research32 min studyLesson 707 in Physics

Slavnov–Taylor identities

Translate BRST invariance into constraints on full Green functions and renormalization constants.

Before you begin

On this page

The question for today

What you’ll understand

Explain the change-of-variables origin of Slavnov-Taylor identities.
Check a transverse two-point tensor without claiming diagram-by-diagram invariance.
Apply a ghost-gauge-vertex renormalization relation to first order.

A gauge-fixed quantum calculation contains unphysical fields, so its consistency is encoded in relations among their correlation functions. Slavnov-Taylor identities express BRST symmetry in that language. They connect propagators, vertices, and ghost contributions rather than declaring every individual diagram invariant. Their power is restrictive: counterterms and approximations must fit a common identity. Their validity also depends on the quantum measure and on the absence or cancellation of the relevant gauge anomaly.

Change variables in the functional integral

Apply an infinitesimal BRST change of variables inside a regulated path integral. If the action and measure are invariant and boundary terms vanish, the integral of a BRST variation vanishes. With sources present, varying the source term gives identities involving insertions of nonlinear field transformations. Differentiating with respect to sources then relates higher correlation functions. The identities are quantum statements derived from a measure, not a replacement for specifying that measure.

Because the transformations are nonlinear, their expectation values cannot generally be replaced by the same polynomial evaluated on mean fields. Introduce external sources K and L for the BRST variations of the connection and ghost, respectively. These sources track composite insertions. After a Legendre transform, the identity becomes a nonlinear functional constraint on the one-particle-irreducible effective action, with graded derivative ordering fixed consistently.

\langle s\mathcal O\rangle=0,\qquad \mathcal S(\Gamma)=\int d^4x\left(\frac{\delta\Gamma}{\delta K_\mu^a}\frac{\delta\Gamma}{\delta A_\mu^a}+\frac{\delta\Gamma}{\delta L^a}\frac{\delta\Gamma}{\delta c^a}+b^a\frac{\delta\Gamma}{\delta\bar c^a}\right)=0

This conventional Slavnov functional uses consistently ordered graded derivatives and sources coupled to sA and sc; matter terms are omitted for pure Yang-Mills theory. The expectation identity assumes an invariant measure and no integration-boundary contribution.↔ Scroll wide equations horizontally.

Use a two-point consequence carefully

In an anomaly-free Lorentz-covariant perturbative treatment, the gauge-field self-energy is transverse after all required contributions and counterterms are included. In Euclidean momentum its tensor structure is a scalar function multiplying k squared times the unit tensor minus the momentum outer product. Transversality constrains the structure but does not make that scalar independent of gauge parameter or subtraction scheme. The full propagator still has a gauge-fixed longitudinal sector.

The non-Abelian vertex identities generally involve ghosts and composite kernels. They are therefore more involved than a naive Abelian replacement asserting that every vertex contraction is merely a difference of inverse propagators. Omitting ghost loops or truncating only one member of an identity can generate an apparent contradiction. A numerical approximation should identify which identity it preserves and how residual violations are monitored, rather than labeling any approximately transverse result fully gauge consistent.

k_\mu\Pi^{ab}_{\mu\nu}(k)=0,\qquad \Pi^{ab}_{\mu\nu}=\delta^{ab}(k^2\delta_{\mu\nu}-k_\mu k_\nu)\Pi(k^2)

This is the transverse perturbative vacuum-polarization form in the stated covariant setting. The scalar Pi is dimensionless in four dimensions and can remain gauge and scheme dependent.↔ Scroll wide equations horizontally.

Constrain counterterms across sectors

Write bare gauge and ghost fields using their respective wave-function factors and define a coupling factor in one consistent renormalization scheme. The ghost-gauge vertex factor must equal the product of the coupling factor, one gauge-field square root, and the two ghost square roots. Slavnov-Taylor consistency relates the coupling inferred this way to that inferred from other vertices. The relation does not say all vertex and field factors are numerically identical.

Order-by-order expansion makes this constraint easy to audit. At first order, products become sums of coefficients and inverse factors subtract their coefficients. Higher orders require the corresponding cross terms; inserting finite factors into a first-order result without labeling the approximation mixes orders. If a regulator breaks the identity, compatible counterterms may restore it when there is no true anomaly. A genuine uncancelled gauge anomaly is an obstruction to this quantum construction, not merely an inconvenient gauge choice.

Z_{1c}=Z_g Z_A^{1/2}Z_c,\qquad Z_g=\frac{Z_{1c}}{Z_A^{1/2}Z_c}

Z-one-c denotes the ghost-gauge vertex factor, Z-c the common ghost and antighost wave-function factor, and Z-A the gauge-field factor. All are dimensionless and refer to the same scheme and normalization.↔ Scroll wide equations horizontally.

Work it through

Audit a first-order counterterm relation

Let a small dimensionless bookkeeping parameter u give $Z_A=1+2u$ , $Z_c=1+3u$ , and $Z_{1c}=1+5u$ to first order. Find Z-g to that order and evaluate the resulting estimate at $u=0.01$ .

The square root of the gauge-field factor is $Z_A^{1/2}=1+u+O(u^2)$ . Multiplying by the ghost factor gives $1+4u+O(u^2)$ .
Its inverse is $1-4u+O(u^2)$ . Thus the vertex relation gives $Z_g=(1+5u)(1-4u)+O(u^2)=1+u+O(u^2)$ . The coefficient is five minus one minus three.
At $u=0.01$ , the first-order estimate is 1.01. This is not the exact finite ratio of the three displayed linear factors, since that ratio includes uncontrolled higher-order terms when interpreted as a perturbative prediction.

The result

The common coupling factor has first-order coefficient one and estimated value 1.01 at the supplied bookkeeping parameter.

The living notebook

Explore the model

Interactive lab

Interactive lab · function

Size of a quadratic remainder

Reset and set quadratic coefficient one and position 0.1; predict y equal to 0.01. Compare a fresh Reset run at coefficient one and position 0.2; predict 0.04. This illustrates quadratic scaling of a possible omitted-order term, without determining its coefficient. The lab is not a Slavnov-Taylor checker, loop-integral evaluator, or renormalization calculation.

y = ax²TangentSigned integral

y at chosen x: 0.608
Tangent slope · 2ax: 1.35
Signed area · ax³/3: 0.182

Coefficient a0.75Position x0.9

t = 0 s

This lab uses only y = ax². Playback traces the curve within the fixed integration interval from 0 to x; it does not estimate numerical error or change the chosen point. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forA first-order expansion must retain a stated remainder; a simple scaling analogy does not estimate quantum counterterms.

Pause & reason

Check your understanding

Interactive checks need JavaScript. You can still read the explanation.

Read the checkpoint explanation

The identity constrains the completed correlation function. Individual diagrams and the scalar self-energy function need not be gauge independent, and the statement does not establish strong-coupling spectral properties.

Your turn

Try it yourself

For one fixed color take Euclidean $k=(M,0,0,0)$ and scalar $\Pi(k^2)=0.2$ . Write the self-energy diagonal and contract with k. Compare the incorrect ansatz $0.2M^2\delta_{\mu\nu}$ .

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

The momentum outer product subtracts the temporal entry in the transverse tensor; retain its factor of k squared.

Compare with the solution

The valid diagonal is $(0,0.2,0.2,0.2)M^2$ . Multiplication by k gives the zero vector, as required. The incorrect ansatz has all four diagonal entries $0.2M^2$ and contracts to $(0.2M^3,0,0,0)$ , violating transversality. This is an algebraic consistency check of a supplied two-point structure, not a computation of its scalar coefficient or a proof of every higher vertex identity.

Keep the idea

Take it with you

Slavnov-Taylor identities follow from a quantum change of variables with explicit measure assumptions.
They constrain complete correlation functions rather than individual diagrams.
Renormalization factors must satisfy one compatible identity across fields and vertices.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.