Linking Number of Wilson Loops

• Linking Number of Wilson Loops is a topological invariant that quantifies the mutual winding of two closed loops in three dimensions.
• It unifies gauge theories by linking Wilson loop correlators with phenomena in Chern–Simons theory, anyonic statistics, and quantized Hall conductivity.
• Its extensions to holomorphic and noncommutative settings offer deeper insights into modern amplitude theory and generalized knot invariants.

The linking number of Wilson loops is a topological invariant central to gauge theory, condensed matter physics, and topological quantum field theory. It characterizes the mutual winding of two closed loops in three dimensions and serves as the fundamental quantity governing the gauge-invariant correlation of Wilson loops in both commutative and noncommutative contexts. The dependence of Wilson loop correlation functions on this integer-valued invariant unifies topological structure with physical phenomena such as anyonic exchange phases and quantized Hall conductivity.

1. Mathematical Definition of the Linking Number

Given two disjoint, oriented, smooth closed curves $C_1, C_2 \subset \mathbb{R}^3$, the Gauss linking number $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$ is defined by the double line integral

$$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$

where $\varepsilon_{\mu\nu\rho}$ is the Levi-Civita symbol and repeated indices are summed over spatial coordinates. This invariant counts, with orientation, the number of times $C_1$ winds around $C_2$. By its construction, $\mathrm{Lk}(C_1, C_2)$ is homotopy invariant, depending only on the topological class of the link. In higher-order or more abstract settings, analogues exist (e.g., holomorphic linking in complex geometry or generalizations in noncommutative geometry) (Obikhod et al., 1 Feb 2026, Bullimore et al., 2011, García-Compeán et al., 2015).

2. Wilson Loops, Topology, and Chern–Simons Theory

In quantum field theory, Wilson loops are nonlocal gauge-invariant observables defined by the holonomy of the gauge field around a closed contour: $W[C] = \exp\left(i q \oint_C A_\mu\,dx^\mu\right)$ for a $U(1)$ gauge field $A$ and charge $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$0. In Chern–Simons theory at integer level $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$1, the expectation value of the correlator of two such Wilson loops $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$2 and $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$3 is

$\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$4

This result follows from the quadratic nature of the Abelian Chern–Simons action, which ensures the path integral is Gaussian and the only two-loop configuration data is encoded in the linking number. In non-Abelian theory, similar structures arise with additional representations and trace structures. The result demonstrates the direct dependence of quantum observables on topological invariants and illustrates how the linking number emerges from the gauge field’s propagator via Wick contractions (Obikhod et al., 1 Feb 2026).

3. Physical Consequences: Anyonic Statistics and Hall Conductivity

The linking number of Wilson loops plays a pivotal role in establishing the topological origin of quantized transport and quasiparticle statistics:

• Fractional Quantum Hall Effect: In the effective Abelian Chern–Simons theory for Laughlin states at filling fraction $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$5, the level $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$6 determines both the quantized Hall conductivity $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$7 and the braiding phase $\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$8 for quasiparticles. The latter is given by

$\mathrm{Lk}(C_1, C_2) \in \mathbb{Z}$9

with $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$0 corresponding to an exchange resulting in anyonic statistics (Obikhod et al., 1 Feb 2026).

• Hall Conductance: Integrating out the Chern–Simons field coupled to an external electromagnetic potential yields an effective action proportional to $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$1, identifying $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$2 through the same topological data that set the Wilson loop correlators.

This unification evidences that the linking number underlies both fractional charge/statistics and macroscopic transport quantization.

4. Holomorphic and Noncommutative Generalizations

Holomorphic analogues of Wilson loops and linking numbers appear in complexified gauge theories, such as holomorphic Chern–Simons theory on twistor space. For two disjoint complex curves $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$3, the holomorphic linking invariant is given by (Bullimore et al., 2011): $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$4 where $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$5 are holomorphic 1-forms. The expectation values of holomorphic Wilson loops encode these invariants, generalizing the real case. This construction directly connects to modern amplitude theory: planar $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$6 super Yang–Mills amplitudes can be interpreted in terms of holomorphic linking by virtue of skein/BCFW recursion relations (Bullimore et al., 2011).

Noncommutative generalizations arise via the Seiberg–Witten map. Here, the noncommutative gauge field $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$7 is expanded in powers of the noncommutativity parameter $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$8, and at each order, new Poincaré-dual 1-cycles emerge. The noncommutative linking number for two loops $$\mathrm{Lk}(C_1,C_2) =\frac{1}{4\pi} \oint_{C_1}\oint_{C_2} \varepsilon_{\mu\nu\rho} \frac{(x^\rho - y^\rho)}{\lvert x-y \rvert^3} dx^\mu\,dy^\nu$$9 is

$\varepsilon_{\mu\nu\rho}$0

where $\varepsilon_{\mu\nu\rho}$1 are trivial cycles dual to higher-order gauge potentials. Each term at order $\varepsilon_{\mu\nu\rho}$2 introduces $\varepsilon_{\mu\nu\rho}$3 such cycles, resulting in a rich hierarchy of linking numbers in the noncommutative field. This structure has physical relevance for phenomena such as $\varepsilon_{\mu\nu\rho}$4-dependent corrections to the Aharonov–Bohm phase and shifts in noncommutative Landau level spectra (García-Compeán et al., 2015).

5. Wilson Loop Linking on Homogeneous Spaces and AdS/CFT

Supersymmetric Wilson loops with contours along Hopf fibers of $\varepsilon_{\mu\nu\rho}$5 represent a geometrically rich setting for linking. Two such fibers always have linking number one. In the context of $\varepsilon_{\mu\nu\rho}$6 SYM, correlators of these loops at both weak and strong coupling reflect the topological linking:

• At weak coupling, the perturbative expansion yields diagrammatic contributions governed by the linking number and explicit propagator integrals.
• At strong coupling, the AdS/CFT correspondence associates the correlator to the area of a classical string worldsheet in AdS$\varepsilon_{\mu\nu\rho}$7 interpolating between the two linked boundary fibers. The renormalized minimal area, and thereby the large-$\varepsilon_{\mu\nu\rho}$8 asymptotics of the correlator, depends smoothly on the base separation but the linking number remains a topological invariant throughout (Griguolo et al., 2012).

Generalizations to multiple fibers or higher winding numbers show the scaling of the effective area—and hence the correlator’s phase—directly with the product of their winding numbers, encapsulating the linking structure at both the perturbative and string-theoretic levels.

6. Role in Knot Invariants, Loop Equations, and Higher-Order Structures

Wilson loop correlators in Chern–Simons theory are the generating functionals for topological knot invariants such as the Jones polynomial. In both commutative and noncommutative settings, expansions of the gauge field or path integral generate linear combinations of linking numbers of the original loop with hierarchies of derived cycles. This construction generalizes to non-Abelian theory and holomorphic settings via Makeenko–Migdal-type and skein/BCFW recursion relations, unifying the computation of amplitudes, loop equations, and topological invariants (Bullimore et al., 2011, García-Compeán et al., 2015).

In noncommutative gauge theories, the path integral expansion yields noncommutative “Jones–Witten” invariants, with each higher-order term corresponding to an increasing number $\varepsilon_{\mu\nu\rho}$9 of auxiliary linking structures, providing a combinatorial and topological backbone for the invariants of noncommutative knots and links.

7. Unified Perspective

The linking number of Wilson loops serves as a foundational topological invariant with broad implications:

• In Abelian and non-Abelian Chern–Simons theories, it encapsulates the full topological content of the two-loop correlation function.
• In condensed matter, it underpins both the quantization of Hall conductance and the statistics of fractionalized excitations.
• In generalized settings—holomorphic, noncommutative, higher-genus, and higher-dimensional—it provides a bridge between topological field theory, knot theory, gauge dynamics, and quantum geometry.
• Its combinatorial proliferation in noncommutative settings hints at a further unification of topological and physical structures in advanced quantum gauge theories (Obikhod et al., 1 Feb 2026, Bullimore et al., 2011, Griguolo et al., 2012, García-Compeán et al., 2015).