V2.433: Edge Mode Area Law — Do Constraints Create Area-Law Edge Modes?

Question

The framework’s prediction R = 0.6877 requires n_grav = 10 (all h_mu_nu components), not n_grav = 2 (TT physical modes). The extra 8 modes are supposed to come from diffeomorphism edge modes at the entangling surface (Donnelly-Wall mechanism). Does this mechanism actually produce area-law-contributing edge modes?

This experiment tests three constraint types on the Srednicki lattice:

1. Algebraic (on-site, q_0(j) = q_1(j)): fiber/internal gauge analog
2. Differential (cross-site, q_0(j) = q_1(j+1)): diffeomorphism analog
3. Gauss derivative (dq_0/dj = dq_1/dj): electromagnetism Gauss law analog

Results

Phase 1-2: Baseline and Algebraic Constraints

System	N_eff	Expected	Status
1 unconstrained chain	1.0000	1	OK
2 unconstrained chains	2.0000	2	OK
3 unconstrained chains	3.0000	3	OK
4 unconstrained chains	4.0000	4	OK
2 chains, 1 algebraic constraint	1.0000	1	OK
4 chains, 2 algebraic constraints	2.0000	2	OK

Algebraic constraints reduce N_eff by exactly 1 per constraint. Confirmed.

Phase 3: Differential Constraints (KEY TEST)

n_sub	S_constrained	S_single	N_eff
5	0.0196	0.0556	0.35
10	0.0588	0.1289	0.46
15	0.0938	0.1838	0.51
20	0.1237	0.2261	0.55

N_eff = 0.47 — LESS than 1, far below 2.

The cross-site constraint q_0(j) = q_1(j+1) over-constrains the system near the entangling surface. It pins the subsystem to the environment through the cross-boundary coupling, REDUCING fluctuations and entropy.

Phase 4: Gauss Derivative Constraint

n_sub	S_constrained	S_single	N_eff
5	0.0556	0.0556	1.000
10	0.1289	0.1289	1.000
20	0.2261	0.2261	1.000

N_eff = 1.0000 exactly. The Gauss derivative constraint removes 1 bulk DOF per constraint, leaves a boundary mode (relative constant), but this boundary mode does NOT contribute to alpha. It contributes to delta only.

This matches the known result for vectors: n_eff(alpha) = 2 per vector field, with edge modes contributing to delta = -1/3 (V2.312, V2.397).

Phase 5: Angular Momentum Dependence

l	Unconstrained	Algebraic	Differential	Gauss_deriv
0	2.000	1.000	0.710	1.000
2	2.000	1.000	0.575	1.000
5	2.000	1.000	0.481	1.000
10	2.000	1.000	0.411	1.000
15	2.000	1.000	0.378	1.000

The differential N_eff decreases with l, approaching ~0.35 at high l. Algebraic and Gauss derivative are l-independent.

Phase 6: Edge Mode Scaling

S_edge = S_diff - S_single grows negatively (linearly in n): S_edge ≈ -0.029 - 0.0034·n

The “edge mode” contribution is NEGATIVE — the cross-site constraint suppresses entropy, it doesn’t add to it.

Critical Interpretation

What this experiment DOES show

1. The Gauss derivative mechanism works correctly for vectors: N_eff = 1 per constraint, matching n_comp = 2 per vector field. Edge modes contribute to delta (-1/3 per vector, V2.312) but not to alpha.

2. Simple cross-site constraints do NOT produce area-law edge modes: The “shift” model q_0(j) = q_1(j+1) gives N_eff < 1, not N_eff = 2.

3. The algebraic constraint is exact: N_eff = 1.0000 per constraint, with zero numerical error at all n_sub and l values.

What this experiment DOES NOT show

The cross-site constraint is NOT a correct model of diffeomorphism constraints.

The actual Donnelly-Wall mechanism requires:

• The constraint moves the entangling surface (not just relates field values)
• The edge modes arise from gauge orbits that cross the boundary
• The boundary DOF are components of the metric perturbation h_mu_nu at the cut

My scalar chain model has no notion of “moving the boundary” — the entangling surface is fixed at site n regardless of the field values. A true diffeomorphism on the lattice would shift the cut location, which is a fundamentally different operation from coupling chain values across sites.

The n_grav question remains open

Evidence FOR n_grav = 10	Evidence AGAINST
Omega_Lambda match at 0.4sigma (V2.325)	This experiment: simple constraints don’t produce it
Lattice SVT: alpha = 10*alpha_s (V2.393)	V2.399: constrained phase space gives 8.92, not 10
Bayesian selection (V2.392)	n=10 counting is circular (extracted from Omega_Lambda)
Fiber vs base manifold argument (V2.395)	No first-principles derivation from Einstein constraints

The strongest evidence for n=10 is the SVT decomposition (V2.393): when you literally set up 10 independent scalar chains representing all h_mu_nu components, you get alpha = 10*alpha_s trivially. But this begs the question: WHY should all 10 components contribute, when 8 are gauge?

The answer — that diffeomorphism constraints create edge modes that restore the gauge DOF at the boundary — is compelling in principle but unverified in a constrained computation. V2.399 got 8.92 (close but not 10), and this experiment shows that generic constraint models don’t produce the effect.

What Would Resolve This

A proper resolution requires:

1. Linearized Einstein constraints on the lattice: implement the actual Hamiltonian and momentum constraints for linearized gravity, not toy models.
2. Symplectic reduction: project to the physical phase space and compute entanglement entropy of the reduced system.
3. Boundary identification: show that the reduced phase space has the same dimension in the bulk but EXTRA DOF at the entangling surface.

This is a substantial numerical GR project, equivalent to building a lattice gauge theory for linearized gravity. It has not been done in the literature.

Honest Assessment

The n_grav = 10 counting is the single biggest vulnerability in the framework. If n_grav = 2 (TT only), R = 0.734 and the prediction is 6.7sigma off. If n_grav = 10, R = 0.688 and it works beautifully.

This experiment shows that simple constraint models cannot produce n_grav = 10. The actual mechanism (Donnelly-Wall surface deformation) requires physics that scalar chains cannot capture. The question is whether the Donnelly-Wall argument is correct for quantum gravity — this is an open problem in the field.

The framework’s prediction stands at R = 0.688 ± 0.01 (theoretical uncertainty dominated by the graviton counting: n_grav = 10 ± 1.4 from V2.328). A proper derivation from the Einstein constraints would either confirm or refute this, and is the highest-priority open problem.