BRST–BV Lagrangian Formulation

• BRST–BV Lagrangian is a field–antifield formulation that extends classical gauge theories by introducing ghosts, antifields, and higher ghosts to secure gauge invariance.
• It defines a master action S satisfying (S,S)=0, which generates a nilpotent BRST differential essential for encoding gauge transformations and higher consistency conditions.
• The formalism underpins diverse applications—from higher-spin and continuous-spin fields to AKSZ models—using techniques like free-field realizations, gauge fixing, and cohomological classification.

The BRST–BV Lagrangian is the field–antifield formulation of a gauge theory in which a classical action $S_0$ is extended by ghosts, antifields, and, in reducible cases, ghosts-for-ghosts, so that a master action $S$ of ghost number zero satisfies the master equation $(S,S)=0$ and defines a nilpotent BRST differential $sF=(S,F)$ (Kaparulin et al., 2011). In the global jet-bundle formulation used for AKSZ field theories, the same structure is expressed in terms of exterior horizontal forms on the infinite order jet space $J^\infty(a)$, with a horizontal $n$-form Lagrangian density $L\in \Omega^{n,0}(J^\infty a)$ satisfying $s_0L=d_H\Theta$; for a trivial $Q$-bundle with flat fiber, the local BRST cohomology is isomorphic to the cohomology of the target-space differential twisted by the de Rham cohomology of the base manifold (Bonavolontà et al., 2013).

1. Canonical field–antifield structure

In a Lagrangian gauge theory with local classical fields $\phi^i(x)$ and infinitesimal gauge transformations $S$0, the BV extension introduces physical fields $S$1 with $S$2, ghosts $S$3 with $S$4, antifields $S$5 with $S$6, and ghost antifields $S$7 with $S$8; reducible gauge algebras require ghosts-for-ghosts together with higher antifields. One often also tracks pure ghost number $S$9 so that $(S,S)=0$0, where $(S,S)=0$1 is antifield number (Kaparulin et al., 2011).

The canonical odd Poisson bracket is the BV antibracket. For local functionals $(S,S)=0$2 and $(S,S)=0$3, with $(S,S)=0$4 and $(S,S)=0$5, it is

$(S,S)=0$6

It has degree $(S,S)=0$7, satisfies graded antisymmetry and the Jacobi identity, and turns the master action into a homological generator (Kaparulin et al., 2011).

In the jet-bundle language of Bonavolontà–Kotov, the fields are sections of a finite-rank graded vector bundle

$(S,S)=0$8

over an $(S,S)=0$9-dimensional oriented manifold $sF=(S,F)$0. The infinite jet bundle $sF=(S,F)$1 carries local jet coordinates $sF=(S,F)$2, and the algebra of forms $sF=(S,F)$3 is bigraded by horizontal degree $sF=(S,F)$4 and vertical degree $sF=(S,F)$5. The purely horizontal forms $sF=(S,F)$6 are equipped with the horizontal differential

$sF=(S,F)$7

while the ghost number is inherited from the graded bundle and extended by $sF=(S,F)$8 (Bonavolontà et al., 2013).

2. Master equation and BRST differential

The extended action $sF=(S,F)$9 is required to satisfy two conditions: $J^\infty(a)$0 Expanding $J^\infty(a)$1 in antifield number gives the standard hierarchy

$J^\infty(a)$2

where the term linear in $J^\infty(a)$3 encodes gauge generators, the term linear in $J^\infty(a)$4 encodes structure functions, and higher terms appear in open or reducible algebras. The master equation at successive antifield numbers yields gauge invariance, structure relations, and higher consistency conditions (Kaparulin et al., 2011).

The BRST differential is defined by

$J^\infty(a)$5

with $J^\infty(a)$6 and $J^\infty(a)$7 if and only if $J^\infty(a)$8 (Kaparulin et al., 2011). In the jet-bundle setting one often splits

$J^\infty(a)$9

where $n$0 is the evolutionary BRST operator and $n$1 is the horizontal differential. For the AKSZ Lagrangian density $n$2,

$n$3

with $n$4 the presymplectic potential or Noether current, so that at the functional level $n$5 (Bonavolontà et al., 2013).

In the AKSZ construction, one starts from a target PQ-manifold $n$6 of degree $n$7, equipped with a symplectic form $n$8 of total degree $n$9 and a Hamiltonian function $L\in \Omega^{n,0}(J^\infty a)$0 of degree $L\in \Omega^{n,0}(J^\infty a)$1 satisfying $L\in \Omega^{n,0}(J^\infty a)$2. On the mapping space $L\in \Omega^{n,0}(J^\infty a)$3, the canonical degree-$L\in \Omega^{n,0}(J^\infty a)$4 functional is

$L\in \Omega^{n,0}(J^\infty a)$5

or, equivalently,

$L\in \Omega^{n,0}(J^\infty a)$6

The classical master equation

$L\in \Omega^{n,0}(J^\infty a)$7

encodes gauge invariance, closure, and all higher relations among the gauge symmetries (Bonavolontà et al., 2013).

3. Local BRST cohomology and global classification

Local BRST cohomology organizes local observables, anomalies, consistent deformations, and conservation laws. In the jet-bundle bicomplex, the local BRST cohomology in ghost number $L\in \Omega^{n,0}(J^\infty a)$8 and form degree $L\in \Omega^{n,0}(J^\infty a)$9 is defined by

$s_0L=d_H\Theta$0

and the physically most relevant group is the top-form sector $s_0L=d_H\Theta$1 (Bonavolontà et al., 2013).

Bonavolontà–Kotov proved that if the target is the formal $s_0L=d_H\Theta$2-manifold $s_0L=d_H\Theta$3 of finite type and the bundle is trivial $s_0L=d_H\Theta$4, then there is a natural isomorphism

$s_0L=d_H\Theta$5

This generalizes the local result of G. Barnich and M. Grigoriev from flat base manifolds to arbitrary base $s_0L=d_H\Theta$6. The proof uses an isomorphism between polynomial horizontal forms and symmetric multilinear differential operators, together with a spectral sequence argument that collapses at $s_0L=d_H\Theta$7 (Bonavolontà et al., 2013).

In the more general setting of non-Lagrangian gauge theory, the local BRST complex is controlled by the Koszul–Tate differential $s_0L=d_H\Theta$8, the longitudinal differential $s_0L=d_H\Theta$9, and their deformations. The group

$Q$0

is identified with Lagrange structures modulo trivial ones, and an integrable Lagrange structure satisfying the Maurer–Cartan condition yields a derived bracket on conservation laws that generalizes the Dickey bracket (Kaparulin et al., 2011). A crucial qualification is that, contrary to the usual BV formalism, such a complex does not always exist for non-Lagrangian dynamics, and when it exists it is by no means unique; the ambiguity and obstructions are controlled by explicit cohomology classes (Kaparulin et al., 2011).

This suggests a precise distinction between two regimes. In ordinary Lagrangian theories, the master action is available as a local solution of $Q$1, and cohomology classes measure standard gauge-theoretic structures. In non-Lagrangian theories, the existence of a BRST–BV complex is itself a cohomological question, mediated by Lagrange anchors and Lagrange structures rather than by an underlying action principle (Kaparulin et al., 2011).

4. Free-field realizations

A large class of BRST–BV Lagrangians is realized as quadratic master actions in oscillator or Fock-space form. In these constructions, gauge fields, ghosts, antighosts, and antifields are assembled into a single master ket or string-vector, and a nilpotent operator $Q$2 encodes both equations of motion and gauge symmetry. Metsaev develops this pattern for massless fields in flat space and AdS, conformal fields, and continuous-spin fields, while Reshetnyak develops it for mixed-antisymmetric and half-integer higher-spin systems (Metsaev, 2015, Metsaev, 2015, Metsaev, 2018, Reshetnyak, 2016, Reshetnyak, 2018).

Theory class	Representative BRST–BV form	Distinctive feature
Massless spin-$Q$3 fields in flat space and AdS	$Q$4, $Q$5, with $Q$6	Built in terms of traceless gauge fields and traceless gauge transformation parameters; Siegel gauge yields decoupled equations (Metsaev, 2015)
Totally symmetric conformal fields	$Q$7	Ordinary-derivative BRST–BV Lagrangian for arbitrary integer spin conformal fields (Metsaev, 2015)
Mixed-antisymmetric fields $Q$8	$Q$9	$\phi^i(x)$0-stage reducible gauge symmetry and minimal BRST–BV action in the minimal sector (Reshetnyak, 2016)
Continuous-spin fields	$\phi^i(x)$1	Siegel-gauge action invariant under global BRST and antiBRST transformations (Metsaev, 2018)

For mixed-antisymmetric higher-spin fields with Young tableaux $\phi^i(x)$2, Reshetnyak constructs irreducible and reducible gauge-invariant Lagrangians in flat $\phi^i(x)$3-dimensional space-time and suggests a minimal BRST–BV action that is a proper solution to the master equation in the minimal sector. The minimal field–antifield multiplet is

$\phi^i(x)$4

and the classical master equation follows from the nilpotency of the underlying BRST operator $\phi^i(x)$5 (Reshetnyak, 2016).

For half-integer totally symmetric higher-spin fields on $\phi^i(x)$6, the constrained BRST–BV formulation uses a minimal triplet $\phi^i(x)$7, a ghost $\phi^i(x)$8, and the corresponding antifields. The constrained master action has the form

$\phi^i(x)$9

with $S$00, and one checks directly that $S$01. The constrained complex incorporates gamma-tracelessness, while the unconstrained quartet formulation adds compensator and multiplier fields (Reshetnyak, 2018).

A related constrained BRST–BV description exists for bosonic fields with continuous spin on $S$02. In that setting the minimal BV action is written in terms of triplet-like fields, level-$S$03 ghosts, and Lagrange multipliers enforcing holonomic constraints, while the stage-one reducibility is represented explicitly through ghost-antifield couplings (Burdik et al., 2019).

5. Gauge fixing, $S$04 extension, and non-minimal sectors

Gauge fixing in the BRST–BV framework can be expressed either by a gauge-fixing fermion in the standard BV setting or, in the $S$05-extended formalism, by a bosonic gauge-fixing functional $S$06. In the $S$07-extended field–antifield formalism, the configuration variables are

$S$08

with Lagrange multipliers $S$09. The quantum action $S$10 satisfies

$S$11

or equivalently

$S$12

The gauge-fixed action is

$S$13

and may also be written as

$S$14

where $S$15 are the right $S$16-differentials (Batalin et al., 2014).

Batalin et al. show that finite field-dependent BRST–BV transformations generate arbitrary finite changes of the gauge-fixing function in the path integral. For a one-parameter family $S$17 defined by

$S$18

the Jacobian obeys an integrated formula involving the matrix $S$19. The compensation equation

$S$20

ensures that the Jacobian induces the desired finite shift $S$21, and the partition function remains unchanged: $S$22 This is a precise statement that the Jacobian of a finite BRST–BV transformation is capable of generating arbitrary finite change of the gauge-fixing function in the path integral (Batalin et al., 2014).

In non-minimal constrained BRST–BV formulations for totally symmetric massless higher-spin fields, the non-minimal BRST operator $S$23 includes additional BFV oscillators, antighosts, and Nakanishi–Lautrup fields. The gauge-fixing procedure is organized by a BFV gauge-fixing fermion $S$24 and a BV gauge-fixing fermion functional

$S$25

followed by the canonical shift

$S$26

This is described in the paper as BFV–BV duality, and the resulting quantum action incorporates a two-parameter $S$27-type gauge condition (Burdík et al., 2020).

Several free-field realizations use particularly simple gauges. In the AdS BRST–BV approach adapted to AdS/CFT, the Siegel gauge

$S$28

sets all antifields to zero and gives the gauge-fixed Lagrangian

$S$29

which yields decoupled equations of motion (Metsaev, 2015). In the continuous-spin case, the condition

$S$30

leads to

$S$31

and the gauge-fixed action is invariant under global BRST and antiBRST transformations (Metsaev, 2018).

6. Deformations, interactions, and recent generalizations

The BRST–BV deformation procedure constructs interactions by preserving the master equation order by order in a coupling constant $S$32. One writes

$S$33

and imposes

$S$34

At first nontrivial order, $S$35 is a cocycle of the free BRST differential, while higher orders test closure and obstructions (Reshetnyak, 2023, Reshetnyak, 2018).

For interacting higher-spin fields on $S$36-dimensional Minkowski space, Reshetnyak develops the BRST–BV approach to construct general off-shell Lorentz covariant cubic, quartic, and $S$37-tic interaction vertices. In oscillator language the cubic contribution is encoded by a vertex operator $S$38 satisfying

$S$39

where $S$40. The free minimal master action is

$S$41

and the cubic deformation is written as a BRST–closed contribution $S$42 built from $S$43 (Reshetnyak, 2023).

The same deformation logic is already present in free higher-spin BRST–BV constructions. The mixed-antisymmetric formulation explicitly states that the minimal BRST–BV action provides objects appropriate to construct interacting Lagrangian formulations with mixed-antisymmetric fields in a general framework (Reshetnyak, 2016). The constrained half-integer formulation similarly proceeds from the free master action to interactions through

$S$44

with the first-order condition $S$45 solved in the constrained field–antifield complex (Reshetnyak, 2018).

A recent generalization extends the BRST–BV framework to free fields of arbitrary masses and symmetry types in the Poincaré patch of AdS. In this approach the master field

$S$46

defines the universal action

$S$47

with

$S$48

The construction covers massless, massive, partially-massless, and continuous-spin fields; constrained and unconstrained formulations; and matching to metric-like Lagrangians formulated in terms of the modified de Donder divergence. It also gives a realization of $S$49 on the space of fields and antifields (Metsaev, 2 Jul 2026).

A recurrent misconception is that the BRST–BV Lagrangian is exhausted by the familiar local master action of ordinary Lagrangian gauge theory. The available results show a broader but sharply delimited scope: the formalism extends to AKSZ sigma models, $S$50-extended quantization, and several constrained higher-spin systems, while in genuinely non-Lagrangian dynamics the existence and uniqueness of the local BRST complex become nontrivial cohomological problems controlled by $S$51 and related classes [(Bonavolontà et al., 2013); (Batalin et al., 2014); (Kaparulin et al., 2011)].