Gribov's No-Pole Condition in Gauge Theories

• Gribov's no-pole condition is a nonperturbative constraint that requires the ghost dressing function to remain free of poles, enforcing 1-σ(k) > 0 for all finite momenta.
• It governs the infrared behavior of ghost and gluon propagators, linking scaling and decoupling solutions and showing dimensional dependence in Yang–Mills theories.
• Implementations in the Gribov–Zwanziger and Coulomb gauge frameworks illustrate its role in controlling gauge-fixing ambiguities, center vortex dominance, and real-time reformulations.

Searching arXiv for recent and foundational papers on Gribov’s no-pole condition, including Euclidean, Coulomb-gauge, and Lorentzian formulations. arxiv_search({"query":"Gribov no-pole condition ghost propagator Landau gauge", "max_results": 10, "sort_by": "relevance"}) arxiv_search(query="Gribov no-pole condition ghost propagator Landau gauge", max_results=10, sort_by="relevance") Searching arXiv for "Gribov no-pole condition" and closely related formulations. Gribov’s no-pole condition is a nonperturbative constraint on the ghost sector of gauge-fixed Yang–Mills theory. In its standard Euclidean Landau-gauge formulation, one writes the ghost propagator as

$G(k^2)=\frac{1}{k^2\,[1-\sigma(k^2)]},$

or equivalently $F(k^2)=1/[1-\sigma(k^2)]$ for the ghost dressing function, and requires that $1-\sigma(k^2)>0$ for all finite $k^2$. This implements the restriction to the first Gribov region $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$, where the Faddeev–Popov operator is positive, and excludes poles of the ghost two-point function away from the origin. Saturation at $k=0$, $1-\sigma(0)=0$, is the horizon condition associated with the Gribov horizon and with the original ghost-enhanced scenario (Cucchieri et al., 2012, Capri et al., 2012).

1. Geometric origin and operator-theoretic meaning

In Landau gauge, the gauge condition is

$\partial_\mu A_\mu^a=0,$

and the Faddeev–Popov operator is

$$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$

Because $M$ is Hermitian in Landau gauge, its spectrum is real, and the first Gribov region is the set of transverse configurations for which all eigenvalues are positive. Its boundary $F(k^2)=1/[1-\sigma(k^2)]$0 is the locus where the lowest eigenvalue vanishes (Cooper et al., 2016, Sainapha, 2019).

The ghost propagator is the inverse of the Faddeev–Popov operator. In a fixed background it admits a spectral representation in terms of Faddeev–Popov eigenmodes, so the positivity of $F(k^2)=1/[1-\sigma(k^2)]$1 is directly tied to the absence of singularities in the ghost sector. In momentum space one introduces the ghost form factor $F(k^2)=1/[1-\sigma(k^2)]$2 through

$F(k^2)=1/[1-\sigma(k^2)]$3

Gribov’s no-pole condition is then

$F(k^2)=1/[1-\sigma(k^2)]$4

while the boundary case

$F(k^2)=1/[1-\sigma(k^2)]$5

corresponds to horizon saturation (Cucchieri et al., 2012, Cooper et al., 2016).

This formulation is necessary for remaining inside $F(k^2)=1/[1-\sigma(k^2)]$6, but it is not sufficient for a unique gauge fixing. Explicit constructions of legitimate topologically trivial Gribov copies inside the first Gribov region in Euclidean $F(k^2)=1/[1-\sigma(k^2)]$7 $F(k^2)=1/[1-\sigma(k^2)]$8 Landau gauge show that positivity of the Faddeev–Popov operator does not eliminate all copies. A precise implication is that restriction to $F(k^2)=1/[1-\sigma(k^2)]$9 solves the wrong-sign determinant problem and controls infinitesimal copies, but the fundamental modular region is smaller than $1-\sigma(k^2)>0$0 and is needed for a fully unique gauge fixing (Landim et al., 2014).

2. Ghost form factor, horizon function, and the Gribov–Zwanziger implementation

Zwanziger’s reformulation replaces the no-pole statement by a condition on the nonlocal horizon function

$1-\sigma(k^2)>0$1

In standard conventions, the Gribov–Zwanziger action is

$1-\sigma(k^2)>0$2

with the Gribov parameter fixed by the gap equation or horizon condition

$1-\sigma(k^2)>0$3

The all-order relation between the two formulations is exact: the zero-momentum ghost form factor satisfies

$1-\sigma(k^2)>0$4

so the no-pole inequality is equivalent to the horizon inequality, and the Gribov–Zwanziger gap equation is the dynamical implementation of Gribov’s requirement (Capri et al., 2012, Cooper et al., 2016).

This equivalence clarifies an issue that was historically treated only at low orders. The exact result is that the physically relevant statement is formulated in terms of the $1-\sigma(k^2)>0$5PI ghost self-energy:

$1-\sigma(k^2)>0$6

and the no-pole condition becomes $1-\sigma(k^2)>0$7. In the infinite-volume limit, the Gribov–Zwanziger horizon condition saturates this bound (Capri et al., 2012).

In the original Gribov–Zwanziger framework, horizon saturation implies an infrared-enhanced ghost and an infrared-suppressed gluon. In refined Gribov–Zwanziger theory, dimension-two condensates modify the infrared sector, allowing a finite ghost and a gluon propagator that is suppressed but finite at zero momentum. This distinction is central to the modern contrast between scaling and decoupling solutions (Capri et al., 2012, Cooper et al., 2016).

3. Landau-gauge infrared structure and dimensional dependence

The clearest modern statement of the no-pole condition’s infrared content is dimension dependent. In Landau gauge one writes the transverse gluon propagator as

$1-\sigma(k^2)>0$8

and the ghost propagator as

$1-\sigma(k^2)>0$9

The one-loop ghost self-energy with a bare ghost–gluon vertex is

$k^2$0

In $k^2$1, exact manipulations of this expression isolate the infrared behavior

$k^2$2

so the no-pole condition can be satisfied only if

$k^2$3

Equivalently, any $k^2$4 decoupling or massive gluon solution with $k^2$5 produces a logarithmic infrared singularity in $k^2$6 and violates $k^2$7 at sufficiently small momentum (Cucchieri et al., 2012, Dudal et al., 2012).

The same conclusion survives beyond the one-loop ghost line. In the full ghost Dyson–Schwinger equation, both the case $k^2$8 and the scaling-like case $k^2$9 were analyzed in $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$0. Under the stated assumptions of infrared-finite ghost–gluon vertices and standard regularity conditions, the logarithmic singularity remains unavoidable if $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$1. The result is therefore not an artifact of the tree-level vertex approximation (Dudal et al., 2012).

By contrast, in $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$2 and $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$3 the same angular analysis and Dyson–Schwinger bounds are infrared finite when $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$4 is finite. The no-pole condition then does not force $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$5, and decoupling or massive gluon propagators with $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$6 are compatible with $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$7 for all finite $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$8. This is the precise sense in which the infrared obstruction present in $\Omega=\{A:\partial\!\cdot\!A=0,\;M[A]>0\}$9 is absent in higher dimensions (Cucchieri et al., 2012, Dudal et al., 2012).

The same work also evaluated $k=0$0 explicitly at one loop using lattice-fitted $k=0$1 gluon propagators in $k=0$2. Those fits exhibit a one-parameter family of ghost behaviors labeled by $k=0$3, with a critical coupling defined by $k=0$4. For the fits used, $k=0$5 in $k=0$6, $k=0$7 in $k=0$8, and $k=0$9 in $1-\sigma(0)=0$0; $1-\sigma(0)=0$1 gives a free-like infrared ghost, while $1-\sigma(0)=0$2 gives infrared enhancement (Cucchieri et al., 2012).

4. Coulomb gauge, horizon saturation, and confinement observables

In Coulomb gauge the relevant Faddeev–Popov operator is spatial,

$1-\sigma(0)=0$3

and the no-pole condition is expressed in terms of the equal-time ghost form factor. Depending on convention one writes

$1-\sigma(0)=0$4

but the infrared condition is the same:

$1-\sigma(0)=0$5

This is the Coulomb-gauge horizon condition, the direct analogue of $1-\sigma(0)=0$6 in Landau gauge (Reinhardt et al., 2016, Burgio et al., 2013).

Within the Hamiltonian variational approach, the ghost loop enters the gluon gap equation through the curvature of the Faddeev–Popov determinant, and the self-consistent solution yields a Gribov-type quasi-gluon energy

$1-\sigma(0)=0$7

so that the equal-time gluon propagator

$1-\sigma(0)=0$8

vanishes as $1-\sigma(0)=0$9. The instantaneous color-Coulomb potential,

$\partial_\mu A_\mu^a=0,$0

inherits its infrared strength from the ghost sector, and its string tension $\partial_\mu A_\mu^a=0,$1 obeys Zwanziger’s inequality $\partial_\mu A_\mu^a=0,$2 (Reinhardt et al., 2016).

Lattice work in Coulomb gauge supports this structure. In the Hamiltonian limit, the Coulomb-gauge ghost dressing behaves as $\partial_\mu A_\mu^a=0,$3 with $\partial_\mu A_\mu^a=0,$4, and the infrared plateau of $\partial_\mu A_\mu^a=0,$5 gives $\partial_\mu A_\mu^a=0,$6 (1311.03908). In the confining wave-functional ensemble used to relate $\partial_\mu A_\mu^a=0,$7 Coulomb gauge to $\partial_\mu A_\mu^a=0,$8 Landau gauge, the ghost dressing shows infrared enhancement

$\partial_\mu A_\mu^a=0,$9

while removing center vortices makes the ghost nearly flat in the infrared, with $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$0, and destroys the horizon condition (Quandt et al., 2010).

This center-vortex connection is one of the most developed physical interpretations of the no-pole condition. Center vortices generate near-zero Faddeev–Popov modes, dominate the infrared behavior of the ghost form factor, and induce both the horizon condition and the confining Coulomb potential. A plausible implication is that, in Coulomb gauge, the no-pole condition is not merely a constraint on a two-point function but a diagnostic of which gauge-field configurations dominate the infrared functional measure (Quandt et al., 2010, Reinhardt et al., 2016).

A complementary spectral analysis of the Faddeev–Popov operator near the first Gribov horizon exhibits two scenarios. In “Type I,” the tuned horizon spectrum behaves as $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$1 and produces an infrared-enhanced ghost dressing; in “Type II,” the spectrum remains $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$2 near $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$3 and the first zero occurs at nonzero momentum, implying a finite ghost dressing at the origin. In the examples studied, Coulomb gauge in $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$4 displayed Type I behavior, whereas Landau gauge in $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$5 displayed Type II behavior (Greensite, 2010).

5. Background fields, exceptional theories, and the limits of the standard picture

The status of the no-pole condition changes substantially when the gauge background, matter content, or kinematics are altered. In the Landau–DeWitt gauge with a constant temporal $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$6 background,

$$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$7

background-covariant momenta are shifted by $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$8, and the one-loop horizon condition becomes a background-dependent gap equation. The explicit result is that the larger the background field, the smaller the Gribov mass parameter. This suggests that the relevance of Gribov copies decreases as the size of the background field increases (Canfora et al., 2016).

There are also theories in which the standard Gribov restriction becomes dynamically irrelevant. In Euclidean $$M^{ab}=-\partial_\mu D_\mu^{ab} = -\big(\partial^2\delta^{ab}-g f^{acb}A_\mu^c\partial_\mu\big).$$9 super Yang–Mills in Landau gauge, the vanishing beta function implies the absence of an RG-invariant scale. The horizon condition then admits only the consistent solution

$M$0

so no nonperturbative Gribov mass can be generated and there is no need to restrict the functional integral to the Gribov region. This agrees with the absence of a confining phase inferred from the Coulombic Wilson loop behavior in AdS/CFT (Capri et al., 2014).

A closely related phenomenon occurs in four-dimensional topological Yang–Mills theory in the Baulieu–Singer anti-self-dual Landau gauges. There, Gribov copies are present, but the associated gap equation has only the trivial solution $M$1. The paper attributes this to the tree-level exactness of the model in that gauge choice, the absence of radiative corrections in both gauge and ghost sectors, and the lack of an RG-invariant mass scale (Dudal et al., 2019).

The no-pole logic can also be extended beyond ordinary non-Abelian Yang–Mills. In noncommutative QED, the Landau-gauge Faddeev–Popov operator develops Gribov ambiguities because the Moyal star commutator mimics non-Abelian structure. A positive Faddeev–Popov operator and a no-pole condition can be formulated, implemented in the path integral, and used to compute the photon propagator. In the approximation developed there, the propagator acquires only a $M$2-dependent rescaling, and the commutative limit reproduces standard QED (Holanda et al., 2021).

These cases delimit the scope of the standard Euclidean Yang–Mills picture. They show that the no-pole condition is neither a universal confinement criterion nor a uniformly dynamical mechanism across gauge theories. Its consequences depend sharply on whether the theory admits a nonperturbative mass scale and on how the ghost sector couples to the background.

6. Lorentzian reformulation and current open problems

For most of its history, the no-pole condition was formulated in Euclidean space, where the Faddeev–Popov operator is elliptic and positivity defines a spectral boundary. In Minkowski spacetime this reasoning fails, because the Faddeev–Popov operator becomes hyperbolic:

$M$3

The Lorentzian replacement proposed recently is therefore not a positivity condition but a real-time boundary-value statement: a background remains inside the first Lorentzian Gribov region if the Faddeev–Popov wave equation admits no nontrivial source-free solution obeying Feynman boundary conditions (Guimaraes, 7 Jun 2026).

The resulting definitions are

$M$4

For backgrounds localized in time, this criterion becomes the injectivity of the negative-frequency block $M$5 of the classical ghost scattering map. For stationary backgrounds, it becomes a spatial bound-state or threshold-resonance problem after Fourier transformation in time (Guimaraes, 7 Jun 2026).

The Lorentzian analysis reveals a structural difference with no Euclidean analogue. A Wronskian identity shows that pure frequency mixing in stable self-adjoint time-dependent channels cannot by itself produce the obstruction, because the corresponding Bogoliubov coefficient $M$6 cannot develop a kernel. Nevertheless, static chromoelectric potentials can reach the horizon at finite, nonvanishing frequency, since $M$7 couples directly to the ghost time derivative. Static chromomagnetic backgrounds instead reproduce the familiar zero-frequency horizon crossing (Guimaraes, 7 Jun 2026).

This reformulation also sharpens the status of real-time Gribov–Zwanziger theory. The exact Lorentzian restriction is a Fredholm determinant,

$M$8

where zeros coincide with the vanishing of the Feynman ghost determinant. The naive local Feynman continuation of the Euclidean Zwanziger horizon functional fails to reproduce this determinant structure, both in critical behavior and in detecting horizons in symmetric static backgrounds. The construction of a genuine local, renormalizable real-time Gribov–Zwanziger action therefore remains an open problem (Guimaraes, 7 Jun 2026).

In this broader perspective, Gribov’s no-pole condition is best understood as a gauge-fixed criterion controlling when the inverse Faddeev–Popov operator remains regular within a restricted functional domain. In Euclidean Landau and Coulomb gauges it organizes the relation between ghost enhancement, horizon saturation, and infrared confinement scenarios; in $M$9 Landau gauge it forbids $F(k^2)=1/[1-\sigma(k^2)]$00; in $F(k^2)=1/[1-\sigma(k^2)]$01 it is compatible with decoupling; in Coulomb gauge it links naturally to the color-Coulomb potential and center-vortex dominance; and in Lorentzian signature it becomes a real-time kernel condition rather than a positivity statement (Dudal et al., 2012, Reinhardt et al., 2016, Guimaraes, 7 Jun 2026).