Holomorphy, Anomalies, and Exact Quantum Constraints

In four-dimensional $\mathcal N=1$ theories, holomorphy can turn symmetry and anomaly data into exact quantum information—but only after the object being constrained has been named precisely. A Wilsonian superpotential, a 1PI vertex, a holomorphic gauge coupling, and a canonically normalized coupling obey different statements. This chapter develops a reliable workflow for deciding what is exact, what remains an undetermined constant or branch choice, and what requires independent dynamical input.

Helpful background. Regulated Jacobians and measure variation explains how anomalous changes of variables become Ward identities. Beta functions and anomalous dimensions supplies the renormalization-group language used for canonical couplings. Fermion zero modes and index selection rules supplies the semiclassical counting used to test possible instanton terms.

Enter this chapter

The central object is a local chiral term in a Wilsonian action,

$$S_F=\int d^4x\,d^2\theta\;\mathcal F(\Phi_i,\lambda_A,\tau)+\text{h.c.},$$

where the dynamical chiral fields are $\Phi_i$, the masses and other couplings $\lambda_A$ may be treated as nondynamical chiral sources, and $\tau$ is a complexified gauge coupling. Chirality makes $\mathcal F$ holomorphic in these variables. Symmetries—including symmetries made formal by transforming the sources—restrict its allowed monomials. Anomalous symmetries still constrain it once the transformation of the path-integral measure, equivalently of the holomorphic scale, is included.

Those facts are powerful but not self-executing. Every exactness argument in this chapter keeps the following questions visible:

1. Which functional? Is the statement about a Wilsonian action at a nonzero scale, a 1PI effective action, or a vacuum expectation value?
2. Which holomorphic coordinates? Are fields and gauge couplings holomorphically or canonically normalized?
3. Which domain and branch? Are masses nonzero, is the theory gapped, and which branch of a fractional power is chosen?
4. Which symmetries survive the regulator? If a symmetry is anomalous, how does $\theta$ or $\Lambda$ transform?
5. What fixes normalization? Holomorphy and charges often determine a functional form only up to a constant. A weak-coupling calculation, instanton measure, decoupling relation, or other controlled limit must fix it.

An argument that does not answer these questions is a useful ansatz, not yet an exact result. Seiberg’s background-coupling formulation makes this logical structure especially clear Seiberg 1993, pp. 469–475.

Choose the controlling obligation

If the proposed statement depends on…	Begin with…	The decisive check is…
masses or couplings treated as chiral variables	Holomorphic couplings and background superfields	whether holomorphy, charges, limits, and regularity actually fix the function
“the superpotential is not renormalized”	Wilsonian, 1PI, and infrared scope	which effective action and which infrared assumptions are meant
an anomalous axial or R-rotation	R-symmetry, anomalies, and the holomorphic scale	the regulated measure and the induced transformation of $\Lambda$
an all-orders gauge beta function	Holomorphic and canonical couplings	the normalization and finite-renormalization scheme
a semiclassical instanton contribution	Instanton zero modes and selection rules	whether precisely two universal fermion modes remain for a superpotential
a dynamically generated superpotential	Nonperturbative superpotentials	symmetries plus zero modes, singularities, decoupling, branches, and one coefficient input
removing or restoring massive matter	Holomorphic decoupling and scale matching	matching the holomorphic scales and solving the correct F-term branch
a quantum relation among chiral operators	Quantum chiral rings and Konishi anomalies	the anomalous Ward identity, contact terms, and the chosen vacuum

The rows are not competing proof techniques. A robust nonperturbative result usually needs several of them. For example, the Affleck–Dine–Seiberg superpotential is first restricted by holomorphy, flavor symmetry, an anomalous axial symmetry encoded in $\Lambda$, and dimension. Its coefficient is then fixed in a semiclassically controlled case and propagated by holomorphic decoupling.

A common normalization

Unless a page states otherwise, examples use a four-dimensional Lorentzian $\mathcal N=1$ theory and the site’s $(+---)$ metric convention. For a simple gauge group $G$,

$$\tau=\frac{\theta}{2\pi}+\frac{4\pi i}{g_h^2}, \qquad b_0=3T(G)-\sum_iT(R_i),$$

with $\operatorname{tr}_{R}(T^aT^b)=T(R)\delta^{ab}$. The perturbative holomorphic scale is defined by

$$\Lambda_h^{b_0}=\mu^{b_0}e^{2\pi i\tau(\mu)}.$$

This definition fixes phases and normalizations for the examples; a finite redefinition of $\tau$ rescales $\Lambda_h$ and correspondingly changes convention-dependent coefficients. For $SU(N_c)$ SQCD with $N_f$ pairs $Q^i,\widetilde Q_i$, $T(\mathbf N_c)=1/2$ and $b_0=3N_c-N_f$.

The symbol $\Lambda_h$ is a holomorphic coordinate, not by itself a physical threshold. Physical masses depend on Kähler normalization and other real data. Likewise, “exact beta function” means an exact relation in a declared coupling scheme, not a scheme-independent function of an unnamed $g$.

Four kinds of evidence

Exact arguments become easier to assess when their ingredients are kept distinct.

Superspace perturbation theory. Supergraph $D$-algebra and locality prove the Wilsonian perturbative nonrenormalization theorem. They do not rule out nonperturbative F-terms or infrared-singular 1PI contributions.

Holomorphy and spurionic symmetry. These restrict allowed functions, sometimes to a one-parameter family. Regularity and asymptotic limits may remove further possibilities, but a singular limit cannot be used as if it were an ordinary Taylor point.

Anomalous Ward identities. A regulated measure supplies the missing transformation of $\tau$, $\Lambda_h$, or a composite operator. Calling an anomalous symmetry “broken” and discarding it loses precisely the information needed for many exact results.

Controlled dynamics. Instanton calculations on a completely Higgsed branch, weakly coupled thresholds, cluster decomposition, or gaugino condensation can fix constants that formal constraints leave free. The source of this input must be stated. The broad synthesis is reviewed in Intriligator and Seiberg 1996, §§ 2–4.

Failure tests

Before accepting an exact claim, try to make it fail in a nearby theory.

• Replace a Wilsonian action by the massless 1PI action. Nonlocal infrared terms can invalidate the naive proof.
• Perform a nonholomorphic field rescaling to canonical normalization. Its anomalous Jacobian changes the gauge coupling relation.
• Count instanton zero modes. Extra unlifted modes can force a higher F-term or correlation function instead of a superpotential.
• Add a large holomorphic mass and decouple a flavor. If scales and exact terms do not match, either the normalization, branch, or proposed function is wrong.
• Examine singular loci such as $\det M=0$. A formula valid on a Higgsed patch need not define a regular global coordinate there.
• Apply a finite coupling redefinition. If an allegedly scheme-independent higher-loop coefficient changes, the claim was stated too broadly.

These tests do more than catch mistakes: they identify the hypotheses that belong in the theorem.

Review the chapter

Consider $SU(N_c)$ SQCD with $0<N_f<N_c$, no tree-level superpotential, and meson matrix $M^i{}_j=Q^i\widetilde Q_j$.

1. Use dimension, flavor symmetry, and the anomalous axial transformation encoded by $\Lambda_h^{3N_c-N_f}$ to determine the possible nonperturbative superpotential up to a constant.
2. Explain why the direct one-instanton coefficient calculation is semiclassically controlled only for $N_f=N_c-1$.
3. Add a mass $mM^{N_f}{}_{N_f}$ and state the scale-matching relation to the $N_f-1$ theory.
4. Identify which parts of the result are Wilsonian and holomorphic, and which physical quantities still require Kähler data.

Solution outline

The unique invariant form on the patch $\det M\neq0$ is

$$W=C_{N_c,N_f} \left(\frac{\Lambda_h^{3N_c-N_f}}{\det M}\right)^{1/(N_c-N_f)}.$$

For $N_f=N_c-1$, Higgs expectation values can completely break the gauge group and lift all but two of the instanton’s fermion zero modes, so a dilute semiclassical calculation fixes $C$. For smaller $N_f$, an unbroken nonabelian subgroup makes a direct four-dimensional one-instanton interpretation inadequate; holomorphic decoupling propagates the answer instead. With the normalization above,

$$\Lambda_{h,N_f-1}^{3N_c-N_f+1} =m\Lambda_{h,N_f}^{3N_c-N_f}.$$

The result is an exact Wilsonian superpotential on a chosen fractional-power branch. Scalar masses, distances on moduli space, and the canonically normalized strong scale additionally depend on the Kähler potential and wavefunction factors.

References

• Kenneth A. Intriligator and Nathan Seiberg, “Lectures on Supersymmetric Gauge Theories and Electric–Magnetic Duality,” Nuclear Physics B Proceedings Supplements 45BC (1996), 1–28, arXiv, DOI.
• Nathan Seiberg, “Naturalness versus Supersymmetric Non-renormalization Theorems,” Physics Letters B 318 (1993), 469–475, arXiv, DOI.