Physicists Map the Hidden Gauge Structure of Non-Abelian Tensor Fields

Antisymmetric tensor fields, known in the trade as p-forms, are among the strangest and most indispensable ingredients of modern theoretical physics. They appear throughout supergravity and superstring and brane theory, and they carry a gauge structure far more intricate than that of the familiar photon or gluon. Now, in a paper published in The European Physical Journal C, theorists I. Buchbinder and N. Kozyrev have taken a long-standing open question about these fields and answered it in detail: when antisymmetric tensor fields interact in a genuinely non-Abelian way, does their characteristic reducible gauge symmetry survive? The answer, according to their analysis, is yes, and the finding matters because that symmetry is precisely what any future quantization of such theories must confront.

To appreciate why the question is hard, it helps to start with the simplest example, the 2-form field with potential B_mu_nu, introduced long ago by Ogievetsky and Polubarinov and later rediscovered in string theory. This field changes under a gauge transformation built from an antisymmetrized derivative of a vector parameter, and its field strength is a 3-index tensor. The peculiarity is that the transformation can be trivial even for a nonzero parameter: if the parameter itself is a gradient, the field does not change at all. Two gauge parameters that differ by such a gradient therefore generate the identical transformation, meaning the parameter is defined only up to its own gauge transformation. This property is called reducibility, and the 2-form is reducible at the first stage, something with no analogue in electromagnetism, Yang-Mills theory or gravity.

For higher-rank forms the pattern deepens. A 3-form transforms with a rank-2 parameter, but that parameter is itself ambiguous up to an antisymmetrized derivative of a rank-1 parameter, which is in turn ambiguous up to a derivative of a scalar. The 3-form is thus reducible at the second stage, and a general p-form is reducible at stage p minus one. This layered structure is a genuine headache for quantization. The standard Faddeev-Popov procedure, which introduces gauge-fixing conditions and ghost fields to construct a path integral over gauge-group orbits, simply fails for reducible theories because the Faddeev-Popov-DeWitt determinant vanishes. The general covariant remedy was developed by Batalin and Vilkovisky, and more recent work has proposed simplified procedures that generalize the Faddeev-Popov approach to reducible gauge theories. Either way, before one can quantize a theory, one must know exactly how reducible its gauge transformations are.

Interactions make the problem dramatically harder. No-go theorems show that the transformation laws of a free form cannot simply be deformed into a non-Abelian pattern if the theory is local and no extra fields are added. Yet the search for non-Abelian tensor models continues, driven largely by the physics of stacks of M5-branes and by the expectation that the maximal superconformal theory, the so-called N=(2,0) theory in six dimensions, should be described by some generalization of the tensor multiplet. Non-Abelian antisymmetric tensors do exist in certain gauged supergravities, such as seven-dimensional supergravity with 3-form fields in the SO(5) vector of an internal symmetry group, or five-dimensional theories with 2-form fields, although in those settings the shift symmetry has been gauge-fixed. Tensor hierarchies arising from non-Abelian forms have also been studied in superconformal theories, exceptional field theory embeddings and formulations of eleven-dimensional supergravity in terms of four-dimensional superfields.

The route taken by Buchbinder and Kozyrev follows a construction pioneered by Scherk and Schwarz: dimensional reduction on a group manifold. The idea is to start with a free Abelian form minimally coupled to gravity in a higher-dimensional spacetime, and to compactify part of the geometry on a compact Lie group. The coordinates split into ordinary spacetime coordinates and internal ones, and every field or parameter carrying an internal index is expanded in the Killing vectors of the group manifold. A crucial feature emerges immediately: the size of the compact manifold is set by a dimensionful parameter kappa, and this parameter, with the dimension of mass, plays the role of the coupling constant of the resulting lower-dimensional theory. Geometry itself generates the interaction.

Working through this reduction for the 3-form case, the authors find that the higher-dimensional field decomposes into a set of lower-dimensional objects: a 3-form, 2-forms carrying a group index, 1-forms carrying two, and scalars carrying three, alongside the Yang-Mills gauge field that arises from the mixed metric components. In the original basis the tensor fields transform inhomogeneously, but a field redefinition, naturally expressed through the inverse vielbein of the compact space, produces fields that transform covariantly under the Yang-Mills symmetry. The resulting gauge algebra is demonstrably non-Abelian: the commutator of two transformations closes on a third transformation whose parameters are built from structure constants contracted with the original parameters, and remarkably the algebra closes without any reference to the equations of motion. The reduced action, a sum of squares of appropriately covariantized field strengths, was previously known to yield consistent equations of motion, making it a sensible non-Abelian tensor model.

The central new result concerns reducibility. Because the Yang-Mills field is dynamical, any transformation with a nonzero group-theoretic parameter is nontrivial, so only transformations with that parameter set to zero can be trivial. Solving directly for the conditions under which all fields are left unchanged, the authors find a complete set of trivial transformations: a vector parameter generates a covariant gradient shift in one gauge parameter, which is compensated by a structure-constant-proportional shift in another, with a further compensating term involving the Yang-Mills field strength in the Abelian sector. For a semisimple group such as SU(N) or SO(N), these are shown to be the only trivial transformations. The upshot is that the non-Abelian transformations are reducible at the first stage, and the full 3-form system is reducible at the second stage, exactly matching the free Abelian case despite the presence of mass terms and interactions.

The 4-form system extends the pattern one rung up the ladder. Reduction produces a tower of tensors up to rank four, and the non-Abelian gauge algebra again closes on structure constants. The trivial transformations now involve a rank-2 parameter, whose own ambiguity is resolved by a rank-1 parameter, whose ambiguity is resolved by a scalar. The non-Abelian part of the transformations is therefore reducible at the second stage, and the Abelian sector at the third, again one stage more than the 3-form case and again mirroring the free theory. A subtle point arises in the scalar sector, where an apparently additional symmetric solution to the triviality conditions turns out to be incompatible with the full set of transformation laws, so no new reducibility appears there.

Two further structural features deserve attention. First, the reduction generates mass terms for some of the non-Abelian tensors, proportional to the coupling kappa, which would ordinarily break the standard form gauge symmetry. The symmetry is rescued by Stückelberg fields, combinations of the other tensors whose transformation laws are pure shifts under the corresponding gauge parameters, and which can in principle be gauged away. It is precisely these mass deformations, absent from the earlier no-go analyses, that allow the construction to bypass those theorems. Second, the authors extend the analysis to a form of arbitrary rank, showing that the reduced system contains non-Abelian antisymmetric tensors of rank p minus one and below, with transformations reducible at stage p minus two, and they contrast this with a related non-Abelian model from the literature whose transformations are only first-stage reducible, unlike the Abelian case.

The authors summarize the message plainly: all the models they consider, obtained by reducing free antisymmetric tensors minimally coupled to gravity on a group manifold, are reducible gauge theories whose reducibility stages coincide with those of the corresponding free Abelian theories, with additional reducibility in the vector and scalar sectors, and the non-Abelian theories necessarily contain Stückelberg fields. The results open several directions. The method should apply to interacting theories and to self-dual actions of the Pasti-Sorokin-Tonin type in higher dimensions, since the construction relied little on the detailed form of the actions. It remains open whether the findings extend to reductions on manifolds other than groups, such as the sphere reductions of ten- and eleven-dimensional supergravity. Looking forward, the authors point to canonical and covariant quantization via BRST-BFV and Batalin-Vilkovisky methods, the study of divergence structure and quantum effective actions, the construction of supersymmetric versions of the models, and the search for dualities among interacting non-Abelian forms, all of which hinge on the specific structure of the reducible gauge transformations that this work has now mapped out.

Subject of Research: The gauge structure and reducibility of interacting non-Abelian antisymmetric tensor field models obtained by dimensional reduction on group manifolds.

Article Title: On interacting non-Abelian antisymmetric tensor field models

Article References: Buchbinder, I., & Kozyrev, N. (2026). On interacting non-Abelian antisymmetric tensor field models. The European Physical Journal C, 86(10), Article 1239. https://doi.org/10.1140/epjc/s10052-026-16512-3

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16512-3

Keywords: gauge theory, antisymmetric tensor fields, p-forms, dimensional reduction, non-Abelian symmetry, reducibility, Stückelberg fields, quantization, supergravity, theoretical physics, interacting, non-Abelian