Anomaly Cancellation and Positive-Hilbert-Space Existence

Anomaly cancellation removes identifiable obstructions to implementing a gauge symmetry, but it does not construct the resulting quantum theory. A positive physical Hilbert space requires more than a nilpotent BRST differential: one needs a representation, a compatible indefinite inner product before reduction, positivity after reduction, controlled limits, and a nontrivial local observable algebra. This page separates those obligations and applies the distinction to one Standard-Model generation.

Required background. The quantum master equation identifies perturbative obstruction classes; local anomaly descent tests infinitesimal gauge transformations; determinant-line holonomy tests large transformations; and Standard-Model anomaly cancellation supplies the chiral representations and charge sums.

Helpful background. Hilbert positivity and unitary evolution state the target positivity properties, while chiral gauge theories on the lattice show why a regulator and its continuum limit are separate questions.

BRST reduction and the positivity condition

Let $(\mathcal K,\langle\cdot,\cdot\rangle)$ be a Krein space carrying a gauge-fixed field algebra, and let $Q$ be a densely defined, Krein-symmetric operator of ghost number $1$. If

$$Q^2=0,$$

then the algebraic candidate for the physical state space is

$$\mathcal H_{\mathrm{phys}}^{\mathrm{alg}} =\ker Q/\operatorname{im}Q.$$

Nilpotence only makes this quotient meaningful. For it to inherit a Hilbert norm, one additionally needs

$$\langle\psi,\psi\rangle\ge 0 \quad(\psi\in\ker Q), \qquad \langle\psi,\psi\rangle=0 \Longleftrightarrow \psi\in\overline{\operatorname{im}Q},$$

followed by completion of the separated quotient. Domain invariance under observables and compatibility with the involution are also required. An arbitrary nilpotent operator on an indefinite space need not satisfy either positivity statement. Thus the implication is directional:

$$\text{BRST positivity hypotheses} \Longrightarrow \text{positive reduced representation},$$

not $Q^2=0\Longrightarrow$ positivity.

Dütsch and Fredenhagen prove a deformation-stability result for this structure in formal power series: if the free BRST system satisfies their positivity and completeness assumptions, its deformation retains a positive physical quotient order by order. They verify the additional hypotheses for QED in finite spatial volume with suitable boundary conditions, and explicitly find an obstruction for periodic boundary conditions of the massless free gauge field Dütsch and Fredenhagen 1998, §2.1 and Theorem 1, pp. 2–4. This is a conditional perturbative theorem, not a construction of a nonabelian chiral theory on infinite Minkowski space.

Cancellation tests for one Standard-Model generation

The exact first application is the one-generation Standard-Model anomaly calculation. With all fermions written as left-handed Weyl fields, one checks the $SU(3)^3$, $SU(3)^2U(1)_Y$, $SU(2)^2U(1)_Y$, $U(1)_Y^3$, and mixed gravitational–$U(1)_Y$ coefficients. One separately counts the $SU(2)$ doublets modulo two: three colored quark doublets plus one lepton doublet give an even number. These computations remove the corresponding four-dimensional perturbative and Witten $SU(2)$ anomalies.

What follows is a list of further obligations, not another cancellation identity. A proposed construction must supply a local regulator or renormalization prescription whose measure is defined over every relevant gauge-field sector; solve the quantum master equation to all orders in its stated expansion; remove ultraviolet and volume regulators with estimates strong enough to define correlation functions or local algebras; construct a positive representation of gauge-invariant observables; prove causal locality and unitary time evolution; control massless infrared behavior; and show that the limit is nontrivial. If the construction uses Euclidean fields, reflection positivity must hold on gauge-invariant observables strongly enough for reconstruction. If it uses a Krein-space BRST representation, the quotient and its completion must satisfy the conditions above.

Even pure four-dimensional Yang–Mills has no accepted axiomatic construction on $\mathbb R^4$ with the required mass gap. The official problem asks for a nontrivial theory with axiomatic properties and a positive gap Jaffe and Witten 2000, problem statement, printed pp. 141–142. Adding chiral matter does not make these analytic existence obligations automatic.

What cancellation proves—and what it does not

The useful conclusion is exact but limited. Vanishing local anomaly classes allows the master identity to be restored by admissible local counterterms, provided no higher-order obstruction appears. Trivial determinant-line holonomy removes the corresponding global phase obstruction. Together they establish consistency conditions on gauge symmetry. They do not select a vacuum, prove convergence of a formal series, produce a continuum measure, establish reflection positivity, or identify a mass gap.

An independent check keeps the algebraic and analytic columns separate. First evaluate every local anomaly coefficient and every known global holonomy for the chosen representation. Then ask whether a regulator, limit theorem, and positivity theorem have actually been supplied. A result in the first column cannot be copied into the second.

The adversarial failure is to infer the existence of an interacting chiral gauge theory solely from vanishing anomaly sums. One Standard-Model generation passes the listed cancellation tests, but those equalities contain no norm estimates, no continuum limiting sequence, and no proof that $\ker Q/\operatorname{im}Q$ completes to a positive Hilbert space. Cancellation is necessary for the proposed symmetry; it is not sufficient for nonperturbative existence.

Exercises

Construct a nilpotent operator whose cohomology is not positive.

Solution

Take $Q=0$ on $\mathbb C^2$ with indefinite form $\langle(x_1,x_2),(y_1,y_2)\rangle=\bar x_1y_1-\bar x_2y_2$. Then $Q^2=0$ and its cohomology is all of $\mathbb C^2$, which contains negative-norm vectors. Nilpotence alone therefore supplies no positivity.

Explain why cancellation of $U(1)_Y^3$ does not prove a continuum limit.

Solution

The cubic sum is a finite algebraic identity among charges. A continuum limit requires uniform estimates as a cutoff is removed, existence of limiting observables, and preservation of the required positivity and locality properties. None is encoded in the charge identity.

References

• Dütsch, Michael, and Klaus Fredenhagen. “Deformation Stability of BRST-Quantization.” AIP Conference Proceedings 453 (1998): 324–333. DOI; Open PDF.
• Jaffe, Arthur, and Edward Witten. “Quantum Yang–Mills Theory.” In The Millennium Prize Problems, 129–152. Providence, RI: American Mathematical Society, 2006; problem formulated 2000. Official PDF.