139 (1978): 1–19. DOI.
• Singer, I. M. “Some Remarks on the Gribov Ambiguity.” Communications in Mathematical Physics 60 (1978): 7–12. DOI.
• Taylor, John C. “Ward Identities and Charge Renormalization of the Yang–Mills Field.” Nuclear Physics B 33 (1971): 436–444. DOI.