Higher-Spin Gaps and Einstein-Regime Obstructions

Large $N$ can suppress bulk loops while an infinite tower of light higher-spin fields remains. Such a theory may be holographic and semiclassical, but it is not described over an improving energy window by Einstein gravity coupled to finitely many fields below the cutoff. The missing datum is a parametrically large higher-spin gap, controlled independently of $N$.

Required background. Sparse Spectra, Large Gaps, and Semiclassical Bulk Criteria separates central-charge, sparsity, and spin-gap requirements.

Helpful background. Weakly Broken Higher-Spin Symmetry supplies the CFT data. Unitarity Bounds and Null States explains why exactly conserved currents sit at a unitarity bound.

From current nonconservation to a bulk mass

In a unitary CFT with $d\ge3$, a symmetric-traceless primary of spin $s\ge1$ obeys $\Delta_s\ge s+d-2$. Saturation,

$$\Delta_s=s+d-2.$$

is equivalent to current conservation away from contact terms and corresponds to a massless spin-$s$ gauge field in AdS. For a symmetric-traceless field with $s\ge1$, the standard mass–dimension relation is

$$m_s^2L^2 = (\Delta_s+s-2) (\Delta_s-s-d+2).$$

When conservation is weakly broken, $\partial\cdot J_s\ne0$. Define the anomalous-dimension excess above the unitarity bound by

$$\gamma_s:=\Delta_s-(s+d-2)\ge0,$$

or equivalently $\Delta_s=s+d-2+\gamma_s$, and normalize the nonconservation equation as

$$\partial\cdot J_s=g_sK_{s-1}.$$

For unit-normalized operators in a vector large-$N$ theory, $g_s=O(N^{-1/2})$ commonly gives $\gamma_s=O(N^{-1})$ at each fixed spin. The normalization-independent statement compares the descendant norm with the two-point norm of $K_{s-1}$; Weakly Broken Higher-Spin Symmetry derives that boundary relation and its mixing qualifications. Substitution into the bulk dictionary gives

$$m_s^2L^2 =\gamma_s \bigl(2s+d-4+\gamma_s\bigr).$$

Thus, at fixed $s$, $\gamma_s\ll1$ makes the bulk field light; uniformly over a spin range one instead needs $\gamma_s(2s+d-4+\gamma_s)\ll1$. A light field is not an absent field. For $\gamma_s=\kappa_s/N+O(N^{-2})$ at fixed $s$,

$$m_s^2L^2 = \frac{(2s+d-4)\kappa_s}{N} +O(N^{-2}),$$

so every such fixed-spin field becomes massless as $N\to\infty$. The estimate is not automatically uniform when $s$ grows with $N$. Giombi and Kirilin 2016, § 1 and § 4 derive the weak-breaking scaling and apply it to the large-$N$ critical $O(N)$ model.

The general dictionary statement should be separated from a famous rigidity theorem. Maldacena and Zhiboedov 2013, § 1 concerns three-dimensional CFTs under hypotheses including a unique stress tensor, finite $C_T$, a finite number of primaries below any fixed dimension, and one exactly conserved higher-spin current. Within that scope, one current forces an infinite tower and current correlators of free-boson or free-fermion type. It is not an unrestricted theorem in every dimension.

An Einstein regime instead needs every spin-above-two single-particle field to decouple from the claimed window. The minimization includes every such primary: placing a light spin-four field inside the retained sector does not evade the obstruction. A physical rest-mass gap for symmetric-traceless fields must use the spin-dependent relation above; mixed-symmetry tensors and other Lorentz representations require their own mass–dimension dictionaries. A large dimension by itself can be caused by large spin. Depending on the argument, the boundary hypothesis may instead be a dimension gap $\Delta_{\mathrm{HS}}$ or a twist gap $\tau_{\mathrm{HS}}$. The preceding page’s distinction must be retained.

Matrix and vector examples

“Large $N$” fixes neither the gap nor the low-energy field content. The comparison below keeps the loop and higher-spin controls separate.

Representative large-N limits with different higher-spin spectra and claim ceilings

Boundary regime	Stress-tensor scale	Higher-spin spectrum	Licensed bulk reading
Planar 𝒩 = 4 SYM at large ’t Hooft coupling	CT ∼ N2	String and higher-spin mass scale MstringL ∼ λ1/4	Ten-dimensional type-IIB supergravity below the string scale; a lower-dimensional finite-field sector needs a consistent truncation
Planar 𝒩 = 4 SYM at zero or small λ	CT ∼ N2	Conserved tower at λ = 0; light, weakly broken tower for small λ	Planar loops are suppressed, but no parametrically separated Einstein window follows
Free O(N) model, singlet sector	CT ∼ N	Exactly conserved bilinear currents at s = 2, 4, 6, …, plus the scalar J0	Candidate classical higher-spin bulk, not a finite-field Einstein EFT
Critical O(N) model, singlet sector	CT ∼ N	γs = O(1/N) and ms2L2 = O(1/N) at each fixed even spin s ≥ 4	Weakly broken higher-spin regime; the tower becomes massless as N → ∞

Strongly coupled matrix example

For planar $\mathcal N=4$ super-Yang–Mills, define $\lambda=g_{\mathrm{YM}}^2N$ and take $N\to\infty$ at fixed $\lambda$ before sending $\lambda\to\infty$. In the standard D3-brane dictionary,

$$\frac{L^2}{\alpha'}=\sqrt\lambda, \qquad M_{\mathrm{string}}L = \frac{L}{\sqrt{\alpha'}} = \lambda^{1/4}.$$

Thus $1/N^2$ suppresses closed-string loops, while $\lambda^{-1/4}$ controls the separation between AdS energies and the string tower. Both limits are needed for a parametrically improving supergravity regime. The compact $S^5$ has radius $L$, so its full Kaluza–Klein tower is not lifted by large $\lambda$; a five-dimensional finite-field description is a consistent sector, not the complete low-energy spectrum. Aharony et al. 2000, § 3.2 review these scale relations and their top-down setting.

At $\lambda=0$, the same large-$N$ scaling of $C_T$ coexists with conserved higher-spin single traces. At small fixed $\lambda$, their planar anomalous dimensions are controlled by $\lambda$, not by $1/N$. Beisert et al. 2004, abstract and §§ 2–5 organize the free single-trace spectrum into higher-spin multiplets and compute representative planar one-loop anomalous dimensions after interactions break the higher-spin symmetry. This supplies an internal counterexample to the claim that matrix factorization or $C_T\sim N^2$ alone selects an Einstein regime.

Vector-model higher-spin example

In the vector rows, “single trace” means the singlet-bilinear analogue rather than a literal matrix trace. In the free three-dimensional $O(N)$ model,

$$\Delta_s=s+1, \qquad s=2,4,6,\ldots,$$

so every even-spin current is conserved and maps to a massless AdS$_4$ gauge field. In the critical model, the stress tensor at $s=2$ remains exactly conserved, while for each fixed even $s\ge4$ the unit-normalized nonconservation has the schematic form $\partial\cdot J_s=N^{-1/2}K_{s-1}+\cdots$. Hence $\gamma_s=\kappa_s/N+O(N^{-2})$ and the mass formula above applies. The limit suppresses bulk loops as $1/N$ while making the fixed-spin tower lighter. It therefore improves a candidate classical higher-spin description rather than producing a finite-field Einstein truncation.

Klebanov and Polyakov 2002, abstract and §§ 1–3 propose the critical vector model/type-A higher-spin duality, and Bekaert et al. 2022, § 2, pp. 4–8; § 3.1, pp. 8–9; § 4, p. 16 review the broader higher-spin framework and its open locality and quantum-completion questions. The spectral and correlator evidence is substantial, but it is not a proof of a unique finite-$N$ quantum bulk theory.

What the comparison licenses

The conventions are fixed $s$ before the large-$N$ limit, unit-normalized boundary operators, and AdS radius $L$. The observables are $C_T$, $\Delta_s$, $\gamma_s$, and normalized correlators. The matrix example has independent loop and string-scale controls, $N^{-2}$ and $\lambda^{-1/4}$; the vector example has loop and current-breaking controls of order $N^{-1}$, but no growing higher-spin mass gap. The inference is restricted to the displayed sectors and limits. Nonuniform large-spin behavior, Kaluza–Klein modes, operator mixing, pseudo-local field redefinitions, and the absence of a complete finite-$N$ higher-spin definition remain explicit uncertainties.

Causality constrains non-Einstein graviton couplings

Higher-derivative corrections to the graviton three-point coupling can cause a high-energy time advance. In the generic setting analyzed by Camanho, Edelstein, Maldacena, and Zhiboedov, new higher-spin physics must enter near the correction scale; their $D>4$ curvature-squared-only case leaves a special loophole involving a finite set of mixed-symmetry states, so an infinite tower is not an unconditional conclusion. Let $\hat\alpha_2:=\alpha_2^{\mathrm{CEMZ}}/L^2$ and $\hat\alpha_4:=\alpha_4^{\mathrm{CEMZ}}/L^4$ denote the dimensionless coefficients of the two non-Einstein on-shell graviton three-point structures, represented schematically by curvature-squared and curvature-cubed terms. In the large-$N$, large-gap CFT$_4$/weakly coupled AdS$_5$ setting of Camanho et al. 2016, § 3.2, pp. 18–19; § 5.4, p. 41; §§ 8–8.1, pp. 47–50, at leading order in the gravitational coupling and under their eikonal and asymptotic-causality assumptions, the parametric argument gives

$$|\hat\alpha_2|^{1/2} \lesssim\Delta_{\mathrm{HS}}^{-1}, \qquad |\hat\alpha_4|^{1/4} \lesssim\Delta_{\mathrm{HS}}^{-1},$$

Here $\Delta_{\mathrm{HS}}$ is the dimension of the lightest single-trace primary with $J>2$. The argument does not determine the order-one constants. These estimates constrain the non-Einstein on-shell $TTT$ structures; they are not bounds on arbitrary off-shell curvature coefficients or every Wilson coefficient. The detailed time-delay argument belongs to Regge Limits, Eikonal Scattering, and Causality.

Within the same large-$N$, large-gap four-dimensional $\mathcal N=1$ SCFT subclass, supersymmetry sets $\hat\alpha_4=0$ and $\hat\alpha_2\propto(a-c)/c$, giving the specialized estimate

$$\left|\frac{a-c}{c}\right| \lesssim \frac{1}{\Delta_{\mathrm{HS}}^2}.$$

The implication is one-way. A large gap suppresses the indicated non-Einstein structure, but $a=c$ does not prove a large gap. $\mathcal N=4$ super-Yang–Mills has $a=c$ throughout its conformal coupling family and nevertheless develops an exactly conserved higher-spin tower at the free point and a light tower nearby. Equality of one anomaly diagnostic therefore cannot replace a spectral measurement.

Hold the gap fixed while N grows

The loop and derivative expansions have different control parameters. Consider a definite family with

$$C_T(N)=c_TN^p, \qquad p>0, \qquad \Lambda_{\mathrm{HS}}L=G,$$

where $c_T$ and the physical higher-spin threshold $G$ are independent of $N$. At fixed $EL$, assume that the coupling-weighted light-species count stays $O(1)$. A fixed gap alone does not determine a matching coefficient, so make the adversarial family explicit: suppose the tower can be integrated out locally and its first nonzero matching coefficient obeys $c_q(N)\to c_q^{(\infty)}\ne0$. Then

$$\epsilon_{\mathrm{grav\,loop}}(E) \sim \frac{N_{\mathrm{light}}(E)}{C_T} (EL)^{d-1}, \qquad \epsilon_{\mathrm{HS}}(E) \sim c_q(N) \left(\frac{EL}{G}\right)^q, \qquad q>0.$$

Here $q$ is the first omitted power in the derivative expansion, and $\epsilon_{\mathrm{HS}}$ is the error from integrating out the higher-spin sector. The full derivative error is controlled by the lowest EFT threshold; the displayed second estimate equals it only when $\Lambda_{\mathrm{HS}}$ is the first cutoff. If the tower cannot be integrated out into a local derivative series, the alternative obstruction is stronger: its degrees of freedom or nonlocal effects must remain in the bulk description.

For a concrete check, take a four-dimensional boundary CFT with $c_T=1$, $N_{\mathrm{light}}=1$, $EL=2$, $G=20$, $q=2$, and $c_q(N)=1$. Then

$$\epsilon_{\mathrm{grav\,loop}} \sim \frac{8}{N^p} \longrightarrow0, \qquad \epsilon_{\mathrm{HS}} \sim \left(\frac{2}{20}\right)^2 =10^{-2}.$$

In this explicit family the fixed derivative error can be numerically useful, but it does not improve as $N\to\infty$. The lesson is that loop suppression does not force derivative or nonlocal effects to improve; it is not that a fixed gap alone fixes their coefficient. The strongest conclusion, once a bulk dictionary is independently established, is a classical bulk retaining a finite-gap stringy or higher-spin sector. The failed hypothesis is a physical higher-spin threshold with $\Lambda_{\mathrm{HS}}L\to\infty$; when the first omitted spin stays bounded, this is equivalently the failure of $\Delta_{\mathrm{HS}}\to\infty$. A finite low-spin Einstein EFT over a parametrically widening window is therefore not licensed. In the vector-model example the obstruction is sharper: the fixed-spin higher-spin masses tend to zero, and the interacting gauge structure does not admit a finite covariant low-spin truncation.

Common pitfalls

Treating large $C_T$ as a locality theorem. It controls a loop coefficient after the dictionary is established; it does not remove light higher-spin particles.

Using $a=c$ as a converse causality test. Equality removes one anomaly diagnostic but does not determine the spectrum.

Calling a fixed numerical gap parametric. A gap of 20 may support a useful finite window, but it does not improve along a family unless it grows with the family’s control parameter.

Exercises

1. Convert weak nonconservation into a mass

At fixed $s$, substitute $\Delta_s=s+d-2+\gamma_s$ into the spin-$s$ mass–dimension relation. What is the leading $m_s^2L^2$ for $\gamma_s\ll1$? For positive $\gamma_s$, also extract $m_sL$.

Solution — anomalous dimension to bulk mass

The second factor is $\gamma_s$ and the first is $2s+d-4+\gamma_s$. Therefore

$$m_s^2L^2 =\gamma_s(2s+d-4)+O(\gamma_s^2).$$

For positive $\gamma_s$ this gives

$$m_sL = \sqrt{(2s+d-4)\gamma_s}\,[1+O(\gamma_s)].$$

At fixed $s$, the mass vanishes continuously with the nonconservation parameter, so a weakly broken current produces a light—not absent—bulk field.

2. Test the failed converse

A four-dimensional large-$N$ CFT has $a=c$. Can the Camanho anomaly estimate alone establish an Einstein regime?

Solution — anomaly equality is not a gap measurement

No. The estimate says that a parametrically large higher-spin gap suppresses $(a-c)/c$ in the stated theory class; it does not say that a vanishing ratio forces a large gap. Weakly coupled $\mathcal N=4$ super-Yang–Mills is the counterexample: $a=c$, while conserved or nearly conserved higher-spin currents obstruct an Einstein window.

3. Compare three large-N limits

Family M has $C_T\sim N^2$ and $M_{\mathrm{string}}L\sim\lambda^{1/4}$. Family V has $C_T\sim N$ and, at every fixed even spin $s>2$, $\gamma_s=\kappa_s/N+O(N^{-2})$ with $\kappa_s>0$. Family F has $C_T\sim N^2$, $\Lambda_{\mathrm{HS}}L=20$, and a leading matching coefficient $c_q(N)\to1$. Which controls improve in each family, and what is the strongest bulk regime each limit licenses?

Solution — matrix, vector, and fixed-gap controls

In M, $N\to\infty$ suppresses closed-string or gravitational loops, while $\lambda\to\infty$ raises the string threshold as $\lambda^{1/4}/L$ and improves the derivative expansion below it. Taking only $N\to\infty$ does not raise that threshold.

In V,

$$m_s^2L^2 = \frac{(2s+d-4)\kappa_s}{N} +O(N^{-2}) \longrightarrow0$$

at every fixed $s$. The same $N\to\infty$ limit suppresses loops and restores a massless higher-spin tower. It licenses a candidate classical higher-spin regime, subject to the dictionary and completion caveats, not a finite-field Einstein EFT.

In F, $N\to\infty$ suppresses loops, but neither the threshold nor the nonzero matching coefficient improves. At fixed $EL$, the associated derivative error approaches a nonzero constant proportional to $(EL/20)^q$. The limit licenses at most a classical bulk with a fixed-gap sector over a fixed window, not a parametrically widening Einstein regime.

This is a necessary-condition test, not a sufficient construction. Continue next to Bulk Interaction Scaling and Effective Cutoffs for the combined EFT error budget, Higher-Spin and Vector-Model Holography for the non-Einstein duality class, and Regge Limits, Eikonal Scattering, and Causality for the dynamical consistency analysis.

Evidence cutoff. Theorem scope and cited literature were reviewed through 28 August 2026.

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

• Aharony, Ofer, Steven S. Gubser, Juan Maldacena, Hirosi Ooguri, and Yaron Oz. 2000. “Large N Field Theories, String Theory and Gravity,” Physics Reports 323, 183–386. Open PDF.
• Beisert, Niklas, Massimo Bianchi, José F. Morales, and Henning Samtleben. 2004. “Higher Spin Symmetry and $N=4$ SYM,” Journal of High Energy Physics 07, 058. Open PDF.
• Bekaert, Xavier, Nicolas Boulanger, Andrea Campoleoni, Marco Chiodaroli, Dario Francia, Maxim Grigoriev, Ergin Sezgin, and Evgeny Skvortsov. 2022. “Snowmass White Paper: Higher Spin Gravity and Higher Spin Symmetry,” arXiv:2205.01567. Open PDF.
• Camanho, Xian O., Jose D. Edelstein, Juan Maldacena, and Alexander Zhiboedov. 2016. “Causality Constraints on Corrections to the Graviton Three-Point Coupling,” Journal of High Energy Physics 02, 020. Open PDF.
• Giombi, Simone, and Vladimir Kirilin. 2016. “Anomalous Dimensions in CFT with Weakly Broken Higher Spin Symmetry,” Journal of High Energy Physics 11, 068. Open PDF.
• Klebanov, Igor R., and Alexander M. Polyakov. 2002. “AdS Dual of the Critical O(N) Vector Model,” Physics Letters B 550, 213–219. Open PDF.
• Maldacena, Juan M., and Alexander Zhiboedov. 2013. “Constraining Conformal Field Theories with a Higher Spin Symmetry,” Journal of Physics A: Mathematical and Theoretical 46, 214011. Open PDF.