Semiclassical Limit and Evidence

Recovering one Regge phase from one large-spin vertex is an important correspondence test, but it is not yet recovery of general relativity. A convincing semiclassical limit must survive sums over internal labels, refinement of the two-complex, coarse graining, constraint dynamics, and the construction of physical observables. At the evidence cutoff of 2026-08-20, loop and spin-foam calculations control several of these ingredients in restricted settings; none controls them all in generic four-dimensional gravity.

Required background. Variational and Symplectic Geometry supplies constrained phase space; Frames, Tetrads, and Spinors supplies tetrad and connection variables.

Helpful background. Flat-Spacetime Quantum Field Theory supplies correlation functions and renormalization; Einstein Equations supplies the target dynamics; Quantum Information supplies boundary states and coarse graining; and Quantum Gravity supplies common standards for correspondence and evidence.

What a semiclassical recovery must show

Let $\Psi_{q,K}$ be a boundary state peaked on an intrinsic three-geometry $q$ and extrinsic curvature $K$. A spin-foam amplitude should have a regime in which

$$Z_{\mathcal C}[\Psi_{q,K}] \sim \exp\!\left(iS_{\mathrm{EH}}[g_{\mathrm{cl}}]\right) \times\left(\det S_{\mathrm{EH}}''\right)^{-1/2} \left[1+\delta_{\mathrm{quantum}}\right],$$

as a schematic correspondence target. The determinant stands for the complete regulated one-loop factor, including gauge, ghost, zero-mode, and contour data; $\delta_{\mathrm{quantum}}$ must be demonstrably small in the stated regime. Its power of $\ell_{\mathrm P}/L$ is to be derived for the observable, not assumed from the vertex asymptotics. The correct boundary term, perturbative degrees of freedom, and composition law must also be recovered. The result should become insensitive to the auxiliary complex $\mathcal C$ after couplings and observables are coarse-grained. Canonically, coherent states must remain semiclassical under physical or relational evolution and reproduce the same observables.

These requirements separate four tests. Kinematical peakedness asks whether areas and holonomies have small relative fluctuations. Stationary phase asks whether a fixed amplitude contains the classical action. Propagation asks whether connected correlations carry the two graviton polarizations with the right long-distance tensor structure. Continuum recovery asks whether refinement produces universal, regulator-independent observables. Passing an earlier test does not entail a later one.

Controlled results on fixed structures

For nondegenerate Lorentzian Regge boundary data, the large-spin EPRL four-simplex amplitude contains conjugate phases $\exp(\pm iS_{\mathrm R})$ with a calculable power-law prefactor Barrett et al. 2010, §6.4, Eq. (53). This identifies the expected discrete classical action at a stationary point. Degenerate or merely vector-geometric boundary data have different asymptotics, and the paired orientations can produce a cosine rather than a single causal phase.

Boundary two-point calculations use coherent spin networks peaked on a flat four-simplex and expand a one-vertex amplitude at large spin. With a carefully chosen boundary state, they reproduce the expected leading tensorial structure and distance scaling of a graviton propagator in that restricted setting Bianchi, Magliaro, and Perini 2009, §§3–5. The background, boundary covariance, one-vertex truncation, and large-distance expansion are essential. The calculation is not a gauge-invariant scattering amplitude and does not include a controlled sum over refined foams.

Canonical complexifier and weave-type states can be peaked on selected classical connections and triads, while symmetry-reduced loop cosmology produces controlled effective evolution in specified models. Those results test semiclassical kinematics or reduced dynamics, not the generic full-theory Hamiltonian. Horizon state counting likewise produces an area-proportional leading term under declared boundary conditions and Barbero–Immirzi/state-counting prescriptions; its thermodynamic interpretation belongs to the dedicated semiclassical and black-hole volumes, and it does not establish bulk continuum dynamics.

A sum-over-spins adverse check

Suppose a fixed internal face contributes only a Regge phase $e^{ij\varepsilon}$, where $\varepsilon$ is a deficit angle. Summing integer labels symmetrically through $N$ gives the normalized kernel

$$K_N(\varepsilon)=\frac{1}{2N+1} \sum_{j=-N}^{N}e^{ij\varepsilon} =\frac{\sin[(N+\tfrac12)\varepsilon]} {(2N+1)\sin(\varepsilon/2)}.$$

For every $N$, $K_N(0)=1$. Away from multiples of $2\pi$, its envelope shrinks like $(2N+1)^{-1}$, and its central peak narrows like $N^{-1}$. Thus a fixed-$j$ phase permits arbitrary curvature, while an unrestricted coherent sum can concentrate on a much smaller set. Face amplitudes, label ranges, quantum corrections, boundary widths, and refinement determine whether that concentration is physical or an artifact. Replacing the sharp cutoff by a Gaussian weight changes the kernel to a periodic Gaussian and changes the width.

This toy sum is not the EPRL partition function. It is an adverse check on a common inference: finding $e^{iS_{\mathrm R}}$ at a fixed vertex cannot by itself show that the internal-spin sum admits generic curved solutions. The so-called flatness issue and related order-of-limits questions are discussed with the necessary qualifications in Perez 2013, §§13.2–13.4.

Results and open questions as of 2026-08-20

Approach or position	Strongest controlled result and source	Regime, adverse evidence, or unresolved step
Single EPRL four-simplex asymptotics	Nondegenerate Lorentzian Regge data yield conjugate Regge-action phases; Barrett et al. 2010, §6.4, Eq. (53)	Fixed complex and large spins; paired orientations, degenerate sectors, and internal sums remain
Boundary graviton two-point function	Correct leading long-distance tensor behavior for a selected coherent boundary state; Bianchi, Magliaro, and Perini 2009, §§3–5	One vertex, chosen background and gauge-sensitive observable; no controlled refinement
Spin-foam coarse graining	Renormalization frameworks and tensor-network calculations find nontrivial flows and fixed structures in simplified models; Dittrich 2017, §§3–5	Truncated representation data and model-dependent moves; no demonstrated generic four-dimensional Einstein universality class
Canonical coherent and reduced sectors	Classical expectation values and controlled evolution exist for selected states or symmetry reductions; Ashtekar and Lewandowski 2004, §§V–VII	Physical Hamiltonian, generic stability, and equivalence to covariant amplitudes are unresolved
Horizon state counting	Leading area scaling can be obtained under isolated-horizon boundary conditions; Ashtekar and Lewandowski 2004, §VIII	Depends on quantization/counting choices and probes a special boundary sector, not bulk continuum propagation

The permissive position reads the agreement of independent kinematical spectra, vertex asymptotics, boundary correlations, reduced sectors, and coarse-graining programmes as a converging semiclassical picture whose missing steps are technical. The conservative position accepts each calculation in its stated regime but withholds a claim of recovered general relativity until the same framework controls refinement, physical observables, constraint closure, and radiative corrections. A stronger skeptical position treats unresolved internal sums or ultralocal dynamics as possible structural failures. Parallel coverage does not make these positions equally supported: the fixed-complex results are firmer than the extrapolation to a continuum theory.

No calculation in the matrix demonstrates recovery of generic four-dimensional general relativity both beyond a fixed discretization and beyond symmetry reduction. A decisive positive result would be a regulator-independent continuum family with the Einstein effective action, two propagating polarizations, correct constraint relations, and matching observables across canonical and covariant descriptions. A decisive adverse result would be a stable extra mode, wrong long-distance correlation, or refinement-independent obstruction.

The conclusions could change after a replicated four-dimensional refinement flow reaches a controlled critical surface, a theorem relates the canonical physical inner product to a converged foam sum, a larger calculation changes the flatness or propagation conclusions, an observable discriminates the programme empirically, or evidence after 2026-08-20 changes the comparison.

Exercises

Derive the expression for $K_N(\varepsilon)$ and estimate the first zero around $\varepsilon=0$. What does the result permit—and not permit—you to infer about a spin-foam continuum limit?

Solution

The geometric series gives

$$\sum_{j=-N}^{N}e^{ij\varepsilon} =e^{-iN\varepsilon} \frac{1-e^{i(2N+1)\varepsilon}}{1-e^{i\varepsilon}} =\frac{\sin[(N+\tfrac12)\varepsilon]}{\sin(\varepsilon/2)}.$$

Dividing by $2N+1$ gives $K_N$. Its first nonzero numerator root is $\lvert\varepsilon\rvert=\pi/(N+\tfrac12)$, so the central width scales as $N^{-1}$. This proves sensitivity of the label sum to the deficit phase under the stated sharp, uniform measure. It does not determine the EPRL face measure, complex saddle structure, boundary state, refinement flow, or order of limits, and therefore cannot prove flatness or a viable continuum limit for the full model.

References

• Ashtekar, Abhay, and Jerzy Lewandowski. “Background Independent Quantum Gravity: A Status Report.” Classical and Quantum Gravity 21 (2004): R53–R152. Open manuscript.
• Barrett, John W., Richard J. Dowdall, Winston J. Fairbairn, Frank Hellmann, and Roberto Pereira. “Lorentzian Spin Foam Amplitudes: Graphical Calculus and Asymptotics.” Classical and Quantum Gravity 27 (2010): 165009. Open manuscript.
• Bianchi, Eugenio, Elena Magliaro, and Claudio Perini. “LQG Propagator from the New Spin Foams.” Nuclear Physics B 822 (2009): 245–269. Open manuscript.
• Dittrich, Bianca. “The Continuum Limit of Loop Quantum Gravity: A Framework for Solving the Theory.” In Loop Quantum Gravity: The First 30 Years (2017): 153–179. Open manuscript.
• Perez, Alejandro. “The Spin-Foam Approach to Quantum Gravity.” Living Reviews in Relativity 16 (2013): 3. Open manuscript.