When scale symmetry becomes local conformal symmetry in four dimensions

← All digests

Family 282 · Mathematical physics
Importance 5/5 · Provisional editorial rating
Sources consulted 2026-10-25

1. The question

A scale-invariant theory looks the same when all lengths are enlarged by a common factor. Conformal symmetry is stronger: it also permits position-dependent changes of scale that preserve angles locally. Why should uniform rescaling force this larger symmetry?

In quantum field theory, the answer depends on which local fields and observables belong to the theory. A formal transformation or a useful two-point function does not automatically produce a physical local operator with well-defined products.

Family 282 claims that scale symmetry implies local conformal symmetry for a specified class of four-dimensional unitary positive-energy theories. The central construction improves the stress tensor to make its trace vanish while preserving translation and Lorentz charges. Its conclusion is a local action through Ward identities. Integrating that action into a global conformal symmetry of the entire bounded observable net remains a separate question.

2. Background and essential definitions

Currents and the trace obstruction

Work in four-dimensional spacetime with metric $\eta=\operatorname{diag}(1,-1,-1,-1)$. The stress tensor $T_{\mu\nu}$ describes energy and momentum. It is symmetric and conserved: $\partial^\mu T_{\mu\nu}=0$. Integrating suitable components across a time slice gives translation and Lorentz generators.

A conserved dilation current has the form

$$D_\mu=x^\nu T_{\mu\nu}-V_\mu, \qquad T^\mu{}_{\mu}=\partial^\mu V_\mu.$$

The local vector field $V_\mu$ is the virial field. The second identity follows from conservation of $D_\mu$. Scale invariance therefore permits a nonzero stress-tensor trace if it is a divergence.

An improvement adds derivatives of local fields to the stress tensor without changing its physical charges. If a local scalar $L$ satisfies $T^\mu{}_{\mu}=\Box L$, where $\Box=\partial^\mu\partial_\mu$, then in four dimensions

$$t_{\mu\nu}=T_{\mu\nu} +\frac13(\partial_\mu\partial_\nu-\eta_{\mu\nu}\Box)L$$

is conserved and traceless. The challenge is not this algebraic calculation; it is establishing that a suitable $L$ actually belongs to the physical local theory.

Local Ward identities

A conformal Killing vector $X$ generates an infinitesimal spacetime transformation whose metric change is proportional to the metric. For a symmetric conserved traceless tensor, the corresponding current is $j_\mu=X^\nu t_{\mu\nu}$.

A Ward identity describes how the flux of that current around a field insertion generates its symmetry transformation. “Unbroken” means the vacuum and correlation functions satisfy the corresponding symmetry identities. Local Ward identities concern compact regions and insertions; they do not by themselves construct globally defined group transformations on every bounded observable.

Fields and bounded local operations

Point fields are distributions: they become operators after smearing with smooth test functions. These operators are often unbounded, so their domains matter. A bounded local net assigns a weakly closed operator algebra $\mathcal M(O)$ to each spacetime region $O$, with commuting algebras at spacelike separation, or the appropriate statistics twist for fermions.

A closed field operator is affiliated with a local algebra when its spectral or graph data are controlled by that algebra. This ties the unbounded field to the bounded local operations. Weak commutation of fields on a small domain does not automatically establish the stronger bounded-algebra locality used in this manuscript.

3. Development history and previous partial results

In two dimensions, Zamolodchikov’s $c$-theorem and Polchinski’s enhancement result established a powerful route from scale to conformal symmetry under suitable assumptions. Four dimensions require different arguments.

Callan, Coleman and Jackiw developed the improved energy-momentum tensor. Perturbative local renormalization-group identities later gave substantial scale-to-conformal results. Flavor-current contributions matter: a cycle in ordinary beta functions need not describe a nonconformal physical theory. Extending perturbative positivity to arbitrary coupling is another issue.

Komargodski and Schwimmer’s four-dimensional $a$-theorem uses anomaly matching and a positive dilaton dispersion relation between conformal endpoints. Luty, Polchinski and Rattazzi developed this approach to constrain possible asymptotic theories and establish conformal behavior in perturbatively controlled regimes. Dymarsky, Komargodski, Schwimmer and Theisen sharpened the vanishing-amplitude argument.

A remaining distinction is between on-shell probe scattering and an off-shell local operator identity. Contact and semi-local terms can survive when scattering amplitudes vanish. Bzowski and Skenderis emphasized their role in carrying scale anomalies, while work on generalized-free traces isolated further alternatives rather than eliminating every improvement problem.

Yonekura proposed constructing a scalar by applying an inverse wave operator to the trace, with locality relying on stated analyticity and crossing assumptions. Sachs treated a restricted class of metric-couplable Lagrangian theories. The present manuscript aims to reconstruct the scalar within a prescribed physical net, without importing an independent scattering or crossing hypothesis.

Its modular methods also have predecessors. Faulkner, Leigh, Parrikar and Wang related deformations of half-space modular Hamiltonians to horizon stress integrals; Casini, Testé and Torroba developed null-plane modular relations. The new manuscript supplies the specific flux limits, domains and reconstruction statements it uses under its operational assumptions.

4. What the new manuscript claims

The theorem assumes a unitary, positive-energy theory on $\mathbb R^{1,3}$, with a Poincaré- and continuously scale-invariant vacuum. Original fields and adjoints generate the Hilbert space and obey Wightman distribution, covariance, spectrum and locality conditions on compatible invariant domains. Uniqueness of the invariant vacuum vector is not assumed.

Scaling dimensions form a real discrete spectrum, bounded below, with no finite accumulation and finite multiplicities after Lorentz components are counted. Every original field is a finite sum of homogeneous scaling fields. Infinite sums of scaling components are not silently admitted as new field labels.

The theory includes a fixed covariant causally commuting bounded local net. Original fields and adjoints have compatible closed affiliated realizations in that same net. A physical stress tensor and virial field obey local flux Ward identities, and the physical dilation current generates the stated scale transformations.

New fields can be admitted only after establishing arbitrarily sharp localization, every finite mixed product with original fields and adjoints, a compatible common invariant domain containing the original vacuum-word domain, and affiliation in the original net. This is an admission criterion, not an assumption that the desired scalar already exists.

Under these hypotheses, the manuscript selects the Hermitian dimension-four stress component $T^{(4)}$ and dimension-three virial component $V^{(3)}$, preserving the affine Ward charges, and constructs a Hermitian physical scalar $L$ of dimension two such that

$$\partial^\mu V_\mu^{(3)}=\Box L, \qquad t_{\mu\nu}=T_{\mu\nu}^{(4)} +\frac13(\partial_\mu\partial_\nu-\eta_{\mu\nu}\Box)L.$$

The improved tensor is symmetric, conserved and traceless, with unchanged translation and Lorentz charges. Its conformal currents satisfy unbroken local Ward identities on the physical field algebra, extending the existing Poincaré and dilation actions.

The Hilbert space and bounded net are preserved. The common field domain is enlarged through reconstructed vacuum words, but the theorem does not identify every newly affiliated extension with every previously chosen closed operator on its larger domain.

The full massless free scalar theory supplies an example: its square field is available and gives the familiar stress improvement. Restricting to a shift-invariant field algebra can remove the needed virial current and square field. The local-current and intrinsic field-membership assumptions are therefore substantive.

5. Why the advance matters

Importance: 5/5, provisional. The family addresses a central symmetry-enhancement problem while explicitly confronting the passage from correlation information to physical local operators. The claimed scalar is reconstructed in the original bounded net, rather than placed in an unspecified larger theory.

The distinction between local and global conclusions is equally important. A traceless stress tensor with conformal Ward identities gives detailed infinitesimal constraints on observables, but a global group action requires additional integration work. Preserving that boundary makes the theorem’s scope assessable.

The rating reflects the significance of the stated enhancement theorem if correct. It does not remove its operational hypotheses or independently verify the long analytic reconstruction.

6. The main idea

Extract information from null boundaries

Modular theory associates canonical operator flows with local algebras and a vacuum. The manuscript uses wedge modular data and inclusions associated with null cuts to obtain positive generators. A bounded-profile Ward identity identifies their response with physical stress-tensor flux.

Comparing two null horizons isolates a possible critical scalar tail. A dispersion comparison and logarithmic test force its positive leakage coefficient to vanish. This improves the scalar spectral estimate needed for the subsequent construction.

Distinguish a vacuum wave from a field

Applying a field to the vacuum gives a vector-valued distribution, or vacuum wave. Dividing trace polarization by invariant mass squared suggests a scalar wave, but that operation alone does not define a local field with arbitrary products.

The manuscript first identifies actual modular actions on time cones and local diamonds. These actions extract a dimension-two scalar wave from the virial field with real localization and bounded approximants in the original net. Short-product Ward identities and positivity on pair states then identify it with the trace-derived scalar wave.

Several actions play different roles: an auxiliary conformal action organizes two-point spectral comparisons, actual modular transformations act on the physical Hilbert space, and a pair-space conformal representation controls residual terms. They are not treated as interchangeable global symmetries.

Reconstruct all products before improving the tensor

Bounded approximants generate analytic products of every finite length. Differential identities, causal wave equations and controlled boundary limits establish the required distributions, spectral support and mixed orderings with original fields.

These products supply a common invariant domain with compatible adjoints. Commutation with the original bounded commutants establishes affiliation in the original local algebras. Only after this work does $L$ become a physical field and the improvement formula become an operator identity with the claimed Ward action.

This reconstruction is the main conceptual burden: a correctly localized two-point object is only the beginning of a local quantum field.

7. Scope, limitations, and open questions

The theorem is not an unconditional statement about every four-dimensional scale-invariant theory. Finite scaling support, the discrete finite-multiplicity dimension spectrum, compatible bounded locality and a physical local dilation current are explicit assumptions.

No Haag duality, split property, phase-space bound, scattering theory or a priori conformal action is assumed. That absence does not erase the assumptions that are present. In gauge descriptions, construction occurs in the chosen gauge-invariant local algebras.

The conclusion is local conformal Ward symmetry. Global conformal covariance of the whole bounded net remains open within the stated result. Other field classes, infinite scaling sums or differently chosen observable nets require separate analysis rather than an automatic application.

8. Sources and further reading

The improvement and dilaton papers linked above provide central predecessors. Polchinski’s Scale and conformal invariance in quantum field theory gives the earlier enhancement context. The source below states the operational hypotheses and local reconstruction theorem, whose claims are reported as stated.

Source manuscripts

• M1. OpenAI, Scale and conformal symmetry in four-dimensional operational quantum field theory (September 26, 2026).

Sources consulted on 25 October 2026 at repository revision 1581d4c. Download the manuscript citations.