Spin-Foam and EPRL Amplitudes: Covariant-Dynamics Proposals

Spin foams assign amplitudes to labeled two-complexes interpolating between spin-network boundaries. The EPRL construction starts from four-dimensional BF theory and imposes simplicity constraints so bivectors can describe tetrad gravity. A single-vertex Regge asymptotic is an important semiclassical check, not a proof of the refined continuum Einstein path integral.

Required background. Ashtekar–Barbero Variables and Connection Dynamics supplies simplicity data; Loop-Quantum-Gravity Kinematics and Spin Networks supplies boundary states.

Helpful background. Stationary Phase, Coalescing Saddles, and Stokes Geometry supplies asymptotics; Hamiltonian Constraints and Quantum Dynamics supplies the canonical comparison.

From BF theory to EPRL labels

On a fixed two-complex $\mathcal C$,

$$Z_{\mathcal C} =\sum_{\{j_f,i_e\}} \prod_fA_f(j_f)\prod_eA_e(j_f,i_e) \prod_vA_v(j_f,i_e).$$

For Euclidean signature and $0<\gamma<1$, linear simplicity maps an $SU(2)$ spin to

$$j_f^\pm=\frac{1\pm\gamma}{2}j_f$$

in $Spin(4)\cong SU(2)_+\times SU(2)_-$. Integrality restricts admissible $j_f$ for fixed $\gamma$. Face weights and measure factors are additional model choices.

First application: one four-simplex

Choose coherent boundary data $(j_f,\mathbf n_{ef})$ satisfying closure and shape matching for a nondegenerate geometric four-simplex. The EPRL vertex can be written as group integrals of coherent-state matrix elements,

$$A_v(\lambda j_f,\mathbf n_{ef}) =\int\prod_e dg_e\, \exp\!\left[\lambda S(g_e;j_f,\mathbf n_{ef})\right].$$

Stationary phase at large common spin $\lambda$ gives

$$A_v\sim\lambda^{-12} \left[ N_+e^{i\lambda S_{\rm Regge}} +N_-e^{-i\lambda S_{\rm Regge}} \right],$$

up to convention-dependent phases and additional critical sectors. The two terms correspond to opposite orientations. Barrett and collaborators derived this nondegenerate asymptotic Barrett et al. 2010; the EPRL amplitude was introduced in Engle et al. 2008.

From a vertex to a theory

The sum over internal spins may diverge, and refinement changes the two-complex. Recovering gravity requires control of degenerate configurations, orientation sectors, radiative corrections, face amplitudes, refinement/renormalization, and boundary observables. Regge behavior at fixed complex tests the phase, not continuum universality.

Adversarial control: alter weights and refine

Change $A_f(j)$ while leaving the one-vertex stationary phase unchanged. Multi-vertex divergences and continuum flows change. Include degenerate boundary data and they can scale as strongly as the desired saddle. Refine a four-simplex while holding physical boundary geometry fixed; stable observables, not the original vertex, must converge.

The EPRL construction is a concrete covariant dynamics proposal with a successful geometric saddle. Its physical inner product, continuum limit, and unique relation to canonical Hamiltonian dynamics remain open tests.

The chapter overview compares these assumptions and limitations.

For the chapter-wide comparison of assumptions, counterevidence, falsifiers, and claim ceilings, see the claim-domain table.

References

• Barrett, John W., Richard J. Dowdall, Winston J. Fairbairn, Henrique Gomes, and Frank Hellmann. “Asymptotic Analysis of the EPRL Four-Simplex Amplitude.” Classical and Quantum Gravity 27, 165009 (2010). DOI. Open PDF.
• Engle, Jonathan, Roberto Pereira, Carlo Rovelli, and Etera Livine. “LQG Vertex with Finite Immirzi Parameter.” Nuclear Physics B 799, 136–149 (2008). DOI. Open PDF.