Ward Identities, Weyl Anomalies, and Contact Terms

Bulk gauge, momentum, and Hamiltonian constraints become the boundary gauge, diffeomorphism, and Weyl Ward identities after the action is renormalized. Source terms, anomalies, and distributional contacts are part of those identities, not defects to be discarded. They provide the decisive consistency test that scalar, current, and stress-tensor counterterms were varied in one compatible scheme.

Required background. Renormalized one-point functions fixes the variational definitions. The trace Ward identity and Weyl anomaly supplies the boundary CFT statement. Helpful background. Trace-anomaly convention translation and momentum-space conformal Ward identities give independent checks.

First application. Derive the trace identity for a sourced scalar coupled to an AlAdS metric and isolate the logarithmic anomaly coefficient.

Constraints as boundary identities

Let the renormalized generating functional depend on boundary metric $g_{(0)ij}$, scalar source $\phi_{(0)}$, and gauge source $a_i$. Define

$$\delta S_{\mathrm{ren}} =\int\mathrm d^d x\sqrt{g_{(0)}} \left( \frac12\langle T^{ij}\rangle\delta g_{(0)ij} +\langle\mathcal O\rangle\delta\phi_{(0)} +\langle J^i\rangle\delta a_i \right).$$

An infinitesimal boundary diffeomorphism generated by $\xi^i$ gives Lie derivatives of every source. Integrating by parts and using invariance yields

$$\nabla_i\langle T^i{}_j\rangle =\langle\mathcal O\rangle\partial_j\phi_{(0)} +F^{(0)}_{ji}\langle J^i\rangle -a_j\nabla_i\langle J^i\rangle,$$

in the displayed source convention. If the background gauge symmetry is nonanomalous and the scalar is neutral, $\nabla_iJ^i=0$ and the last term vanishes. Charged sources replace it by the corresponding operator variation. The same equation follows from the radial momentum constraint, which tests the complete sign translation.

Gauge variation $\delta a_i=\partial_i\lambda$ gives current conservation or the declared anomaly. A bulk Chern–Simons term can generate a boundary consistent anomaly; adding a Bardeen counterterm redistributes it among currents but does not remove the invariant anomaly polynomial.

Weyl response and the scalar trace term

Under a boundary Weyl transformation,

$$\delta_\sigma g_{(0)ij}=2\sigma g_{(0)ij}, \qquad \delta_\sigma\phi_{(0)}=(\Delta-d)\sigma\phi_{(0)},$$

for a source coupled as $\int\sqrt g\,\phi_{(0)}\mathcal O$. Therefore

$$\langle T^i{}_i\rangle =(d-\Delta)\phi_{(0)}\langle\mathcal O\rangle +\mathcal A[g_{(0)},\phi_{(0)},a],$$

up to beta-function terms when the source is promoted to a running coupling. The local density $\mathcal A$ is the coefficient of the logarithmic counterterm. In even dimension it includes curvature and background-field invariants; in odd dimension there is no local bulk Weyl anomaly on a smooth closed boundary, although finite contact and boundary anomalies may remain.

For a resonant scalar, the logarithmic radial coefficient fixes the scalar contribution to $\mathcal A$. This is the promised application: compute the coefficient that replaces the singular Fefferman–Graham recursion, vary its local logarithmic counterterm under $\mu$, and insert it in the trace identity. The anomaly coefficient obtained from these three routes must agree.

Contact terms are required by differentiation

Differentiate the one-point Ward identity with respect to a source. For example, at vanishing scalar source,

$$\nabla_i\langle T^i{}_j(x)\mathcal O(y)\rangle =-\delta^{(d)}(x-y)\,\partial_j\langle\mathcal O(y)\rangle +\cdots,$$

with the exact derivative-of-delta structure determined by the operator’s tensor character and source convention. Such terms express the transformation of the insertion. Removing them because separated points obey a simpler equation makes the distributional identity false.

Finite local counterterms shift contact terms simultaneously across correlators. A curvature-scalar term changes $\langle\mathcal O\rangle$, $\langle T_{ij}\rangle$, and their mixed contacts. Scheme comparison must transform the whole family before testing a Ward identity.

Explicit breaking, spontaneous response, and running

A nonzero source creates explicit breaking through $(d-\Delta)\phi_{(0)}\langle\mathcal O\rangle$. A normalizable scalar profile with vanishing source can describe spontaneous response and does not produce that explicit term, although the state may break the symmetry. If the source is a running coupling $\lambda^a$, the trace relation is organized as

$$\langle T^i{}_i\rangle =\beta^a(\lambda)\langle\mathcal O_a\rangle +\mathcal A +\nabla_iV^i,$$

where improvement and virial terms require their own hypotheses. Identifying the radial derivative of one background profile with an exact field-theory beta function is an additional dictionary claim, not a consequence of this algebraic identity.

A compact verification suite

1. Derive each identity both from boundary source variation and from the corresponding radial constraint.
2. Differentiate once and retain every delta-function contact.
3. Change a finite local counterterm and verify that only the allowed local structures shift.
4. Compare the scale derivative of $S_{\mathrm{ren}}$ with the logarithmic Fefferman–Graham coefficient.
5. Turn off all sources and recover conservation; on a flat boundary recover the expected trace statement.

The result is a set of renormalized distributional identities in a stated scheme. It does not prove that arbitrary boundary source data admit a smooth bulk solution.

The anomaly and Ward-identity structure used here is derived in de Haro, Solodukhin, and Skenderis 2001; finite local counterterms still move the allowed contact terms.

The chapter overview contains the structure diagram and validity and failure diagram. They are embedded there once so that their shared chapter-level context is not repeated on every article.

For the chapter-wide comparison of assumptions, counterevidence, falsifiers, and claim ceilings, see the claim-domain table.

References

• Bianchi, M., Freedman, D. Z., and Skenderis, K. “How to Go with an RG Flow.” Journal of High Energy Physics 2001, 041 (2001). DOI. arXiv.
• de Haro, S., Solodukhin, S. N., and Skenderis, K. “Holographic Reconstruction of Spacetime and Renormalization in the AdS/CFT Correspondence.” Communications in Mathematical Physics 217 (2001): 595–622. DOI. arXiv.
• Henningson, M., and Skenderis, K. “The Holographic Weyl Anomaly.” Journal of High Energy Physics 1998, 023 (1998). DOI. arXiv.