Koszul–Tate Resolution and the BRST Bicomplex

The Koszul–Tate differential resolves the algebra of functions on the stationary surface: it kills the equations of motion and every Noether identity among them, including higher reducibility identities. A longitudinal differential then resolves gauge orbits. Under explicit regularity and completeness hypotheses, their filtered total complex has ghost-number-zero cohomology equal to gauge-invariant on-shell functions. The conclusion is homological and local; it neither supplies a gauge-fixed measure nor proves positivity of a quantum state space.

Required background. Elliptic gauge complexes and Gribov obstructions distinguish a local gauge complex from a global slice; chains, homology, and exactness supply resolutions and contracting homotopies; and the BRST differential supplies the physical gauge-system notation used below.

Helpful background. Derived critical loci and gauge quotients explain why the extra homological directions record equations and stabilizers rather than new classical particles.

The resolution behind the BRST bicomplex

Let $\mathcal F$ be a graded-commutative algebra of local functions of fields $\phi^i$ and finitely many derivatives. The Euler–Lagrange expressions $E_i=\delta S_0/\delta\phi^i$ generate an on-shell ideal $\mathcal I\subset\mathcal F$. Simply quotienting by $\mathcal I$ loses relations $R^i_{\alpha}E_i=0$. Instead introduce antifields $\phi_i^*$ of antifield number one, ghost antifields $c^*_{\alpha}$ of antifield number two, and further generators for every reducibility stage. The Koszul–Tate differential $\delta$ lowers antifield number by one and begins as

$$\delta\phi_i^*=E_i, \qquad \delta c^*_{\alpha}=R^i_{\alpha}\phi_i^*, \qquad \delta\phi^i=\delta c^{\alpha}=0.$$

Signs are fixed by using left derivatives and by requiring $\delta^2=0$. Nilpotence on $c^*_{\alpha}$ is precisely the Noether identity. A Koszul–Tate resolution means more than nilpotence: $H_0(\delta)\cong\mathcal F/\mathcal I$ and $H_k(\delta)=0$ for $k>0$. The acyclicity statement requires a regular stationary surface, a complete set of Noether identities, and all reducibility generators. Barnich, Brandt, and Henneaux prove this local-form statement by changing jet coordinates into contractible pairs Barnich, Brandt, and Henneaux 2000, Theorem 5.1 and §5.2, pp. 33–35.

Give ghosts pure ghost number one and define total ghost number by

$$\operatorname{gh}=\operatorname{pgh}-\operatorname{afn}.$$

The longitudinal differential $\gamma$ raises pure ghost number and differentiates along gauge orbits. For a closed irreducible algebra, the total BRST differential starts with $s=\delta+\gamma$; for open or reducible algebras, higher-antifield-number terms are generally required. Filtering by antifield number gives a spectral sequence whose first page is $H(\delta)$. Only after the positive-antifield homology vanishes may one identify the next differential with gauge action on the stationary surface. Thus $s^2=0$ alone is not the converse of the resolution theorem.

Maxwell theory as an exact test

For a Maxwell potential on a contractible region, take

$$S_0=-\frac14\int F_{\mu\nu}F^{\mu\nu}\,d^4x, \qquad F_{\mu\nu}=\partial_\mu A_\nu-\partial_\nu A_\mu .$$

With $\operatorname{afn}(A^{*\mu})=1$ and $\operatorname{afn}(c^*)=2$, one convenient convention is

$$\delta A^{*\mu}=\partial_\nu F^{\nu\mu}, \qquad \delta c^*=-\partial_\mu A^{*\mu}, \qquad \gamma A_\mu=\partial_\mu c, \qquad \gamma c=0.$$

The identity $\partial_\mu\partial_\nu F^{\nu\mu}=0$ gives $\delta^2c^*=0$, while $\delta\gamma+\gamma\delta=0$. Hence $s=\delta+\gamma$ is nilpotent. At ghost number zero, $s$-closed functions reduce, modulo $s$-exact terms, to functions of $F_{\mu\nu}$ and its derivatives restricted by Maxwell’s equations. This is the concrete Abelian construction developed physically on the BRST differential page. The independent checks are transparent: $sF_{\mu\nu}=0$, the divergence of the Euler–Lagrange expression vanishes identically, and every displayed differential has the advertised bidegree.

The calculation is also a small spectral-sequence check. At the first stage, $\delta$ removes the Maxwell equations and the identity among them; at the next, $\gamma$ removes longitudinal gauge dependence. Because the filtration is bounded below at each fixed total degree and the relevant local expressions have finite antifield number, the sequence stabilizes in this finite calculation. In a completed algebra of nonlocal functionals, convergence of the filtration would be an additional hypothesis, not a consequence of the Maxwell example.

What fails when reducibility is incomplete

For a two-form $B$, the gauge variation $\delta_\Lambda B=d\Lambda$ is reducible because $\Lambda=d\rho$ acts trivially. A complex containing a ghost one-form but no ghost-for-ghost for $\rho$ leaves a nonzero cycle at positive antifield number. That residual homology is not a subtle global correction: it is the direct certificate that the proposed Koszul–Tate complex is not a resolution. Adding the missing stage repairs this algebraic defect locally, but still does not remove global zero modes, Gribov phenomena, or boundary cohomology. Those require their own domains and complexes.

The directional conclusion is therefore precise: regularity plus a complete tower of Noether and reducibility data gives a local resolution, and that resolution licenses the BRST cohomology calculation. Vanishing of a few low-degree groups does not conversely prove completeness of the full tower, nor does local acyclicity prove a global quotient is smooth.

Exercises

For Maxwell theory, verify that $\delta^2c^*=0$ without using the equations of motion.

Solution

Apply $\delta$ once more: $\delta^2c^*=-\partial_\mu\partial_\nu F^{\nu\mu}$. The two derivatives are symmetric in $\mu,\nu$, whereas $F^{\nu\mu}$ is antisymmetric, so the expression vanishes identically. This is a Noether identity, not an on-shell equality.

Show that omitting the ghost antifield $c^*$ prevents the Maxwell complex from resolving the equation-of-motion ideal.

Solution

The combination $\partial_\mu A^{*\mu}$ is $\delta$-closed because its image is $\partial_\mu\partial_\nu F^{\nu\mu}=0$. Without $c^*$ there is no generator whose $\delta$-image is this cycle, so positive-antifield homology survives.

References

• Barnich, Glenn, Friedemann Brandt, and Marc Henneaux. “Local BRST Cohomology in Gauge Theories.” Physics Reports 338 (2000): 439–569. DOI; Open PDF.
• Gomis, Joaquim, Jordi París, and Stuart Samuel. “Antibracket, Antifields and Gauge-Theory Quantization.” Physics Reports 259 (1995): 1–145. DOI; Open PDF.