Charges, Screening, and Long-Range Forces

Gauss law does two jobs at once: locally it is the gauge constraint, while after integration it relates charge to boundary flux. In a gapless Maxwell-like channel, an unscreened charge leaves radial flux at arbitrarily large distance and, in four spacetime dimensions, a Coulombic $1/r$ interaction. In a screened medium the microscopic Gauss law remains exact, but induced charge cancels the source so that the flux through a distant surface vanishes. Confinement is different again: flux can be collimated into a tube rather than exponentially dispersed.

Required background. Dynamical gauge fields and matter supplies the action and source convention; gauge orbits, Gauss constraints, and stabilizers explains why the local Gauss operator generates redundancy.

Helpful background. Gauge-invariant dressed observables supplies explicit charged dressings; vacua, states, and representations supplies the representation-theoretic meaning of sectors.

Gauss law from local constraint to boundary flux

Begin with the Abelian equation in the source convention $\nabla\!\cdot\!\mathbf E=J^0$. On a spatial region $V$ with boundary $S$,

$$Q_J(V) \equiv\int_V d^3x\,J^0 =\int_V d^3x\,\boldsymbol\nabla\!\cdot\!\mathbf E =\oint_S \mathbf E\!\cdot d\mathbf S.$$

This equality is exact. It says neither that the flux is long-ranged nor that $Q_J(V)$ equals the bare external charge: $J^0$ includes every dynamical contribution inside $V$. If a source polarizes the vacuum, the induced charge changes the right-hand side as the surface expands.

In canonical quantization the local version is a constraint on physical states,

$$\widehat{\mathcal G}(x)|\Psi_{\mathrm{phys}}\rangle =\left(\boldsymbol\nabla\!\cdot\!\widehat{\mathbf E}-\widehat J^0\right)(x) |\Psi_{\mathrm{phys}}\rangle=0.$$

The corresponding non-Abelian equation is $(D_iE^i)^a=J^{a0}$. It cannot be converted into a gauge-invariant “color vector” by simply integrating the adjoint index: field values at different points transform in different local frames. A boundary charge requires parallel transport, boundary conditions, and a specification of which transformations at the boundary act physically. This distinction between the local constraint and asymptotic charge is part of the canonical analysis in Weinberg 1996, §15.4, pp. 14–17.

Coulomb and Yukawa response derived

For a static point source $Q$ in vacuum Maxwell theory,

$$-\nabla^2 A_0(\mathbf x)=Q\delta^{(3)}(\mathbf x), \qquad A_0(r)=\frac{Q}{4\pi r}.$$

Hence

$$\mathbf E(r)=\frac{Q}{4\pi r^2}\widehat{\mathbf r}, \qquad \oint_{S_r}\mathbf E\!\cdot d\mathbf S=Q.$$

The flux survives at every radius, and two static probes have a $1/r$ potential. Equivalently, the momentum-space propagator has a pole at $\mathbf p^2=0$; Fourier transformation produces the power law.

Now suppose the gauge-invariant response contains a vector mass $m$. At distances where linear response is adequate, the static equation becomes

$$(-\nabla^2+m^2)A_0(\mathbf x)=Q\delta^{(3)}(\mathbf x),$$

with solution

$$A_0(r)=\frac{Qe^{-mr}}{4\pi r}, \qquad \mathbf E(r)= \frac{Qe^{-mr}}{4\pi r^2}(1+mr)\widehat{\mathbf r}.$$

The outward flux is therefore

$$\Phi_E(r)=Qe^{-mr}(1+mr)\longrightarrow0 \qquad (r\to\infty).$$

There is no violation of Gauss law: rewriting the massive equation as $\nabla\cdot\mathbf E=Q\delta^{(3)}-m^2A_0$ exhibits an induced charge density whose integral tends to $-Q$. The screening length is $m^{-1}$. A Proca equation is enough to derive the profile, but in a fundamental gauge theory the mass must arise consistently—for example through a Higgs regime—rather than by discarding gauge redundancy.

These two solutions give an operational test. A nearly massless pole produces power-law response and persistent flux; a nonzero screening mass produces exponential response and vanishing distant flux. The test should be applied to gauge-invariant correlators or to the energy of dressed external probes.

Screening classes and dressed charge

A bare field $\psi(x)$ changes under a gauge transformation, so the state $\psi(x)|0\rangle$ is not physical. A dressing supplies the flux required by Gauss law. In an Abelian theory it may be Coulombic, string-like, or terminate on a physical boundary. Distinct dressings can have different energies and radiation content even though they carry the same asymptotic charge.

With dynamical matter, an external representation is screenable if it can combine with dynamical excitations and gauge fields to form a neutral representation. For $SU(N)$:

• adjoint gauge bosons have zero $N$-ality, so they can change a representation without changing its center charge;
• adjoint matter therefore leaves $N$-ality as a useful electric screening class;
• dynamical fundamental matter carries one unit of $N$-ality and can eventually screen every external electric $N$-ality.

This statement depends on the global gauge group, not only on the Lie algebra. The quotient that defines the group changes the allowed electric and magnetic probes, while the dynamical representations determine which lines can end. The relation among global form, genuine lines, and screening is developed in Gaiotto et al. 2015, §4.2, pp. 17–19.

The resulting long-distance possibilities should not be collapsed into a single “charge/no charge” dichotomy.

Response of a dressed external pair	Flux and static energy	What can be concluded
Coulombic	Radial flux reaches infinity; $V(r)\sim c/r$	A gapless gauge response survives in this channel
Yukawa screened	Flux decays as $e^{-mr}$; $V(r)\sim ce^{-mr}/r$	The channel has a finite screening length
Flux tube without string breaking	Flux remains narrow; $V(r)\sim\sigma r$	The chosen unscreenable probe has string tension $\sigma$
String breaking	Linear rise crosses over to two screened bound states	The probe can be screened by dynamical matter; its asymptotic line law is not an area law
Finite-volume neutralization	No asymptotic sphere exists	Use boundary flux, finite-size scaling, and neutral-pair energies rather than an infinite-volume charge label

In a compact lattice gauge theory with dynamical matter, open strings can terminate on matter and a Wilson loop that shows an intermediate area-like regime can cross over to perimeter behavior at sufficiently large scale. Fradkin gives an explicit Hamiltonian account of this screening and phase structure in Fradkin 2013, §9.10, pp. 315–318.

When charge labels define sectors

An asymptotic charge can label a superselection sector when all local gauge-invariant observables preserve it. Gauss law explains why: a local operation in a bounded region cannot change flux measured at spatial infinity. This conclusion must be qualified in three ways.

1. Screening: dynamical fields can identify external charge labels that differ by a screenable representation.
2. Infrared structure: in a theory with massless gauge bosons, a charged state carries an inseparable soft cloud. Sharp one-particle mass eigenstates can be replaced by infraparticle behavior.
3. Boundaries: a boundary, defect, or external reservoir can absorb flux, so the algebra and allowed operations differ from those in empty infinite space.

Thus “the charge sector” is shorthand for a sector of a specified observable algebra with specified asymptotic conditions. The fully rigorous construction is deferred to Gauss-law infrasectors and asymptotic charge classes.

Independent checks and failure diagnoses

• Integrated constraint: compute both $\int_VJ^0$ and the boundary flux. A mismatch usually signals an omitted induced charge, boundary term, or sign convention.
• Pole/profile agreement: a pole at $\mathbf p^2=-m^2$ should give the same screening length $m^{-1}$ as the real-space exponential.
• Representation check: tensor the external representation with the available dynamical representations. If a singlet appears, an asymptotic unscreened line cannot be assumed.
• Volume scaling: repeat the diagnostic at increasing volume. A finite box can turn a power law into an apparent gap or force total charge to vanish.
• Energy test: distinguish a low-overlap operator from a true flux tube by extracting the lowest static energy with a basis that includes broken-string states.

Common pitfalls

Saying screening “breaks” Gauss law. Screening changes the total source inside an expanding surface. The local constraint continues to hold exactly.

Using a bare colored field as a state. A gauge-variant insertion does not define a physical isolated charge. Specify its dressing, endpoints, or boundary support.

Reading confinement from a short linear window. A potential can look linear before pair creation becomes favorable. The asymptotic diagnosis must include all dynamical screening channels.

Treating the Lie algebra as the complete theory. The groups $SU(N)$ and $SU(N)/\mathbb Z_N$ share a Lie algebra but admit different genuine line operators and topological sectors.

Continue with Coulomb, Higgs, and confining regimes for the full gauge-invariant phase classification, or go directly to line-operator diagnostics when the probe lattice is already known.

References

• Fradkin, Eduardo. Field Theories of Condensed Matter Physics. 2nd ed. Cambridge University Press, 2013, §9.10, pp. 315–318. DOI.
• Gaiotto, Davide, Anton Kapustin, Nathan Seiberg, and Brian Willett. “Generalized Global Symmetries.” Journal of High Energy Physics 2015, no. 2 (2015): 172, §4.2, pp. 17–19. DOI. Open PDF.
• Weinberg, Steven. The Quantum Theory of Fields. Vol. II: Modern Applications. Cambridge University Press, 1996, §15.4, pp. 14–17. DOI.