---
title: Higher Chern–Simons Action Extensions
url: https://www.emergentmind.com/topics/higher-chern-simons-action
type: topic
---

# Higher Chern–Simons Action Extensions

Higher Chern–Simons action denotes a family of Chern–Simons-type functionals that extend ordinary Chern–Simons theory beyond the standard three-dimensional Lie-algebra-valued connection. In the modern categorified formulation, the basic field is a higher connection valued in a Lie \(2\)-algebra or, more generally, an \(L_\infty\)-algebra, and the action is a transgression of a higher invariant polynomial such as \(\langle \mathcal F^n,\mathcal H\rangle\) [2605.29282]. In differential-cohomological and higher-stack formulations, the same structure is expressed as the holonomy of a universal differential characteristic morphism \(\mathbf L:\mathrm{Fields}\to \mathbf B^nU(1)_{\mathrm{conn}}\), so that the exponentiated action is obtained by fiber integration over spacetime [1207.5449] [1301.2580]. This suggests that the expression “higher Chern–Simons action” is not tied to a single formalism: in the literature it covers categorified higher gauge theory, higher-dimensional transgression theories, and, in some papers, algebraic generalizations that remain ordinary \(3\)-dimensional Chern–Simons actions but use higher-spin or \(3\)-algebraic gauge data [1511.01482] [1311.4589].

## 1. Ordinary Chern–Simons theory as the template

Ordinary Chern–Simons theory starts from a Lie algebra \(\mathfrak g\), a connection \(A\), curvature \(F=dA+\tfrac12[A,A]\), and an invariant polynomial such as \(\langle F^{n+1}\rangle\). The Chern–Simons form is a transgression form whose exterior derivative reproduces the characteristic class. This transgression viewpoint persists in essentially every higher formulation.

One direct categorified analogue replaces the ordinary curvature polynomial by
\[
\langle \mathcal F^n,\mathcal H\rangle_{\mathfrak v_0\mathfrak v_1},
\]
where \(\mathcal F\) is a \(2\)-form fake curvature and \(\mathcal H\) is a \(3\)-form higher curvature. In this setting, the characteristic class has degree \(2n+3\), while the higher Chern–Simons form has degree \(2n+2\); the paper on semistrict higher Chern–Simons theory states the comparison explicitly as: ordinary Lie algebra Chern–Simons has characteristic form in degree \(2n+2\) and Chern–Simons form in degree \(2n+1\), whereas semistrict \(2\)-Chern–Simons has characteristic form in degree \(2n+3\) and Chern–Simons form in degree \(2n+2\) [2605.29282].

A different, globally defined template arises in Deligne–Beilinson or differential cohomology. There the field is not a globally defined form but a differential cohomology class, and the action is the differential cup square. For abelian theories this produces a nontrivial action only in dimensions \(4l+3\),
\[
CS_k(w^{[2l+1]})=2i\pi k\int_M w^{[2l+1]} *_D w^{[2l+1]},
\]
with integral level \(k\) [1207.1270]. The higher-stack version packages the same idea into a universal map \(\mathbf{B}G_{\mathrm{conn}}\to \mathbf{B}^3U(1)_{\mathrm{conn}}\) or, more generally, \(\mathbf L:\mathbf{B}G_{\mathrm{conn}}\to \mathbf B^nU(1)_{\mathrm{conn}}\), whose transgression yields the action, prequantum bundle, and WZW object in lower codimension [1301.2580].

## 2. Differential cohomology and higher-stack formulations

In the differential-cohomological approach, the higher Chern–Simons action is defined globally on the full field space, including topologically nontrivial sectors. For higher abelian theories on a closed \((4k+3)\)-manifold \(\Sigma_{4k+3}\), a field is a differential cohomology class
\[
\hat a\in \hat H^{2k+2}(\Sigma_{4k+3};\mathbb Z),
\]
and the exponentiated action is
\[
\exp(iS(\hat a))=\exp\!\left(2\pi i\int_{\Sigma_{4k+3}} \hat a\cup_{\mathrm{conn}}\hat a\right).
\]
This formulation automatically includes instanton sectors and explains the integrality of both the level and the charges of generalized Wilson \((2l+1)\)-loops [1207.5449] [1207.1270].

The higher-stack formulation sharpens this further by replacing the moduli space of gauge-equivalence classes with the full higher smooth moduli stack of fields. For ordinary \(3\)-dimensional Chern–Simons theory, the universal differential characteristic morphism
\[
\hat{\mathbf c}:\mathbf{B}G_{\mathrm{conn}}\to \mathbf{B}^3U(1)_{\mathrm{conn}}
\]
is the extended Lagrangian. Transgression along a closed oriented \(k\)-manifold \(\Sigma_k\) gives
\[
[\Sigma_k,\mathbf{B}G_{\mathrm{conn}}]\to \mathbf{B}^{3-k}U(1)_{\mathrm{conn}},
\]
so that \(k=3\) yields the action functional, \(k=2\) the prequantum line bundle, and \(k=1\) the WZW gerbe [1301.2580]. The same paper treats this as the prototype of a general machine: any differential characteristic map \(\mathbf L:\mathrm{Fields}\to \mathbf B^nU(1)_{\mathrm{conn}}\) defines an \(n\)-dimensional higher Chern–Simons-type theory by holonomy.

This framework is especially important because it distinguishes the local Lagrangian density from the globally defined exponentiated action. It also makes clear why higher Chern–Simons theories are naturally extended prequantum field theories: transgression produces \(U(1)\)-\((n-k)\)-bundles with connection on moduli stacks in codimension \(k\) [1207.5449] [1301.2580].

## 3. Categorified gauge theory: Lie \(2\)-algebras and \(L_\infty\)-algebras

In categorified higher gauge theory, the algebraic input is commonly a semistrict Lie \(2\)-algebra
\[
\mathfrak v=(\mathfrak v_0,\mathfrak v_1,\alpha,[\cdot,\cdot],[\cdot,\cdot],[\cdot,\cdot,\cdot]),
\]
where the trilinear bracket \([\cdot,\cdot,\cdot]\) is the Jacobiator. A \(2\)-connection is a pair
\[
(A,B),\qquad A\in \Omega^1(M,\mathfrak v_0),\quad B\in \Omega^2(M,\mathfrak v_1),
\]
with curvatures
\[
\mathcal F=dA+\tfrac12[A,A]-\alpha(B),\qquad
\mathcal H=dB+[A,B]-\tfrac16[A,A,A].
\]
The higher Chern–Weil form
\[
\mathcal P_{2n+3}(\mathcal F_t,\mathcal H_t)=\langle \mathcal F_t^{\,n},\mathcal H_t\rangle_{\mathfrak v_0\mathfrak v_1}
\]
is closed, and Cartan homotopy produces the transgression form \(\mathcal Q_{2n+2}\). Setting one endpoint of the interpolation equal to zero defines the \((2n+2)\)-dimensional semistrict higher Chern–Simons form
\[
\mathcal{CS}_{2n+2}(A,B)=\int_0^1dt\Big(n\langle A,\mathcal F_t^{n-1};\mathcal H_t\rangle_{\mathfrak v_0\mathfrak v_1}+\langle \mathcal F_t^n;B\rangle_{\mathfrak v_0\mathfrak v_1}\Big),
\]
which satisfies
\[
d\mathcal{CS}_{2n+2}(A,B)=\langle \mathcal F^n,\mathcal H\rangle
\]
within the transgression setup [2605.29282].

In the \(4\)-dimensional semistrict theory of balanced Lie \(2\)-algebras, the action is written as
\[
\mathrm{CS}_2(\omega,\Omega_\omega)=\kappa_2 \int_N \left[ \frac12\big(2f+\partial \Omega_\omega,\Omega_\omega\big) -\frac1{24}\big(\omega,[\omega,\omega,\omega]\big) \right],
\]
with
\[
f=d\omega+\frac12[\omega,\omega]-\partial\Omega_\omega,\qquad
F_f=d\Omega_\omega+[\omega,\Omega_\omega]-\frac16[\omega,\omega,\omega].
\]
Its Euler–Lagrange equations are
\[
f=0,\qquad F_f=0,
\]
so classical solutions are flat higher connections [1406.2197].

The homotopy-algebraic formulation generalizes this further. For a cyclic \(L_\infty\)-algebra \(L\), the universal higher Chern–Simons action is
\[
S=\sum_{i\in\mathbb N}\frac{1}{(i+1)!}\langle \ell,\mu_i(\ell,\ldots,\ell)\rangle_L,
\]
and its equation of motion is the homotopy Maurer–Cartan equation
\[
\sum_{i\in\mathbb N}\frac{1}{i!}\mu_i(\ell,\ldots,\ell)=0.
\]
In a \(5\)-dimensional \(3\)-term case with field \(\hat a=A+B+C\), this yields higher curvatures \(F\), \(H\), and \(G\); in the holomorphic ambitwistor-space model, the same pattern gives a holomorphic higher Chern–Simons action on \(L^{5|6}\) whose classical equations are equivalent to those of maximally supersymmetric Yang–Mills theory on \(\mathbb C^4\) [1702.04160].

## 4. Gauge variation, transgression, and higher Wess–Zumino–Witten terms

A distinctive feature of higher Chern–Simons theory is the way gauge variation produces higher Wess–Zumino–Witten terms. In the semistrict Lie \(2\)-algebra construction, a finite higher gauge transformation is parametrized by
\[
(g,\sigma_g,\Sigma_g,\tau_g),
\]
with \((\sigma_g,\Sigma_g)\) a flat \(\mathfrak v\)-connection. The gauge-transformed \(2\)-connection obeys
\[
A^{g}=g_0(A-\sigma_g),\qquad
B^{g}=g_1\bigl(B-\Sigma_g+\tau_g(A-\sigma_g)\bigr)-\frac12 g_2(A-\sigma_g,A-\sigma_g),
\]
and the central variation formula is
\[
\mathcal{CS}_{2n+2}(A^g,B^g)-\mathcal{CS}_{2n+2}(A,B)
=
-\mathcal{CS}_{2n+2}(\sigma_g,\Sigma_g)+d\beta_{2n+1}.
\]
The term
\[
-\mathcal{CS}_{2n+2}(\sigma_g,\Sigma_g)
\]
is the higher WZW term [2605.29282].

The paper identifies the source of this term with the nonzero Jacobiator \([\cdot,\cdot,\cdot]\). Because the flatness equation for \(\Sigma_g\) contains
\[
d\Sigma_g+[\sigma_g,\Sigma_g]-\frac16[\sigma_g,\sigma_g,\sigma_g]=0,
\]
the three-bracket appears already in the Maurer–Cartan-type structure of the gauge parameter. The conclusion is explicit: in the semistrict case the higher WZW term is nontrivial, while in the strict crossed-module limit, where \([\cdot,\cdot,\cdot]=0\), it vanishes [2605.29282].

The earlier \(4\)-dimensional semistrict model exhibits the same pattern in a slightly different language. There the action transforms as
\[
\mathrm{CS}_2({}^g\omega,{}^g\Omega_\omega)=\mathrm{CS}_2(\omega,\Omega_\omega)-\kappa_2 Q_2(g),
\]
with higher winding number
\[
Q_2(g)=\frac14\int_N\left[2(d\sigma_g,\Sigma_g)-(\sigma_g,d\Sigma_g)\right]
=
\kappa_2^{-1}\mathrm{CS}_2(\sigma_g,\Sigma_g).
\]
Canonical quantization then leads to higher WZW Ward identities and explicit higher WZW actions obeying higher Polyakov–Wiegmann laws [1406.2197].

These results correct a common simplification. In higher gauge theory the gauge variation is not only an exact boundary term: in semistrict models it can contain a genuinely nontrivial topological contribution, and that contribution is controlled by the semistrict Jacobiator rather than by strict crossed-module data [2605.29282].

## 5. Concrete models and later refinements

Several concrete higher Chern–Simons models instantiate the abstract structure.

A prominent \(4\)-dimensional example is special \(2\)-Chern–Simons theory. It is built from the skeletal semistrict Lie \(2\)-algebra \(\mathfrak v_k(\mathfrak g)\) associated with a compact connected Lie group \(G\) with nontrivial center and a chosen central element \(k\). Its field content is a special \(G\)-\(2\)-connection \((\omega,\Omega_\omega)\) together with a fixed background closed \(3\)-form \(H\), and the full action is
\[
\mathrm{CS}_2(\omega;H)= \kappa_2 \int_N \left\{ \big(d\omega+\tfrac12[\omega,\omega],\Omega_\omega\big) -\frac16(\omega,k)(\omega,[\omega,\omega]) +8\pi^2(\omega,k)\,H \right\}.
\]
The equations of motion are
\[
f=0,\qquad F_f+8\pi^2Hk=0.
\]
Under a homotopically nontrivial \(1\)-gauge transformation \(\gamma\), the action is not strictly invariant but shifts the background as \(H\mapsto H+w(\gamma)\), where \(w(\gamma)\) is the winding number density; the paper therefore treats \(\kappa_2\) as a continuous parameter rather than imposing an established level quantization [1512.05977].

Another strict line of development uses crossed modules and \(2\)-crossed modules together with generalized differential calculus. In this approach a \(2\)-connection \((A,B)\) is packaged as a type-\(N=1\) generalized \(1\)-form \(\mathcal A=A+B\xi\), while a \(3\)-connection \((A,B,C)\) is packaged as a type-\(N=2\) generalized \(1\)-form. The paper states that this produces a \(2\)-Chern–Simons theory in dimension \(4\) and a \(3\)-Chern–Simons theory in dimension \(5\), together with higher second Chern forms and higher Chern–Weil theorems [2212.04667].

A later refinement addresses the fake-flatness obstruction. “Adjusting Higher Chern-Simons Theory” argues that naïve higher Chern–Simons theories based on homotopy Maurer–Cartan data do not yield a satisfactory off-shell higher gauge theory unless one imposes fake-flatness. The proposed remedy is an adjusted higher connection, but the paper proves that any cyclic adjusted skeletal \(n\)-term \(L_\infty\)-algebra with \(n>1\) is Abelian. To bypass this no-go statement it introduces half-adjusted higher Chern–Simons theories by passing to the shifted cotangent \(L_\infty\)-algebra and using the action
\[
S=\int_M \mathcal A_A^*\,\mathcal F^A.
\]
In the \(4\)-dimensional \(2\)-term case the action becomes
\[
S=\int_M\{A^*(H)+B^*(F)\},
\]
with adjusted curvatures \(F\) and \(H\). The resulting theory has well-defined kinematic data and the expected transgression property, but higher gauge transformations in cotangent directions must be excluded manually; that restriction is the source of the term “half-adjusted” [2507.02082].

## 6. Scope, adjacent uses, and mathematical significance

The literature uses “higher Chern–Simons” in several adjacent senses. One direction is higher-dimensional but not categorified: transgression forms for AdS gravity in dimension \(2n+1\) are written as
\[
\mathcal T_{2n+1}(A,\bar A)=\mathcal C_{2n+1}(A)-\mathcal C_{2n+1}(\bar A)-dB_{2n}(A,\bar A),
\]
or equivalently
\[
\mathcal T_{2n+1}(A,\bar A)=(n+1)\int_0^1dt\,\langle \Delta A\,F_t^{\,n}\rangle.
\]
These actions are strictly gauge invariant when both connections are transformed, automatically supply boundary terms, and yield finite Noether charges and Euclidean action for asymptotically AdS configurations with finite AdS gauge curvature [1407.6032].

A second adjacent usage is algebraic rather than categorified. The Nambu–Chern–Simons action is a \(3\)-dimensional action for the Nambu \(3\)-algebra of functions on \(S^3\),
\[
L_{\mathrm{SDiff}(S^3)}=\oint d^3\sigma\, e \left[ ds^i \wedge A_i - \frac{1}{3}\epsilon_{ijk} s^i \wedge s^j \wedge s^k \right],
\]
and the paper states explicitly that it is not a Lie \(2\)-algebra or \(2\)-connection theory; “higher” there refers to the underlying \(3\)-algebraic structure [1511.01482]. Likewise, higher-spin Chern–Simons theories of anyons remain ordinary \(3\)-dimensional Chern–Simons theories for enlarged higher-spin/fractional-spin gauge algebras, with master connection \(\mathbb A\) and flatness equation \(d\mathbb A+\mathbb A^2=0\) [1311.4589].

This suggests a persistent terminological ambiguity. In one strand, higher Chern–Simons action means a genuinely categorified gauge theory for higher connections on higher principal bundles, often controlled by Lie \(2\)-algebras or \(L_\infty\)-algebras. In another, it means a higher-dimensional transgression theory. In a third, it denotes a \(3\)-dimensional Chern–Simons action whose internal algebraic input has been enlarged from a Lie algebra to a higher-spin or \(3\)-algebraic structure. The sources agree, however, on a common structural core: a higher Chern–Simons action is characterized by transgression, curvature polynomials, and topological or quasi-topological gauge variation, with WZW-type boundary or gauge-parameter functionals appearing as natural descendants [2605.29282] [1301.2580].

Mathematically, the topic sits at the intersection of Chern–Weil theory, differential cohomology, higher stacks, and higher gauge theory. Physically, the same constructions are tied in the cited literature to anomalies, extended objects, topological field theory, higher WZW functionals, String and Fivebrane structures, and higher-dimensional effective actions [1207.5449] [1702.04160].

Source: https://www.emergentmind.com/topics/higher-chern-simons-action