---
title: Dipolar Symmetries in Gauge and Topological Theories
url: https://www.emergentmind.com/topics/dipolar-symmetries
type: topic
---

# Dipolar Symmetries in Gauge and Topological Theories

Searching arXiv for recent and foundational papers on dipolar symmetries and closely related dipole-symmetry field theories.
arxiv_search({"query":"all:dipolar symmetry OR all:dipole symmetry", "max_results": 10, "sort_by": "submittedDate", "sort_order": "descending"})
arxiv_search({"query":"ti:\"dipole symmetry\" OR ti:\"dipolar symmetry\" OR abs:\"dipole symmetry\" OR abs:\"dipolar symmetry\"", "max_results": 10, "sort_by": "submittedDate", "sort_order": "descending"})
arxiv_search({"query":"id:2403.08158 OR id:2305.02492 OR id:2311.04962 OR id:2503.15834 OR id:2309.10036 OR id:2304.12911 OR id:2304.01181 OR id:2303.04479 OR id:2304.09927", "max_results": 20, "sort_by": "submittedDate", "sort_order": "descending"})
Dipolar symmetry is a multipole or modulated symmetry in which ordinary charge conservation is supplemented by conservation laws for dipole moments or for coordinate-weighted charge combinations. In continuum formulations it appears through relations such as
\[
\partial_\mu J_\mu=0,\qquad \partial_\mu\!\left[J_\mu^a-(x^a-x_0)J_\mu\right]=0,
\]
while on lattices it is often represented by a charge generator \(Q\) and a dipole generator \(D\) whose algebra with translations is nontrivial, for example \(T^{-1}QT=Q\) and \(T^{-1}DT=QD\). This structure distinguishes dipolar symmetry from ordinary onsite internal symmetry and from standard higher-form symmetry, and it is the source of restricted mobility, rank-2 gauge descriptions, modulated boundary anomalies, and position-dependent braiding phases [2403.08158][2311.04962][2508.06604].

## 1. Algebraic structure and modulated character

Dipolar symmetries are a prime example of modulated symmetries: internal symmetries that do not commute with spatial symmetries. In the defect-network formulation, the full symmetry group is a semidirect product
\[
G = G_{\rm int}\rtimes G_{\rm sp},
\]
so that each spatial symmetry \(S\in G_{\rm sp}\) acts on the internal symmetry by an automorphism \(S(g)=SgS^{-1}\). In one spatial dimension, the characteristic relation is \(T(D)=QD\), while in two dimensions one has separate dipole generators \(D_x,D_y\) with \(T_x(D_x)=QD_x\), \(T_y(D_x)=D_x\), and the analogous relations for \(D_y\) [2508.06604].

For continuous \(U(1)\) charge, the dipolar generator can be written as
\[
U(\alpha)=\exp\!\left(i\alpha\sum_x x\,q_x\right),
\]
where \(q_x\) is the charge density. In two dimensions,
\[
U_x(\alpha_x)=\exp\!\left(i\alpha_x\sum_{x,y}x\,q_{x,y}\right),\qquad
U_y(\alpha_y)=\exp\!\left(i\alpha_y\sum_{x,y}y\,q_{x,y}\right),
\]
and translation changes these generators by a charge factor. For discrete \(\mathbb{Z}_N\) charge, the corresponding dipole generator is
\[
D=\exp\!\left(\frac{2\pi i}{N}\sum_x x\,q_x\right),
\]
again with \(T(D)=QD\) and \(T(Q)=Q\) [2508.06604].

In one-dimensional dipolar SPT constructions this algebra is implemented explicitly by
\[
Q_g=\prod_x U_g(x),\qquad D_g=\prod_x [U_g(x)]^x,
\]
with
\[
T^{-1}Q_gT=Q_g,\qquad T^{-1}D_gT=Q_gD_g.
\]
For \(G=\mathbb{Z}_N\), a concrete realization is
\[
Q=\prod_j X_j,\qquad D=\prod_j (X_j)^j.
\]
This translation-entangled structure is what prevents the usual edge fractionalization of charge operators and leads instead to mixed charge-dipole projective data [2311.04962].

A recurring subtlety is that modulated generators are generically natural on open chains or infinite systems, but can fail to exist globally on a ring. For dipolar symmetry on a periodic chain of length \(L\), \(D\) is strictly periodic only when \(L\) is a multiple of \(N\); otherwise only a power \(D^k\), with \(k=N/\gcd(L,N)\), survives. This nonexistence of a global generator under periodic boundary conditions motivates the bundle-symmetry formulation discussed below [2309.10036].

## 2. Conservation laws and continuum formulations

In \(2+1\) dimensions, dipolar symmetry supplements ordinary charge conservation by tying charge transport to dipole transport. In the dipolar Chern-Simons precursor and in dipolar BF theory, the fundamental gauge fields are an ordinary one-form \(A_\mu\) and dipole-labeled one-forms \(A_\mu^a\), with gauge transformations
\[
A_\mu^a\to A_\mu^a+\partial_\mu\beta^a,\qquad
A_\mu\to A_\mu+\partial_\mu\alpha+\beta^\mu,
\]
together with the constraints \(\beta^t=0\) and \(A_\mu^t=0\). Gauge invariance forces
\[
\partial_\mu J_\mu=0,\qquad
\partial_\mu J_\mu^x=J_x,\qquad
\partial_\mu J_\mu^y=J_y,
\]
or equivalently
\[
\partial_\mu\!\left[J_\mu^x-(x-x_0)J_\mu\right]=0,\qquad
\partial_\mu\!\left[J_\mu^y-(y-y_0)J_\mu\right]=0.
\]
The conserved quantities are therefore charge-dipole combinations relative to a reference point, not independent conserved charge and dipole currents in the ordinary sense [2403.08158].

Dipolar BF theory extends this structure by introducing BF partners \(b_\mu\) and \(b_\mu^a\), plus additional currents \(K_\mu\) and \(K_\mu^a\). Gauge invariance then yields, besides \(\partial_\mu J_\mu=0\), the conservation laws
\[
\partial_\mu K_\mu^a=0,\qquad
\partial_\mu\!\left[K_\mu+(x-x_0)K_\mu^x+(y-y_0)K_\mu^y\right]=0.
\]
These equations encode conservation of charge, intrinsic dipole currents in the \(K\)-sector, conserved dipole charges of the \(J\)-sector, and a modified charge built from \(K\) and \(K^a\) [2403.08158].

The same continuity structure can be derived from a more general bosonic field-theory framework. If the action is written in Hamiltonian form with symplectic potential \(\omega_a(\phi)\), then in \(d=2\) the scalar charge density is
\[
\rho = J^0 = \varepsilon^{ij}\partial_i\omega_j,
\]
and the symmetric tensor current
\[
J^{ij}=\frac{1}{2}\left(\varepsilon^{ik}\sigma^j{}_k+\varepsilon^{jk}\sigma^i{}_k\right)
\]
descends from the stress tensor \(\sigma^{ij}\). One obtains the dipole continuity equation
\[
\partial_t\rho+\partial_i\partial_j J^{ij}=0.
\]
Under suitable decay conditions this implies conservation of
\[
Q=\int d^2x\,\rho,\qquad D^i=\int d^2x\,x^i\rho,
\]
and, when \(J^{ij}\) is traceless, also of \(X=\int d^2x\,|x|^2\rho\) [2303.04479].

A common misconception is that dipolar symmetry is simply a higher-form symmetry in disguise. The continuum formulations explicitly reject this identification: the coordinates appear in the conserved quantities because the symmetry constrains multipole moments, not because it measures the flux of an ordinary \(p\)-form current. This is why dipolar symmetry is described as a subsystem or multipole symmetry rather than a standard higher-form symmetry [2403.08158].

## 3. Dipolar gauge theories, braiding, and topological response

The central topological field-theory result is that dipolar BF theory is equivalent to the rank-2 tensor BF theory used as an effective theory for the rank-2 toric code, but the dipolar BF formulation makes the dipole symmetry explicit. Under the field redefinition
\[
A_{ab}=\partial_aA_b-A_a^b,
\]
the dipolar theory maps to rank-2 gauge fields \(A_{ab}\) and electric fields \(E\), and the resulting Gauss laws involve the composite charge density
\[
\tilde J_t = J_t+\partial_xJ_t^x+\partial_yJ_t^y.
\]
This expresses a key physical statement: a mobile scalar charge is not a bare charge, but a composite charge-dipole object [2403.08158].

The quasiparticles of dipolar BF theory are either charge-like or dipole-like. Dipole quasiparticles can move freely within their sector, but bare charges cannot move without inducing orbital dipole currents. The braiding statistics therefore departs sharply from ordinary topological phases. For a charge braided around enclosed dipoles, the phase contains explicitly position-dependent contributions,
\[
\varphi_{J\circlearrowleft \text{dipoles}}
=
-\frac{2\pi}{N}\left[\sum_i(x_i-x_0)+\sum_i(y_i-y_0)\right]
+\int_{\rm area} F_{xy}\,dx\,dy,
\]
with
\[
F_{xy}=\partial_xA_y-\partial_yA_x+A_y^x-A_x^y.
\]
By contrast, braiding of same-orientation dipoles remains ordinary anyonic:
\[
\varphi_{K^a\circlearrowleft K^a}=\frac{2\pi}{N}\times(\text{number of enclosed }K^a\text{-dipoles}).
\]
This position dependence is the defining hallmark of dipolar braiding [2403.08158].

Dipolar Chern-Simons theory gives a complementary anomaly-based perspective. In this formulation both \(U(1)\) and dipole gauge fields are ordinary one-forms, and the gauge transformation
\[
A_\mu\to A_\mu+\partial_\mu\alpha+\delta_{\mu a}\xi^a,\qquad
A_\mu^a\to A_\mu^a+\partial_\mu\xi^a
\]
implies that an ordinary \(U(1)\) Chern-Simons term is not gauge invariant in the presence of dipole symmetry. The resulting conclusion is that only the highest multipole symmetry can support a ’t Hooft anomaly. In \(1+1\) dimensions the anomaly inflow from the \(2+1\)-dimensional bulk produces a chiral cubic boundary mode,
\[
\omega=-\frac{2C}{\chi}k_x^3,
\]
while in \(3+1\) dimensions point-group invariant mixed Chern-Simons terms yield boundary anomalies whose chirality and dispersion depend on the boundary orientation [2305.02492].

## 4. Dipolar SPT phases, bundle symmetry, and translation enrichment

For finite abelian \(G\), one-dimensional dipolar SPT phases are classified by
\[
H^2[G\times G,U(1)]/H^2[G,U(1)]^2.
\]
The quotient is a direct consequence of the dipole-translation algebra: in a dipole-symmetric matrix product state, the virtual charge action must be linear, so purely charge projective classes are forbidden, and the diagonal subgroup projectivity is fixed by the mixed charge-dipole data. The surviving phases are therefore genuinely dipolar rather than ordinary onsite SPT phases [2311.04962].

For \(G=\mathbb{Z}_N\), an explicit commuting-projector realization is
\[
a_j=(Z_{j-1}Z_j^\dagger)^\eta\,X_j\,(Z_j^\dagger Z_{j+1})^\eta,\qquad
H=-\sum_j(a_j+a_j^\dagger).
\]
On an open chain, the edge operators satisfy
\[
L_Q L_D = e^{2\pi i\eta/N} L_D L_Q,
\]
and the minimal protected edge degeneracy is
\[
\frac{N}{\gcd(N,\eta)}.
\]
This algebra is the direct edge signature of dipolar SPT order [2311.04962].

Periodic boundary conditions require a further refinement. In modulated-symmetry chains, a global dipole generator may fail to exist even though the phase remains nontrivial. The bundle-symmetry construction addresses this by defining symmetry sections patchwise. On overlaps, the sections are glued by transition functions, and the modulated SPT order survives as long as these local sections commute with the Hamiltonian on their support. In this sense, dipolar SPT order can persist on a ring even when the global dipole symmetry is obstructed [2309.10036].

In two dimensions, the dipolar cluster state and the dipolar topological state sharpen this structure. The dipolar cluster state is an SPT protected by charge \(\mathbb{Z}_N\times\mathbb{Z}_N\), dipole bundle symmetry \(g_b(A_\alpha)\), and three \(1\)-form symmetries. Its boundary symmetry operators obey anomaly relations such as
\[
g_h^u g_{X2}=\omega\, g_{X2}g_h^u,\qquad
g_v^u g_{X1}=\omega\, g_{X1}g_v^u,\qquad
g_d^u g_{X3}=\omega\, g_{X3}g_d^u,
\]
implying \(N^3\) boundary zero modes. The post-measurement dipolar topological state is obtained by simultaneous gauging of two charge symmetries and one dipole symmetry; sequential gauging fails because the relevant gauge symmetry operators share the same plaquette gauge field. The resulting phase is translation symmetry-enriched: translations permute anyon species, and the dipole symmetry intertwines with charge symmetries under translation [2503.15834].

A defect-network reformulation extends these conclusions. For modulated symmetries treated as \(G_{\rm int}\rtimes G_{\rm sp}\), a decorated-cell construction is anomaly-free only if the cohomology class on each top-dimensional cell satisfies
\[
S^*[\omega]=[\omega]
\]
for every spatial symmetry \(S\). This criterion reproduces the low-dimensional classifications of translation plus dipolar symmetry and shows explicitly how some defect networks that are non-anomalous for unmodulated internal symmetries become anomalous after modulation [2508.06604].

## 5. Symmetry breaking, low-energy modes, and spectral constraints

When dipolar symmetry is spontaneously broken in fermionic models, single-particle dispersion arises through dipole condensation. The broken-symmetry phase contains Goldstone modes whose generators fail to commute trivially with translations,
\[
[P_j,Q_i]=i\delta_{ij}Q,
\]
so the Goldstones couple strongly to the Fermi surface at zero momentum. In \(d=2\), the low-energy patch theory is analogous to a Fermi surface coupled to a transverse \(U(1)\) gauge field. The Goldstone self-energy is Landau damped,
\[
\Pi_{ab}(i\omega,\mathbf p)\approx -\gamma\frac{|\omega|}{|\mathbf p|}
\left(\delta_{ab}-\frac{p_ap_b}{|\mathbf p|^2}\right),
\]
and the fermion self-energy becomes
\[
\Sigma(i\omega)\propto -\,i\,\mathrm{sgn}(\omega)\,|\omega|^{2/3}.
\]
If dipolar symmetry is weakly broken explicitly, the theory crosses over to Fermi-liquid behavior at a scale \(T^*\sim r_0^{3/2}\) [2304.01181].

A distinct large-\(N\) bosonic construction realizes the monopole-dipole-momentum algebra in a discretized internal space coupled to a constant background \(V_x\). In its quantum dipole-breaking phase, monopole charge remains unbroken while the phase of a composite dipole order parameter becomes the low-energy field. The effective Nambu-Goldstone action is
\[
\mathcal L_{\rm NG}
=
\frac{V_x^2}{240\pi \tilde m_\phi^4}
\left[
10\tilde m_\phi^2(\partial_t\theta)^2
-(\partial_x\partial_t\theta)^2
+(\partial_t^2\theta)^2
\right].
\]
Because every term contains at least one time derivative, the action has the emergent subsystem symmetry
\[
\theta(t,x)\to \theta(t,x)+f(x),
\]
and the physical branch is exactly flat, \(\omega^2=0\). The resulting fractonic Nambu-Goldstone mode is immobile, and the equal-time order-parameter correlator does not decay, so the Coleman-Hohenberg-Mermin-Wagner argument is avoided in \(1+1\) dimensions [2304.12911].

Dipolar symmetry also constrains quantum Hall kinematics. After projection to the lowest Landau level, translational invariance and particle conservation combine into dipole symmetry, with
\[
D_i=-\ell_B^2\varepsilon_{ij}P_j.
\]
The projected density obeys
\[
\bar\rho_{\mathbf k}=Q-i k_j D_j+O(k^2),
\]
and the Ward identities \([\,\bar H,Q\,]=[\,\bar H,D_i\,]=0\) force the LLL \(f\)-sum rule to scale as \(|k|^4\). Under the assumptions stated for the projected problem, this quartic behavior forbids spontaneous microscopic \(U(1)\) breaking at zero temperature in two dimensions. If the magnetic field becomes spatially inhomogeneous, the dipole symmetry disappears and the constraint is lifted [2304.09927].

A broader geometric statement follows from the symplectic-field-theory derivation of dipole conservation. When the symplectic form is closed but not exact, the translation algebra is centrally extended,
\[
\{P_i,P_j\}=-\varepsilon_{ij}Q,
\]
and a well-defined local momentum density can fail to exist. For a translationally invariant microscopic theory, this obstruction implies that the low-energy description must contain additional light modes that cure the anomaly [2303.04479].

## 6. Terminological range in dipolar matter

The literature also uses the phrase “dipolar symmetries” in a technically different sense: not as multipole charge conservation, but as symmetry classification induced by anisotropic dipole-dipole interactions. In quasi-two-dimensional dipolar spinor condensates, the relevant symmetry objects are combined real-space and spin-space operations
\[
g=(M,t,O)\in E(2)\times O(3),
\qquad
\hat n^i(x)\to O^{ij}\hat n^j(Mx+t).
\]
Imposing the constraints of a rectangular lattice, \(D_2\) point-group symmetry, vanishing net magnetization, vanishing net skyrmion charge, and full coverage of spin space restricts the allowed space groups to \(p2mm\), \(p2mg\), or \(p2gg\), with 11 compatible spin groups in total. Within these symmetry classes, the minimal-energy textures are neutral skyrmion stripe crystals rather than conventional uniform spin states [1008.2240].

A second example is microwave-dressed dipolar molecular gases, where the effective interaction is parameterized by \((\epsilon_0,\epsilon_2)\). Here the real-space symmetry is generically \(D_{2h}\), but it is enhanced to \(C_{\infty v}\) on the lines \(\epsilon_2=0\), \(\epsilon_0=\epsilon_2/\sqrt3\), and \(\epsilon_0=-\epsilon_2/\sqrt3\). In parameter space, the interaction is invariant under a \(D_3\) action generated by coordinate permutations, so each point belongs to a sextet of degenerate interaction parameters related by axis relabeling. The same symmetry analysis classifies prolate, oblate, and triaxial self-bound droplets, with thermodynamic-limit stability when
\[
\epsilon_0+\sqrt3|\epsilon_2|>1
\qquad\text{or}\qquad
\epsilon_0<-\frac12.
\]
This is a symmetry of the interaction potential rather than a dipole-moment conservation law [2510.04634].

This range of usages suggests that the phrase “dipolar symmetries” presently covers two distinct technical structures. In one, the defining ingredient is a modulated charge-dipole algebra with translations, leading to multipole conservation, restricted mobility, and topological response. In the other, the defining ingredient is the anisotropic symmetry of dipolar interactions themselves, which organizes textures, phases, and equilibrium shapes. The distinction is conceptually important, because only the first sense carries the continuity equations, bundle symmetries, and rank-2 gauge-theory constructions associated with dipole conservation.

Source: https://www.emergentmind.com/topics/dipolar-symmetries