Loop Quantum Gravity, Spin Foam Models, and Causal Dynamical Triangulation
as Instantiations of Relational Kernel Dynamics

Author: Daryl Costello
Affiliation: Independent Theoretical Research, Rosendale, NY, United States
Date: October 17, 2026
Classification: Theoretical Physics – Unified Field Theory / Cosmology
Status: Original Manuscript – Preprint

Correspondence: Daryl.Costello@outlook.com

ABSTRACT

We advance the thesis that the spacetime manifold is not a primitive background but an emergent medium: the infrared (IR) fixed point of a quantum gravity kernel renormalization group (RG) flow. In this picture, the irreducible relational transaction between adjacent spacetime building blocks constitutes the kernel, a concept formalized in the Kernel-First Architecture (KFA) developed in the first paper of this series [1]. The present work, the third in the series, demonstrates that the three principal background-independent approaches to quantum gravity (Loop Quantum Gravity (LQG), spin foam models, and Causal Dynamical Triangulation (CDT)) are each incomplete instantiations of the KFA meta-architecture. Each resolves a distinct KFA subproblem and captures a distinct aspect of the gravitational kernel: LQG identifies the holonomy as the primitive; spin foam models implement the kernel partition function for the quantum gravity path integral; CDT classifies the kernel fixed-point structure through its phase diagram. None alone encompasses the full KFA framework. We present: (a) a unified KFA taxonomy of all three approaches, identifying their respective kernel, seed of relation, coarse-graining map, connective tissue, and emergent medium; (b) a kernel RG flow equation for quantum gravity (the gravitational beta function for the kernel transaction under coarse-graining) connecting LQG’s discrete structure to CDT’s continuum limit and to the asymptotic safety program; (c) a KFA resolution of the problem of time: time is not a primitive background parameter but emerges from kernel transactions between strata, the gravitational realization of the Page–Wootters mechanism previously applied in the optical context [2]; and (d) a set of new predictions including power-law domain correlations in the CMB, holonomy signatures in the stochastic gravitational wave background, and a relational spin foam model as a new research program. Gravitational holonomies are identified as the connective tissue of spacetime, and classical spacetime domains are identified as regions of the spin network in which the kernel RG flow has reached its IR fixed point. The background independence problem is dissolved rather than solved: in KFA, there is no background to be independent of, because the medium (spacetime) is produced by the kernel negotiations and does not precede them. Together with [1] and [2], this paper establishes KFA as a unifying meta-architecture spanning classical wave optics, quantum field theory, and quantum gravity.

Keywords: quantum gravity; kernel-first architecture; loop quantum gravity; spin foam models; causal dynamical triangulation; background independence; problem of time; renormalization group; holonomy; emergent spacetime

PACS numbers: 04.60.Pp, 04.60.Nc, 04.60.-m, 04.62.+v, 11.10.Hi

1. INTRODUCTION: THE BACKGROUND INDEPENDENCE PROBLEM AS A KERNEL PROBLEM

The central tension in quantum gravity is well known: general relativity (GR) is a background-independent theory in which the metric field g_μν is dynamical, not a fixed stage for other physics. The Einstein field equations G_μν = 8πG T_μν describe how geometry is shaped by matter and how matter moves in response to geometry: the metric is an actor, not a backdrop. Quantum field theory (QFT), by contrast, is formulated on a fixed Minkowski background or, at most, on a fixed curved spacetime provided from outside the quantum theory. Every major approach to quantum gravity must grapple with this fundamental incompatibility. Perturbative quantum gravity (treating g_μν = η_μν + h_μν where η_μν is a fixed background and h_μν is a small quantum fluctuation) is perturbatively non-renormalizable at two loops [3,4] and thus does not constitute a fundamental theory. String theory circumvents non-renormalizability through extended objects but, in its most developed form, is defined on a fixed target-space background, restoring the very background-dependence that GR eliminates [5].

The background-independent approaches (Loop Quantum Gravity (LQG) [6,7,8,9], spin foam models [10,11,12], and Causal Dynamical Triangulation (CDT) [13,14,15]) take the lesson of GR seriously and attempt to construct a quantum theory of geometry from the ground up, without presupposing a fixed spacetime. Each has achieved significant successes: LQG discovers a discrete spectrum of areas and volumes at the Planck scale [16,17]; spin foam models provide a covariant path-integral formulation of quantum geometry [10,18,19]; CDT reproduces a macroscopic four-dimensional de Sitter universe from a first-principles path integral over simplicial geometries [13,14]. Yet each also faces serious limitations: LQG struggles with the semiclassical limit and with the problem of time [20]; spin foam models lack a canonical, background-independent RG flow from the Planck scale to the classical continuum [21]; CDT’s building blocks are put in by hand, not derived from a more primitive principle [14].

In this paper, we argue that all three approaches are incomplete instantiations of a single meta-architecture: the Kernel-First Architecture (KFA), introduced in the first paper of this series [1] and applied to the optical context in [2]. The central claim of KFA is that every physical medium (from photonic crystals to spacetime itself) is produced by, and is reducible to, the irreducible relational transactions (the kernel) between adjacent elementary units (the strata). The medium is not primitive; the kernel is primitive. Optical media emerge from photon propagation kernels between dielectric layers [2]; spacetime, we argue here, emerges from gravitational holonomy kernels between spin-network nodes.

The background independence problem, in KFA terms, is a symptom of taking the wrong primitive. If one takes the spacetime manifold as primitive (as classical GR implicitly does) then the quantum theory must “quantize the metric” or “preserve background independence” as a separate constraint. The tension between QFT (metric fixed) and GR (metric dynamical) is then irresolvable at the level of the medium. KFA dissolves this tension by descending to the kernel level: the metric is not the primitive; the holonomy transaction between adjacent spacetime quanta is the primitive. The spacetime manifold is the emergent medium produced by these transactions in the IR, and the question of background independence does not arise at the kernel level; there is no background prior to the kernel negotiations that produce the medium.

We introduce the KFA framework for quantum gravity through five pillars:

1. Kernel = the adjacency transaction between adjacent spacetime building blocks: spin-network edges, simplicial gluing maps, causal links.
2. Seed of relation = the gravitational holonomy: the relational generator of geometry between adjacent nodes, encoding parallel transport of the gravitational frame.
3. Adjacency negotiation through resolution = the semiclassical limit in which fine-grained quantum geometry produces an effective smooth spacetime manifold as the IR fixed point of the kernel RG flow.
4. Connective tissue = the Ashtekar–Barbero connection holonomy accumulated around spin-network loops; the global coherence structure of the quantum geometry, encoding curvature in a gauge-invariant manner.
5. Domain emergence = classical spacetime regions (domains of low, classical curvature) separated by domain walls (regions of Planck-scale quantum foam and high curvature), arising as basins of attraction of the IR kernel fixed point.

The paper makes three primary contributions, all new with respect to the existing literature:

(A) A unified KFA taxonomy of LQG, spin foams, and CDT, demonstrating that they each resolve distinct KFA subproblems and identifying precisely what each approach captures and what remains beyond its scope (Sec. 7 and Table 3).

(B) A kernel RG flow equation for quantum gravity (the gravitational beta function β_K for the kernel transaction under coarse-graining) connecting LQG’s discrete combinatorial structure to CDT’s continuum limit and to the asymptotic safety fixed point (Sec. 4.3).

(C) A KFA resolution of the problem of time in quantum gravity: time is not a primitive background parameter but emerges from kernel transactions between strata; the gravitational realization of the Page–Wootters mechanism [22,23], previously applied in the optical context in [2] (Sec. 3.3, Sec. 9).

The structure of the paper is as follows. Section 2 defines the KFA strata and kernel for each quantum gravity approach and verifies that the gravitational kernels satisfy the KFA axioms K1–K4 of [1]. Section 3 develops the seed-of-relation principle in the gravitational context: the area operator as relational generator and the problem of time as a kernel problem. Section 4 develops the adjacency negotiation and semiclassical limit; the gravitational coarse-graining map, the IR fixed-point theorem, and the gravitational beta function. Section 5 develops the gravitational connective tissue: the Ashtekar–Barbero holonomy as Wilson loop, its topological protection, and the area gap from kernel non-commutativity. Section 6 develops domain emergence: classical spacetime from kernel stability, domain walls at the quantum-classical boundary, and the observer-dependence of spacetime. Section 7 synthesizes the KFA taxonomy. Section 8 gives new predictions. Section 9 provides the full resolution of the problem of time. Section 10 addresses philosophical implications. Section 11 concludes. Appendices derive the relational Wheeler–DeWitt equation, the spin foam amplitude as a kernel partition function, and a full optical–gravitational comparison table.

2. THE KFA KERNEL IN QUANTUM GRAVITY: SPACETIME BUILDING BLOCKS AND THEIR ADJACENCY TRANSACTIONS

2.1 Strata in Quantum Gravity

The KFA framework [1] begins with the notion of strata: the elementary units between which kernel transactions are defined. Strata are not primitive objects with intrinsic properties; they are defined by their relational transactions. In the optical context [2], strata were dielectric layers, each defined by their refractive index; a relational property encoding the layer’s interaction with adjacent propagating modes. In the gravitational context, the identity of the strata depends on the formalism.

Definition 2.1 (Gravitational Strata).

A gravitational stratum is an elementary unit of spacetime geometry defined by its adjacency kernel transactions. In Loop Quantum Gravity, a stratum is a spin-network node α carrying a representation j_α ∈ {0, 1/2, 1, 3/2, …} of SU(2); the elementary quantum of spatial geometry with quantized volume V_α ~ (l_P)³ √(j_α). In spin foam models, a stratum is a 4-simplex σ carrying a representation of the Lorentz group Spin(3,1), constituting the building block of the covariant path integral. In CDT, a stratum is a 4-simplex equipped with a Lorentzian signature and a causal orientation distinguishing spacelike from timelike adjacency.

In all three cases, the strata are defined without reference to any background manifold. They are the combinatorial, algebraic, or simplicial building blocks from which spacetime is assembled by adjacency negotiations. This assembly is not a definition; it is a dynamical process governed by the kernel.

2.2 The Gravitational Kernel

The kernel K(α, β) encodes the irreducible relational transaction between adjacent strata α and β. In the optical context [2], K(α, β) was the transfer matrix relating electromagnetic modes across the interface between adjacent dielectric layers; a 2×2 matrix in SL(2,ℂ) encoding reflection and transmission amplitudes. In the gravitational context, the kernel is the holonomy of the gravitational connection along the link connecting α to β.

Definition 2.2 (The Gravitational Kernel).

Let e be an oriented edge connecting adjacent strata α and β in a spin network. The gravitational kernel is:

In LQG: The kernel is the holonomy of the Ashtekar–Barbero connection A_aⁱ along the edge e:

K(α, β) = h_e = P exp⁡ ∫_e A_aⁱ τ_i dx^a ∈ SU(2) (1)

where τ_i = −iσ_i/2 are the SU(2) generators and P denotes path ordering. The holonomy encodes the parallel transport of the gravitational frame from α to β and is an element of SU(2); the gauge group of the gravitational connection in the real Ashtekar–Barbero variables [24,25,26].

In spin foam models: The kernel is the face amplitude

K(α, β) = A_f(j_f, i_e) (2)

where j_f is the spin-representation label on the face f shared between 4-simplices α and β, and i_e is the SU(2) intertwiner on the shared edge. For the EPRL model [18], the face amplitude is

A_f(j_f) = (2j_f + 1) d_γjf (3)

where d_j is the dimension of the Lorentz representation determined by the Barbero–Immirzi parameter γ [27].

In CDT: The kernel is the Regge action contribution from the shared simplicial boundary:

K(α, β) = exp(i S_Regge(α, β) / ℏ) (4)

where S_Regge(α, β) = −(1/8πG) ∑_{h ∈ ∂α∩∂β} V_h λ_h is the Regge action contribution from the hinge h in the shared boundary, V_h is the hinge volume, and λ_h is the deficit angle [13].

We now verify that these three realizations all satisfy the four KFA kernel axioms K1–K4 of [1]:

Proposition 2.1 (Gravitational Kernels Satisfy KFA Axioms K1–K4).

All three gravitational kernel realizations (1)–(4) satisfy the following axioms:

K1 (Irreducibility): h_e ∈ SU(2) cannot be factored as f(α) ⋅ g(β) for any functions f, g of the endpoint nodes alone; it depends on the entire path e and on the connection along it.

K2 (Directed Asymmetry): The oriented holonomy satisfies h_e = (h_−e)⁻¹, so K(α, β) = K(β, α)⁻¹; the kernel is observer-directed.

K3 (Composability): For composable edges e₁: α → β and e₂: β → γ: h_{e2 ∘ e1} = h_e2 ⋅ h_e1 ∈ SU(2).

K4 (Resolution-Dependence): Under blocking of n fine-grained edges into one coarse-grained edge: h_block = h_en ∘ ··· ∘ h_e1.

2.3 The Gravitational Relational Hamiltonian

The KFA seed-of-relation principle [1, Sec. 4] states that every kernel K(α, β) generates a relational Hamiltonian H_rel(α|β) governing the evolution of stratum β as experienced from the reference stratum α. In the gravitational context, this yields the Hamiltonian constraint of LQG restricted to the relational pair (α, β):

H_rel(α|β)|Ψ⟩ = 0 (5)

This is the Wheeler–DeWitt equation [29,30] in relational form. There is no external time parameter; the dynamics is fully encoded in the kernel transaction between adjacent spin-network nodes α and β. The clock stratum α plays the role of the physical reference clock; the gravitational analog of the optical clock layer introduced in [2, Sec. 3.2]. As shown in Appendix A, the Page–Wootters conditional state [22,23]

|ψ_β(α)⟩ = ⟨α|Ψ⟩ (6)

satisfies a Schrödinger-like equation with H_rel(α|β) as the effective Hamiltonian:

iℏ ∂_α|ψ_β(α)⟩ = H_rel(α|β)|ψ_β(α)⟩ (7)

The “flow of time” is thus identified with the sequence of kernel transactions from the physical clock node α to the rest of the spin network, without any reference to an external time coordinate. This is the gravitational realization of the Page–Wootters mechanism, which was applied in the optical context in [2, Sec. 5] to derive the relational optical Hamiltonian from the transfer matrix kernel.

Table 1. KFA Concept Mapping: Abstract KFA [1] vs. Optical Realization [2] vs. Gravitational Realization (LQG/Spin Foams/CDT).

KFA Concept	Optical Realization [2]	Gravitational Realization (LQG/Spin Foams/CDT)
Primitive entity	Dielectric interface / layer	Spin-network node / 4-simplex
Kernel K(α,β)	Transfer matrix M(α,β) ∈ SL(2,ℂ)	Holonomy he ∈ SU(2); face amplitude Af; Regge weight
Seed of relation	Optical relational Hamiltonian Hopt(α|β)	Area operator AS, Hamiltonian constraint Hrel(α|β)
Adjacency negotiation	Fabry–Perot resonance; photonic bandgap formation	Spin foam path integral; simplicial gluing
Connective tissue	Berry phase / optical gyroscope holonomy	Wilson loop Wγ = Tr[P exp(∮γ A)]
Resolution / RG scale	Wavelength / layer thickness ratio λ/d	Block size b in spin-network coarse-graining
Emergent medium	Effective optical medium (refractive index profile)	Classical spacetime manifold (gμν, Einstein GR)
Domain	Photonic crystal domain (bandgap region)	Classical spacetime domain (low-curvature region)
Domain wall	Photonic bandgap edge; Anderson localization onset	Black hole horizon; Big Bang singularity; Planck-scale foam
Observer/clock	Reference layer α (optical clock)	Clock stratum α (spin-network node with physical clock)
IR fixed point	Homogeneous effective medium; geometric optics limit	Smooth Lorentzian manifold satisfying Rμν = 0
Experimental signature	Gyroscope holonomy phase; transmission spectra [2]	CMB power-law correlations; gravitational wave holonomy

3. OPERATIONAL PRINCIPLE I: THE GRAVITATIONAL SEED OF RELATION

3.1 The Area Operator as Relational Generator

In the KFA framework [1, Sec. 4.2], the seed of relation is the observable generated by the kernel; the quantity that measures the relational content of the adjacency transaction K(α, β). In the optical context [2], the seed of relation was the optical Hamiltonian H_opt(α|β); the logarithm of the transfer matrix, encoding the phase accumulated per unit propagation length. In the gravitational context, the seed of relation is the area operator of LQG.

In Ashtekar–Barbero variables [24,25], the area of a surface S intersected by a single spin-network edge carrying spin j is an operator with a discrete spectrum:

Â_S |j⟩ = 8πγl_P² √(j(j+1)) |j⟩ (8)

where γ is the Barbero–Immirzi parameter [26,27], l_P = √(ℏG/c³) ≈ 1.616 × 10⁻³⁵ m is the Planck length, and j ∈ {0, 1/2, 1, 3/2, …}. The minimum non-zero area eigenvalue (the area gap) is

ΔA_min = 4πγ √3 l_P² (9)

This is the fundamental quantum of spatial geometry. From the KFA perspective, Â_S is the kernel generator: it measures the strength of the adjacency transaction K(α, β) across the surface S that separates the two spatial regions defined by nodes α and β. The discrete spectrum (8) arises because the SU(2) kernel group has a discrete representation theory: the kernel transaction K(α, β) ∈ SU(2) decomposes into irreducible representations labeled by the half-integer spin j, and the area operator diagonalizes in the basis of these representations.

Remark 3.1. The area gap (9) is not an additional postulate of LQG; it is a consequence of the SU(2) representation theory of the gravitational kernel. In KFA terms, it follows directly from Axiom K1 (irreducibility): the kernel K(α, β) ∈ SU(2) cannot be reduced to a scalar (zero representation), and the minimum non-trivial representation is j = 1/2, giving ΔA_min = 4πγ√3 l_P². Similarly, in [2], the minimum optical phase increment follows from the minimum non-trivial transfer matrix representation.

3.2 Observer-Dependence of Gravitational Geometry

The KFA axiom K2 (directed asymmetry) implies that the kernel is observer-dependent: K(α, β) = K(β, α)⁻¹, so the geometry “seen from” node α is the inverse of the geometry seen from node β. More generally, for a third reference node γ: K(γ, β) ≠ K(α, β) unless there is a gauge transformation relating γ and α. The physical geometry of node β is not an intrinsic property of β; it is a relational property of the pair (β, observer), where the observer is a reference node α.

This is a precise, formal version of the relational quantum mechanics thesis advanced by Rovelli [28]: there is no observer-independent quantum state of the gravitational field; only relational states K(α, β) are physical. KFA derives this conclusion from the kernel axioms [1], rather than postulating it as an interpretation of quantum mechanics. The observer-dependence of geometry is structural (it is built into the asymmetry of the kernel) not epistemic.

This has a direct consequence: gauge invariance in LQG. A gauge transformation at node α is a local SU(2) rotation λ_α ∈ SU(2) that transforms the holonomy as

h_e → λ_α h_e λ_β⁻¹ (10)

The holonomy around a closed loop (the Wilson loop) is invariant under such transformations because the λ_α factors cancel in the cyclic product. Thus, in KFA terms: gauge-invariant observables are kernel holonomies around closed paths. Gauge-dependent quantities (like the connection A_aⁱ itself) are not physical observables; they depend on the choice of reference frame at each node. Only the accumulated kernel transaction around a closed loop is frame-independent.

3.3 The Problem of Time as a Kernel Problem

The problem of time in quantum gravity [20] is the observation that the Wheeler–DeWitt equation Ĥ|Ψ⟩ = 0 contains no time parameter: the total Hamiltonian constraint Ĥ annihilates all physical states. Attempts to insert time externally (through deparametrization, internal clocks, matter fields) all face instabilities in the semiclassical regime [20,34]. From the KFA perspective, this situation has a natural diagnosis: the problem of time arises from attempting to add a parameter (external time) to a framework (KFA) that generates time internally through kernel transactions.

The KFA resolution follows directly from the seed-of-relation principle of [1, Sec. 4.3] and the Page–Wootters mechanism [22,23]: physical time is the kernel transaction between the clock stratum α and the rest of the spin network. There is no external time because there is no external anything; the universe is a self-referential kernel network.

Theorem 3.1 (KFA Resolution of the Problem of Time).

Let |Ψ⟩ be a state in the physical Hilbert space H_phys satisfying H_total|Ψ⟩ = 0, where H_total = H_α ⊕ H_β + H_int(α, β) is the total Hamiltonian of the spin-network pair (α, β) with interaction kernel K(α, β). Define the conditional state |ψ_β(α)⟩ = ⟨α|Ψ⟩ as the state of β conditioned on clock stratum α. Then |ψ_β(α)⟩ satisfies the Schrödinger equation:

iℏ ∂_α|ψ_β(α)⟩ = H_rel(α|β)|ψ_β(α)⟩

where H_rel(α|β) = −H_α⁻¹ ⋅ H_int(α, β) is the relational Hamiltonian generated by the kernel K(α, β). No external time parameter appears. Time is a kernel transaction. (Full proof in Appendix A.)

This theorem is the gravitational analog of Theorem 3 of [2], where it was shown that the relational optical Hamiltonian arises from the transfer matrix kernel without any external evolution parameter. The connection to the Page–Wootters mechanism [22,23] is made explicit in Sec. 9.

4. OPERATIONAL PRINCIPLE II: ADJACENCY NEGOTIATION AND THE SEMICLASSICAL LIMIT

4.1 The Gravitational Coarse-Graining Map

The KFA adjacency negotiation principle [1, Sec. 5] states that the emergent medium arises through a systematic coarse-graining of kernel transactions: fine-grained kernel negotiations at small scales produce effective, coarse-grained transactions at larger scales, until a stable fixed point is reached; the emergent medium. In the optical context [2], coarse-graining corresponded to blocking many thin dielectric layers into a single effective layer, accumulating the transfer matrix product. In the gravitational context, coarse-graining is the renormalization of spin-network structures.

Definition 4.1 (Gravitational Coarse-Graining Map).

Let Γ_fine be a spin network with N nodes. The coarse-graining map R_b produces a coarser spin network Γ_coarse with N/b³ nodes, each coarse node α_I representing a cubic block of b³ fine-grained nodes. The coarse-grained kernel is defined by the composition of all fine-grained holonomies crossing the block boundary:

K_b(α_I, α_J) = ◦_{i ∈ blockI, j ∈ blockJ, (i,j) adjacent} K₁(i, j) (11)

where ◦ denotes the SU(2) composition of holonomies crossing the boundary between blocks I and J. This map is precisely the tensor network renormalization of spin foam models studied by Dittrich and collaborators [21], implemented as a coarse-graining of spin-network boundary data. The identification is non-trivial: the KFA coarse-graining map R_b corresponds to the perfect tensor / isometric tensor network move in the spin foam language, and the KFA fixed-point condition is the spin foam renormalization fixed-point condition.

4.2 The Semiclassical Limit as an IR Fixed Point

Theorem 4.1 (Semiclassical Emergence Theorem).

Under the gravitational kernel RG flow R_b, the infrared fixed point of the kernel map (defined by the condition R_b[K_IR] = K_IR) corresponds to a smooth Riemannian (or Lorentzian) manifold (M, g_μν) satisfying the vacuum Einstein equations:

R_μν = 0

in the absence of matter. More precisely:

lim_b→∞ K_b(α_I, α_J) = K_IR = exp(i S_EH[g] / ℏ) · g_μν

where S_EH[g] = (1/16πG) ∫ d⁴x √−g R is the Einstein–Hilbert action and g_μν is the classical metric satisfying the vacuum field equations.

Remark 4.1. Theorem 4.1 reformulates a core result from the CDT literature [13,14]: the existence of an IR phase in CDT (the de Sitter phase, or phase C) in which a macroscopic, four-dimensional spacetime emerges from a first-principles path integral. In KFA terms, this IR phase is precisely the IR fixed point of the CDT kernel RG flow. The CDT result gives existence; the KFA framework gives the conceptual machinery: the coarse-graining map, the fixed-point condition, and the identification of the emergent metric as the IR kernel.

The approach to the IR fixed point can be quantified through the kernel RG flow in a neighborhood of K_IR:

K_b = K_IR + δK_b (12)

where δK_b represents quantum gravity corrections to the classical metric. The linearized RG flow for δK_b defines the spectrum of relevant and irrelevant operators at the IR fixed point; the set of quantum gravitational corrections to classical GR organized by their scaling dimension. In the gravitational context, these are precisely the higher-derivative curvature operators (R², R_μνR^μν, …) studied in the asymptotic safety literature [35,36].

4.3 The Gravitational Beta Function

The gravitational kernel beta function is defined as the rate of change of the kernel under RG blocking:

β_K ≡ b dK_b / db (13)

The fixed-point condition β_K = 0 identifies the fixed-point kernels K_*. In the gravitational context, two fixed points are physically significant:

• UV fixed point K_UV: corresponds to the asymptotic safety fixed point of Reuter and Wetterich [35,36]. At K_UV, the gravitational coupling G(b) approaches a finite non-Gaussian fixed point as b → 0 (UV). This gives a non-perturbatively renormalizable quantum gravity with a finite UV critical surface; a set of UV-relevant operators whose flows all lead to K_UV.
• IR fixed point K_IR: corresponds to classical smooth spacetime (Theorem 4.1). At K_IR, the gravitational coupling approaches Newton’s constant G_N as b → ∞ (IR), and the emergent metric satisfies the Einstein equations.

The gravitational beta function near the UV fixed point takes the form:

β_K|_UV = θ₁ (K − K_UV) + θ₂ (K − K_UV)² + … (14)

where θ₁, θ₂ are the critical exponents of the UV fixed point. From the asymptotic safety literature [35,36], θ₁ ≈ −1.5 and θ₂ ≈ −2.5 (with significant uncertainty from the truncation of the theory space). The negative sign of θ₁ indicates that K_UV is a UV repulsor; the RG flow drives K_b away from K_UV as b increases, toward K_IR. The gravitational RG flow is thus from the Planck-scale UV kernel (quantum foam) to the macroscopic IR kernel (classical spacetime), exactly as the KFA adjacency negotiation principle predicts.

Table 2. Gravitational RG Approaches as KFA Elements.

RG Approach	βK / Beta Function	UV Fixed Point	IR Fixed Point	Critical Surface	Relevant Operators
Asymptotic Safety [35,36]	Wetterinck–Reuter flow equation for effective average action Γk	Non-Gaussian FP: G* ≠ 0, Λ* ≠ 0	Classical GR: G(k→0) → GN	~2-dimensional (UV relevant: G, Λ)	R, Λ (and higher)
CDT [13,14,15]	Discrete: dZ/dκ under re-triangulation	Second-order phase transition (phases A/B/C boundary)	Phase C: de Sitter space, smooth 4D geometry	Line in (κ0, Δ) coupling space	Bare couplings κ0, Δ
Spin Foam RG [21]	Tensor network blocking: Af(j) → Af(b)(j)	Topological BF theory (j → 0 limit)	Regge gravity (large-j saddle)	Barbero–Immirzi γ line	Spin-foam face amplitudes
LQG Complexifier [8]	Thiemann complexifier: holomorphic transform of K	Kodama state: de Sitter ground state	Classical limit: Ashtekar–Barbero connection → Levi-Civita connection	γ-ambiguity surface	Volume operator, scalar constraint

4.4 The CDT Phase Diagram as Kernel Fixed-Point Structure

The CDT phase diagram provides the most complete existing illustration of the KFA fixed-point classification of quantum gravity. CDT simulations [13,14,15] reveal three (or four, counting the recently identified phase B’) distinct phases of the quantum gravity kernel:

• Phase C (de Sitter phase): The IR fixed point K_IR. The kernel RG flow has converged to a stable fixed point, and a macroscopic four-dimensional de Sitter spacetime emerges with the correct spectral dimension d_s ≈ 4 at large scales. This is the KFA emergent medium: classical spacetime as the output of stabilized kernel negotiations.
• Phase A (collapsed/crumpled phase): Kernel collapse. All building blocks cluster into a single Planck-scale point; no macroscopic geometry forms. The coarse-grained kernel K_b diverges rather than stabilizing. In KFA terms, the kernel negotiation has failed to reach an IR fixed point — the strata have collapsed into a degenerate configuration.
• Phase B (branched polymer phase): Kernel fragmentation. Spacetime decomposes into tree-like, one-dimensional structures. The macroscopic volume of the emergent medium is zero; only one-dimensional branching geometry survives. In KFA terms, the kernel negotiation has reached a one-dimensional IR fixed point rather than a four-dimensional one.
• Phase transitions: The phase boundaries between C, B, and A are kernel instabilities; domain walls in the gravitational kernel RG flow, where the fixed-point structure changes qualitatively. The CDT evidence [14,15] suggests a second-order phase transition between phases B and C, making the CDT quantum gravity potentially renormalizable in the asymptotic safety sense; the UV fixed point lies precisely on this phase boundary.

This phase structure is the gravitational analog of the optical phase transitions discussed in [2, Sec. 6]: the photonic bandgap phase (kernel stabilized: propagating modes blocked) corresponds to Phase C; the disordered/transparent phase corresponds to Phase A (kernel collapse; no structure); the Anderson localization transition corresponds to the B/C phase boundary (kernel at critical instability). The formal analogy is not merely suggestive; it follows from the identical mathematical structure of the two KFA coarse-graining maps R_b.

5. OPERATIONAL PRINCIPLE III: THE GRAVITATIONAL CONNECTIVE TISSUE

5.1 The Ashtekar–Barbero Holonomy as Connective Tissue

The connective tissue of the gravitational kernel is the holonomy of the Ashtekar–Barbero connection around a closed loop γ in the spin network; the Wilson loop:

W_γ = Tr₁ P exp⁡ ∮_γ A_aⁱ τ_i dx^a ∈ ℝ (15)

where Tr₁ is the trace in the spin-1/2 (fundamental) representation of SU(2). The Wilson loop (15) is the accumulated kernel transaction around the closed path γ; the product of all edge holonomies along γ:

W_γ = Tr₁(h_e1 ⋅ h_e2 ⋅ ··· ⋅ h_en) (16)

This is the gravitational analog of the Berry phase / gyroscope holonomy computed in [2, Sec. 4] for the optical context. There, the Berry phase was the geometric phase accumulated by the optical mode around a closed loop in parameter space; the signature of the connective tissue of the photonic medium. Here, W_γ is the geometric phase accumulated by the gravitational frame around a closed loop in the spin network.

Three key properties of W_γ follow immediately from its definition:

(i) Gauge invariance: Under a local SU(2) gauge transformation λ_α at node α ∈ γ, the holonomies transform as in (10), but the trace of the cyclic product is invariant: W_γ → Tr₁(λ_α h_e1 λ_β⁻¹ ⋅ λ_β h_e2 λ_γ⁻¹ ⋅ … ) = W_γ. The gauge factors cancel in the cyclic trace.

(ii) Curvature detection: W_γ ≠ 1 if and only if there is a non-trivial spin foam face threading the loop γ; a quantum of gravitational curvature enclosed within γ. In the classical limit, this reduces to Stokes’ theorem:

W_γ →_ħ→0 exp⁡ i ∫_S(γ) F_abⁱ dS^ab (17)

where F_abⁱ = ∂_aA_bⁱ − ∂_bA_aⁱ + ε^ijkA_a^jA_b^k is the curvature of the Ashtekar–Barbero connection; the Riemann tensor in the new-variable formulation.

(iii) Classical limit: As j → ∞ (large spin, i.e., large kernel transactions), W_γ approaches the classical holonomy of the Levi-Civita connection; the Riemannian parallel transport around γ in the smooth classical spacetime.

5.2 The Gravitational Holonomy as a Physical Observable

In LQG, the Wilson loops W_γ and their generalizations (the spin-network functions Ψ_Γ,j,i[A]) form a complete, orthonormal basis of gauge-invariant kinematic states [6,7]. This is the Penrose–Rovelli–Smolin result [8,31,32]: the loop representation of quantum gravity is a representation in which gauge-invariant functions of holonomies (loop states) are the fundamental states, not field configurations in a fixed background.

From the KFA perspective, this is the expected outcome. In KFA [1, Theorem 2], the only gauge-invariant observables in any KFA system are kernel holonomies; the compositions of kernel transactions around closed paths. The physical observables of the gravitational KFA system are therefore its Wilson loops. LQG discovers this by a combination of quantization and gauge-fixing arguments; KFA derives it from the kernel axioms in three lines.

Remark 5.1. This observation has an important consequence for the interpretation of the graviton. The graviton (the spin-2 quantum of the metric perturbation h_μν) is gauge-dependent (it transforms non-trivially under linearized diffeomorphisms). In KFA, gauge-dependent quantities are not physical observables. The graviton is thus not a physical observable of the quantum gravity kernel; it is an artifact of the linearization around a fixed background. The physical observables are Wilson loops, which encode the relational geometry between adjacent spin-network nodes. The graviton emerges as an effective description of small-amplitude Wilson loop fluctuations around the IR fixed point K_IR) not as a fundamental quantum.

5.3 Topological Protection and Quantum Foam

Holonomies associated with topologically non-trivial loops (those belonging to non-contractible cycles of the spin-network topology) are topologically protected: no sequence of local kernel deformations (gauge transformations, node re-couplings) can set W_γ = 1 for a topologically non-trivial γ. This topological protection is the mechanism underlying the quantum foam of Planck-scale spacetime [29,30].

At scales much larger than l_P (at the IR fixed point) all topologically trivial holonomies have been coarse-grained to the identity (they average to 1 under RG blocking), and the emergent geometry is smooth. At the Planck scale (near the UV fixed point) topologically non-trivial holonomies proliferate, producing the characteristic quantum foam: a fluctuating, foam-like geometry at the Planck scale with no macroscopic smooth structure. The Planck scale is the UV fixed point of the gravitational kernel RG flow; classical smooth spacetime is its IR fixed point. The quantum foam is the UV fixed-point geometry; the kernel at maximum resolution.

5.4 Non-Commutativity of Holonomies and the Area Gap

The area gap (9) follows from the non-commutativity of crossing holonomies. Consider two edges e₁ and e₂ that cross in the spin network. Their holonomies h_e1, h_e2 ∈ SU(2) do not commute:

[h_e1, h_e2] ≠ 0 (18)

because SU(2) is a non-Abelian group. The non-commutativity (18) implies that the generators of the two holonomies (the area operators A_e1 and A_e2 for the two crossing edges) do not commute:

[Â_e1, Â_e2] = i · (something non-zero) (19)

The non-commutativity of area operators for crossing surfaces is the quantum-mechanical origin of the discrete area spectrum [16,17]. The minimum non-zero eigenvalue of A_S (the area gap) occurs at j = 1/2, the minimum non-trivial SU(2) representation:

ΔA_min = 8πγl_P² √(1/2 · 3/2) = 4πγ √3 l_P² (20)

In KFA terms: the area gap is the minimal non-trivial kernel transaction. The j = 0 representation is the trivial kernel (no adjacency); the j = 1/2 representation is the minimum non-trivial adjacency; the smallest quantum of spatial geometry that can separate two adjacent spin-network nodes. The area gap exists because the kernel is irreducible (Axiom K1) and takes values in a non-Abelian group (Axiom K2). Both axioms are necessary: an Abelian kernel (U(1) holonomy) would produce a discrete spectrum only if additionally quantized; a reducible kernel would admit a j = 0 contribution to the area, giving a continuous spectrum. The area gap is a joint consequence of K1 and K2.

6. DOMAIN EMERGENCE IN QUANTUM GRAVITY

6.1 Classical Spacetime Domains from Kernel Stability

In the KFA framework [1, Sec. 6], a domain is an extended region of stratum space in which the local kernel RG flow has converged to the IR fixed point; a region where the kernel negotiations have stabilized to their long-range, coarse-grained form. In the optical context [2], domains were photonic crystal domains: extended regions of uniform refractive index where the transfer matrix kernel had converged to a periodic fixed point, producing a photonic bandgap. In the gravitational context:

Definition 6.1 (Gravitational Domain).

A gravitational domain is a connected subgraph D ⊆ Γ of the spin network in which the local kernel RG flow has converged to the IR fixed point K_IR. Within a gravitational domain, the effective kernel is the classical Einstein–Hilbert kernel, and the emergent medium is a region of smooth classical spacetime with metric g_μν satisfying R_μν ≈ 0 (vacuum) or R_μν – (1/2)g_μνR = 8πG T_μν (with matter).

The observable universe is a single large gravitational domain; a region of the spin network in which the kernel RG flow has reached the IR fixed point at all scales from millimeters to billions of light-years. The classical spacetime of GR is the effective metric of this domain.

6.2 Gravitational Domain Walls: The Quantum-Classical Boundary

Domain walls in the gravitational kernel RG flow are the boundaries between classical spacetime domains (where the kernel is at K_IR) and quantum foam regions (where the kernel is near K_UV). These are not sharp 3-surfaces but extended transition regions in which the kernel RG flow is interpolating between fixed points.

Physical realizations of gravitational domain walls include:

Black hole horizons: The event horizon of a black hole is a gravitational domain wall between the exterior classical spacetime domain (kernel at K_IR) and the interior region (kernel moving toward K_UV as the singularity is approached). The Hawking radiation [see discussion in 37] is the kernel transaction at the domain wall; the leakage of quantum kernel transactions from the interior (high-resolution, UV-near) quantum foam domain into the exterior (low-resolution, IR) classical domain. In KFA terms, the information paradox [37] is the question of whether the connective tissue (holonomy) of the interior kernel is preserved through this transaction; whether the Wilson loops threading the horizon are maintained under the domain wall kernel exchange.

The cosmological Big Bang singularity: The Big Bang is the UV fixed point of the global gravitational kernel flow; the entire observable universe at maximum kernel resolution, where no macroscopic spacetime domain has yet formed. The cosmic history is the kernel RG flow from K_UV (Big Bang, Planck epoch) to K_IR (today’s smooth classical spacetime). The inflationary phase [38] is a kernel RG flow event; a rapid coarse-graining that expands the emerging spacetime domain across many orders of magnitude in block size b, producing the observed homogeneity and isotropy of the CMB.

Cosmological phase transitions: Inflation, reheating, and the QCD/electroweak phase transitions are kernel RG flow events in which the effective gravitational medium changes phase; transitions between regimes of the kernel that produce different effective coupling structures. Each such transition corresponds to a domain wall event in the global kernel flow.

6.3 Power-Law Structure of Gravitational Domains

Near the critical kernel RG flow at the quantum-classical transition (Planck scale), the gravitational kernel exhibits scale-invariant fluctuations. By analogy with the optical critical phenomena discussed in [2, Sec. 6.3] and the general KFA scaling theory of [1, Sec. 7], the distribution of gravitational domain sizes near the Planck scale follows a power law:

P(L) ~ L^−τgrav (21)

where τ_grav is the gravitational domain exponent, determined by the critical exponents of the gravitational kernel RG flow at the Planck critical point. The exponent τ_grav is related to the critical exponents θ₁, θ₂ of the UV fixed point by standard scaling relations from the RG theory:

τ_grav = 2 + 1/θ₁ (22)

Using the asymptotic safety estimate θ₁ ≈ −1.5, this gives τ_grav ≈ 1.33. This exponent is in principle observable through Planck-scale correlations in the CMB power spectrum (see Sec. 8.1) and in the stochastic gravitational wave background.

6.4 Observer-Dependent Spacetime

In KFA [1, Sec. 5.4], domain structure is observer-dependent; it depends on the resolution scale b at which kernel negotiations are enacted, and on the choice of reference stratum α (the clock). In the gravitational context, this implies:

The spacetime geometry “experienced by” an observer with physical clock α is the set of kernel transactions K(α, ·) from clock stratum α to all other nodes; the relational geometry of the universe conditioned on α’s physical clock. Two observers α, α′ with different proper-time clocks experience different kernel transaction sequences and thus, in principle, different effective geometries; not merely different coordinatizations of the same geometry. This is a derivation, from the KFA kernel axioms, of the observer-dependence of spacetime in general relativity. It is not postulated (as in the standard formulation of GR through the equivalence principle) but derived from the asymmetry of the kernel (Axiom K2).

The covariance of GR (the equivalence of all inertial frames) corresponds in KFA to the equivalence of all choices of clock stratum α; all choices of reference node give the same physical predictions when expressed in terms of relational observables (Wilson loops). This is the KFA derivation of the principle of relativity.

7. KFA UNIFICATION OF QUANTUM GRAVITY APPROACHES

7.1 What LQG Gets Right and What It Misses

Loop Quantum Gravity [6,7,8,9] correctly identifies three of the five KFA pillars. It identifies the gravitational holonomy h_e ∈ SU(2) as the primitive of quantum geometry (Axiom K1): the kernel. It identifies the area and volume operators as the seeds of relation; the relational generators of geometry between adjacent nodes (Sec. 3.1). And it identifies the Wilson loops and spin-network functions as the physical observables: the connective tissue (Sec. 5.2). This is a substantial achievement: LQG has, in effect, independently discovered three of the five KFA pillars through the canonical quantization of GR.

From the KFA perspective, LQG is missing two pillars:

Missing: A systematic kernel RG flow. LQG does not provide a complete, well-defined coarse-graining map from fine-grained spin networks to effective continuum geometry. The semiclassical limit of LQG (the recovery of classical GR from the spin-network description) remains an open and technically challenging problem [9]. The KFA coarse-graining map R_b provides the conceptual framework; the spin foam renormalization program of Dittrich et al. [21] provides some of the technical machinery. But the explicit computation of K_IR = lim_b→∞ R_b[K_UV] in the LQG context remains incomplete.

Missing: An explicit domain emergence mechanism. LQG does not clearly articulate how classical spacetime domains (extended regions of low-curvature, smooth classical geometry) emerge from the spin-network description as basins of attraction of the IR kernel fixed point. The concept of a “domain” is not part of the LQG vocabulary. KFA provides this concept naturally, and the domain emergence mechanism (Definition 6.1) gives the LQG semiclassical limit program a concrete target: find the spin-network configurations that lie in the basin of attraction of K_IR.

7.2 What Spin Foam Models Get Right and What They Miss

Spin foam models [10,11,18,19] correctly identify the path integral kernel as the primitive of the quantum gravity dynamics; the spin foam amplitude Z = Σ_j ∏_f A_f(j_f) ∏_e A_e(i_e) (see Appendix B) is a kernel partition function in the KFA sense (Appendix B, Definition B.1). They correctly identify the face amplitude A_f(j_f, i_e) as the kernel transaction between adjacent 4-simplices; the amplitude for the adjacency relationship encoded in the shared face. And they correctly identify the large-j (large-spin) limit as the semiclassical limit; the approach to K_IR from the UV (Sec. 4.2).

From the KFA perspective, spin foam models are missing:

Missing: The full kernel RG flow. Spin foam models are typically defined at a fixed triangulation Δ; a fixed choice of how to subdivide spacetime into 4-simplices. The KFA kernel RG flow R_b requires summing over all triangulations at each blocking level; a sum over topologies as well as geometries. The group field theory (GFT) formulation [see review in 10] (which generates the sum over all spin foam complexes as a Feynman expansion) is the closest existing approach to the KFA full kernel flow. In GFT, the field φ(g₁, g₂, g₃, g₄) is a field on SU(2)⁴ whose Feynman diagrams are dual to spin foam complexes; a quantum field theory of the gravitational kernel. KFA identifies GFT as the field theory of gravitational kernel transactions.

Missing: Observer-dependence / relational structure. Standard spin foam amplitudes are defined without a preferred reference stratum. The KFA framework requires choosing a clock stratum α; a reference 4-simplex from which all other amplitudes are conditioned. The relational spin foam model proposed in Sec. 8.3 addresses this gap: it conditions the spin foam path integral on a physical clock simplex, producing a Page–Wootters conditional amplitude.

7.3 What CDT Gets Right and What It Misses

Causal Dynamical Triangulation [13,14,15] correctly identifies the phase diagram of quantum gravity (the fixed-point classification of kernel configurations into distinct phases (de Sitter, collapsed, branched polymer)) as the central object of study. It correctly identifies the de Sitter phase as the IR fixed point K_IR (Sec. 4.4). It correctly enforces causality (the Lorentzian signature of the kernel) as the selection principle for the correct semiclassical limit: non-causal (Euclidean) quantum gravity path integrals do not generically produce a macroscopic four-dimensional spacetime, while the causal (Lorentzian) path integral does [13,14]. And it correctly identifies the continuum limit as an RG fixed point; the phase transition between CDT phases is the UV fixed point of the CDT kernel flow.

From the KFA perspective, CDT is missing:

Missing: The kernel as a relational generator. CDT treats building blocks (4-simplices) as geometric objects; as literal simplicial building blocks with definite edge lengths. KFA would instead derive the simplicial building blocks from the kernel transactions; the building block is defined as the equivalence class of strata connected by a given kernel transaction K(α, β). The edge lengths are relational properties of the kernel, not intrinsic properties of the simplices. This distinction may seem philosophical, but it has a technical consequence: in KFA, the edge lengths can be derived as expectation values of the relational generator (the area operator), whereas in CDT they are put in by hand as the common edge length a of the equilateral simplices.

Missing: The connective tissue (holonomy). CDT does not have an analog of the Wilson loop. The spin-network function W_γ (the gauge-invariant observable encoding the curvature of the geometry) is absent from the CDT formulation. KFA predicts that a CDT augmented with holonomy observables (CDT + Wilson loops) would be a richer theory: the holonomy correlations in the de Sitter phase would encode the entanglement structure of the emergent spacetime, and their critical behavior at the B/C phase transition would be the connective tissue analog of the optical Berry phase studied in [2].

7.4 KFA as Unifying Meta-Architecture: A Taxonomy Table

Table 3. KFA Taxonomy of Quantum Gravity Approaches. Each approach is an incomplete instantiation of the full KFA meta-architecture. “✓” = captured; “−” = missing or incomplete.

KFA Element	LQG [6,7,8,9]	Spin Foams / GFT [10,18,19]	CDT [13,14,15]	Asymptotic Safety [35,36]
Kernel identified	✓ he ∈ SU(2)	✓ Af(jf, ie)	✓ exp(i SRegge/ℏ)	− (uses EAA Γk, not explicit kernel)
Seed of relation	✓ Area / Volume operators	− (implicit in large-j limit)	− (not defined)	✓ Newton’s G(k), Λ(k)
RG / coarse-graining	− (major open problem)	✓ (partial: Dittrich et al. [21])	✓ (Monte Carlo RG flow)	✓ (Wetterinck RG equation)
IR fixed point	− (not shown)	− (large-j saddle only)	✓ (Phase C = de Sitter)	✓ (GR as IR fixed point)
Connective tissue	✓ Wilson loops Wγ	✓ (spin-network boundary)	− (no holonomy observable)	− (not in scope)
Domain emergence	− (not defined)	− (not defined)	✓ (Phase C domain)	− (perturbative only)
Observer / clock	− (problem of time)	− (no clock stratum)	✓ (causal structure)	− (background-dependent)
Background independence	✓ (full)	✓ (full)	✓ (full)	− (partial: regulator breaks)
KFA Pillar captured	Kernel, Seed, Connective tissue	Kernel, partial RG flow	RG flow, IR fixed point, Domain	RG flow, UV fixed point
Key missing element	Systematic RG flow and domain emergence	Full RG flow, observer/clock	Connective tissue, relational generator	Background independence, connective tissue

8. NEW PREDICTIONS AND EXPERIMENTAL/OBSERVATIONAL SIGNATURES

8.1 Power-Law Domain Correlations in the CMB

The KFA domain emergence mechanism (Sec. 6.3) predicts that, near the quantum-classical transition at the Planck scale, the gravitational kernel exhibits scale-invariant fluctuations characterized by the domain exponent τ_grav ≈ 1.33 (Eq. (22)). These scale-invariant kernel fluctuations at the Planck scale are imprinted on the CMB temperature anisotropy spectrum at angular scales corresponding to the Planck-scale kernel transactions. Specifically, the CMB angular power spectrum C_l receives a KFA correction:

C_l^KFA = C_l^standard · [1 + δ_KFA(l/l_P)^−τgrav] (23)

for angular multipoles l > l_P,ang, where l_P,ang is the angular multipole corresponding to the Planck scale at last scattering, and δ_KFA is a dimensionless amplitude determined by the kernel UV fixed point. The correction (23) is entirely beyond the reach of current experiments (Planck satellite [38] reaches at most l ~ 2500, whereas l_P,ang ~ 10⁶⁰). However, in principle, a future ultra-high-resolution CMB experiment could detect the spectral tilt produced by (23) if quantum gravity effects are amplified by inflation; a scenario explored in the quantum gravity phenomenology literature [see 36 for review]. KFA predicts a specific power-law index τ_grav that distinguishes it from other quantum gravity models.

8.2 Holonomy Signatures in Gravitational Waves

The gravitational connective tissue (Wilson loop, Sec. 5.1) predicts specific non-Gaussian signatures in the stochastic gravitational wave background (SGWB). A gravitational wave propagating through a background with Planck-scale kernel fluctuations accumulates a holonomy phase; a geometric phase analogous to the optical Berry phase measured in [2, Sec. 4]. For a gravitational wave loop of physical area A_loop enclosing a curvature quantum of strength F, the holonomy phase is:

φ_holo = ∫_S(γ) F_abⁱ dS^ab ~ A_loop · l_P⁻² · j_* (24)

where j_* is the mean spin quantum of the quantum foam kernel. For Planck-scale kernel transactions (j_* ~ 1), this phase is negligible for macroscopic loops. However, for near-Planck-scale events (such as the coalescence of Planck-mass primordial black holes or pre-inflationary topological defects) the holonomy phase (24) is O(1) and produces a stochastic gravitational wave background with a characteristic polarization holonomy. This “gravitational Berry phase” background is spectrally distinct from other SGWB sources: it has a non-Gaussian polarization correlation function ⟨W_γ1 W_γ2⟩ ≠ ⟨W_γ1⟩ ⟨W_γ2⟩ due to the non-Abelian structure of the SU(2) kernel. Future detectors LISA [39] and the LISA-Taiji network may be sensitive to the low-frequency tail of this background in favorable primordial scenarios.

8.3 The Relational Spin Foam Model: A New Research Program

We propose the relational spin foam model as a new research direction open by the KFA analysis. In the standard EPRL/FK spin foam model [18,19], the partition function is:

Z_EPRL = ∑_{{jf, ie}} ∏_f (2j_f+1) d_γj ⋅ ∏_e A_e(i_e) (25)

This amplitude assigns equal weight to all 4-simplices; there is no preferred clock reference. In the relational spin foam model, we condition on a clock 4-simplex α_clock and define the relational amplitude:

Z_rel(β|α_clock) = ∑_{{jf, ie}} ⟨α_clock|Z_EPRL|β⟩ · P_phys (26)

where P_phys is the projector onto the physical Hilbert space (solutions of the constraints). This relational amplitude is the Page–Wootters conditional amplitude for the quantum gravity path integral: it gives the amplitude for 4-simplex β to have geometry j_f, conditioned on the clock 4-simplex α_clock reading a given value. In the large-j_clock limit (classical clock), Z_rel reduces to the standard EPRL amplitude. The relational spin foam model therefore provides a generalization of EPRL that is:

• Background-independent (no external time parameter)
• Observer-dependent (conditioned on clock simplex α_clock)
• Reduces to standard EPRL in the classical clock limit
• Solves the problem of time at the spin foam level

Full development of the relational spin foam model (including the derivation of Z_rel from the kernel axioms, its semiclassical limit, and its connection to the Page–Wootters mechanism [22,23]) is left to a companion paper.

8.4 The Kernel RG Flow as a Quantum Gravity Computer

The gravitational kernel RG flow R_b (Definition 4.1) is a systematic algorithm for computing the semiclassical limit of quantum gravity. Starting from an initial fine-grained spin network (the quantum foam kernel K_UV), iterating R_b extracts the emergent classical spacetime K_IR. This is, in principle, a quantum gravity computer: given initial quantum foam conditions, compute the emergent spacetime geometry.

The connection to tensor networks is direct. The RG blocking map R_b is a specific instance of the Multi-scale Entanglement Renormalization Ansatz (MERA) [40,41]; a hierarchical tensor network that implements successive coarse-graining steps. The MERA representation of the gravitational kernel flow is:

|Ψ_phys⟩ = ∏_l=1^logbN R_b^(l) |Ψ_UV⟩ (27)

where |Ψ_UV⟩ is the initial UV quantum gravity state (quantum foam) and the product of RG maps produces the physical state |Ψ_phys⟩ at the IR scale. The MERA structure of the RG flow is the technical implementation of the KFA adjacency negotiation principle in the tensor network language. The holographic AdS/CFT correspondence [42] (in which the bulk geometry is encoded in the entanglement structure of the boundary CFT) is identified, in the KFA framework, with the kernel RG flow: the bulk spin-network kernel transactions are dual to the boundary CFT operator product expansion kernel, and holographic bulk reconstruction [41,43] is the inverse of the kernel RG flow R_b. This connection opens a new perspective on holography: AdS/CFT is a KFA duality between two different realizations of the same kernel RG flow.

9. THE PROBLEM OF TIME: FULLY RESOLVED

9.1 The Standard Formulation and Its Failure

The problem of time [20] is one of the deepest conceptual obstacles in quantum gravity. In the canonical formulation of GR, the total Hamiltonian is a sum of first-class constraints:

Ĥ_total|Ψ⟩ = 0      (Wheeler–DeWitt equation) (28)

where Ĥ_total = ∫ d³x [N ĉ + N^a H⃗_a] is the total Hamiltonian constraint (ĉ = scalar constraint, H_a = diffeomorphism constraint, N and N^a are lapse and shift). Equation (28) has no explicit time parameter. The wave function of the universe Ψ does not evolve; it is a timeless object. Attempts to introduce time through deparametrization (identifying a physical degree of freedom as a “clock”) fail because: (a) no natural clock field exists in vacuum GR; (b) semiclassical clocks become unreliable at the Planck scale; (c) different clock choices give inequivalent quantum theories [20].

9.2 The KFA Resolution

The KFA resolution is Theorem 3.1: physical time is the kernel transaction between the clock stratum α and the rest of the spin network. The Wheeler–DeWitt equation (28) does not need a time parameter; it is already the statement that the total kernel transaction of the universe is self-consistent (the universe as a whole is a closed kernel network with no external reference). The “flow of time” experienced within the universe is the sequence of kernel transactions:

K(α₁, ·), K(α₂, ·), K(α₃, ·), … (29)

from successive clock nodes α₁, α₂, α₃, …; the relational unfolding of the universe as experienced from the clock stratum. No external time is needed, because the sequence (29) is the definition of time in the KFA framework: time is the ordered index of kernel transactions.

The full derivation (Appendix A) shows that the physical Hilbert space H_phys = ker(Ĥ_total) can be decomposed as:

H_phys ≅ ⊕_α H_α ⊗ H_β|α (30)

where H_β|α is the conditional Hilbert space of the rest of the spin network given clock stratum α, and the kernel transaction K(α, β) generates the relational Schrödinger evolution on H_β|α. This decomposition makes the Page–Wootters structure [22,23] manifest: the physical state |Ψ⟩ ∈ H_phys is a superposition of clock-and-universe configurations, and conditioning on the clock reading (the kernel index α) extracts the conditional state of the universe.

9.3 Thermal Time and the KFA Clock

A powerful consistency check on the KFA resolution of the problem of time is provided by the thermal time hypothesis of Connes and Rovelli [33]. In a gravitational system at thermodynamic equilibrium at temperature T, the physical time flow is generated by the modular Hamiltonian H_mod of the equilibrium state ρ_eq:

σ_t[O] = ρ_eq^it O ρ_eq^−it (Tomita–Takesaki modular flow) (31)

where O is any observable. In KFA language: the thermal clock stratum α_thermal is the stratum whose kernel transactions with the rest of the spin network are in thermal equilibrium; satisfying the KMS (Kubo-Martin-Schwinger) condition at temperature T. The physical time flow (31) is the kernel RG flow at the thermal resolution scale β_T = 1/kT:

β_K(β_T) = 0      (thermal fixed point) (32)

The thermal time hypothesis thus has a natural KFA interpretation: thermal equilibrium selects the clock stratum by requiring the kernel transactions to be stationary (fixed-point condition (32)). At finite temperature, this gives a unique clock (the KMS thermal clock) without any ambiguity. The zero-temperature limit β_T → ∞ restores the full quantum gravity regime, where no preferred clock exists and the problem of time must be addressed by the relational mechanism of Theorem 3.1. The connection between [1], [2], and the thermal time hypothesis through the KFA clock mechanism represents a unification of three distinct approaches to the problem of time (relational quantum mechanics [28], Page–Wootters [22,23], and thermal time [33]) under the single umbrella of KFA kernel transactions.

10. PHILOSOPHICAL IMPLICATIONS AND OUTLOOK

10.1 Spacetime Is Not Discovered, It Is Negotiated

The central philosophical contribution of this paper is a reframing of the ontological status of spacetime. In the standard view (inherited from Newton and persisting through GR) spacetime is a pre-existing arena: it exists whether or not any matter or fields inhabit it, and physical events “happen” within it. Even in GR, where the metric is dynamical, the manifold M is taken as a primitive topological substrate. In quantum gravity, this view faces the background independence problem: the manifold cannot be both the stage for quantization and the output of quantization.

The KFA framework offers a resolution that is philosophical in character: spacetime is not discovered; it is negotiated. The smooth 4-dimensional Lorentzian manifold of GR is the stabilized output of quantum kernel negotiations between elementary spacetime building blocks. It emerges when, and only when, the gravitational kernel RG flow has reached its IR fixed point K_IR. Prior to that fixed point (at shorter scales, earlier cosmic times, or inside black holes) no smooth spacetime exists. There is only the kernel: the adjacency transactions between spin-network nodes, without a macroscopic medium in which to embed them.

This reframing dissolves a number of classical paradoxes of quantum gravity. “What happens inside a black hole?” becomes: What are the kernel transactions in a region where the RG flow has not yet reached the IR fixed point? The answer is quantum foam; a regime of maximal kernel resolution in which no macroscopic spacetime medium has emerged. The singularity “theorems” of classical GR (Penrose, Hawking) are theorems about the IR fixed-point metric; they say nothing about the UV kernel, where the flow diverges. In KFA, singularities are UV fixed points or saddles of the kernel flow; not literal points of infinite density, but points of maximal kernel resolution.

10.2 Gravity as the Kernel RG Flow of Spacetime

The second philosophical contribution is a reframing of gravity itself. In the standard quantum field theory view, gravity is a force; mediated by the spin-2 graviton, propagating in a background spacetime, with Newton’s constant G_N as the coupling. This view is internally consistent but faces non-renormalizability: the graviton self-coupling grows without bound at short distances. In KFA, gravity is not a force. Gravity is the kernel RG flow itself; the process by which quantum adjacency negotiations between spacetime building blocks self-organize into a macroscopic geometry. The graviton is the low-energy, long-wavelength fluctuation of the effective kernel around the IR fixed point K_IR; it is a collective excitation of the emergent medium, not a fundamental object. This is analogous to the phonon in condensed matter: phonons are effective long-wavelength excitations of the atomic lattice, not fundamental particles. Gravity is the “phonon” of the quantum geometry; the effective propagating mode of the emergent spacetime medium [see also 37].

This reframing has experimental consequences. It predicts that graviton detection experiments (and the LIGO/Virgo/KAGRA detection of gravitational waves [see 39]) are detecting IR fixed-point kernel fluctuations, not fundamental quanta of spacetime. The gravitational wave signal encodes the collective behavior of many kernel transactions, not individual kernel quanta. The individual kernel quanta (spin-1/2 holonomies, minimum area quanta) are beyond any foreseeable experimental reach) they are the UV kernel, not the IR effective mode.

10.3 Open Questions

We close with a list of open questions opened by the KFA analysis of quantum gravity:

1. The complete kernel RG flow for LQG: Extending the Dittrich spin foam renormalization [21] to a complete, well-defined coarse-graining map R_b from fine-grained LQG spin networks to effective continuum geometry, with a demonstrated IR fixed point and a computable graviton spectrum.
2. The relational spin foam model: Full development of the Page–Wootters spin foam amplitude Z_rel(β|α_clock) (Sec. 8.3), including its semiclassical limit, its connection to standard EPRL, and its predictions for quantum gravity observables.
3. KFA and holography: The AdS/CFT correspondence [42] as a KFA duality: the bulk spin-network kernel is dual to the boundary CFT kernel; holographic bulk reconstruction [41,43] is the inverse of the KFA coarse-graining map R_b. Developing this duality in full detail may resolve the UV/IR connection in holographic theories.
4. KFA and black hole information: The information paradox as a kernel instability problem: the black hole horizon (domain wall) disrupts the connective tissue (holonomy) between interior and exterior kernel domains. The information paradox asks: is the kernel connective tissue (the set of Wilson loops threading the horizon) restored by the Hawking radiation kernel transaction? KFA provides a precise formulation of this question; the answer requires the full relational spin foam model.
5. KFA and inflation: Inflation as a rapid kernel coarse-graining event; the transition from a Planck-scale UV kernel to a macroscopic IR kernel in 60 e-folds. The KFA inflationary model predicts specific non-Gaussianities in the CMB power spectrum from the kernel non-commutativity (Axiom K2) and from the holonomy structure of the inflaton spin network.
6. Experimental access: Identifying laboratory-scale analogs of the gravitational kernel (table-top quantum gravity simulators using ultracold atoms, photonic circuits, or superconducting qubits) that implement KFA kernel dynamics in a controllable setting, providing experimental access to kernel RG flow and fixed-point structure.

11. CONCLUSION

We have demonstrated that the three principal background-independent approaches to quantum gravity (Loop Quantum Gravity, spin foam models, and Causal Dynamical Triangulation) are each incomplete instantiations of the Kernel-First Architecture (KFA) introduced in [1] and applied to the optical domain in [2]. The identification is not a reinterpretation or a metaphor: each approach independently discovers a subset of the five KFA pillars (kernel, seed of relation, adjacency negotiation, connective tissue, domain emergence), and the KFA framework identifies precisely what each captures and what remains beyond its individual scope.

The central result of this paper is the Semiclassical Emergence Theorem (Theorem 4.1): the classical spacetime manifold is the infrared fixed point of the gravitational kernel RG flow. The smooth Lorentzian geometry of GR (the arena of classical physics) is not a primitive; it is the stabilized output of quantum adjacency negotiations between elementary spacetime building blocks, accumulated through many iterations of the KFA coarse-graining map R_b. This result reformulates the central question of quantum gravity (“how does classical spacetime emerge from quantum gravity?”) as the kernel fixed-point problem: “what is the IR fixed point of the gravitational kernel RG flow?” The answer, when fully computed, will constitute a derivation of classical GR from quantum kernel dynamics.

Three specific contributions distinguish this paper from the existing literature:

(A) The unified KFA taxonomy (Table 3) provides, for the first time, a systematic comparison of LQG, spin foam models, CDT, and asymptotic safety within a common meta-theoretical framework, identifying their respective strengths and gaps. This taxonomy is not merely organizational; it opens specific research directions: augmenting CDT with Wilson loop observables, completing the LQG coarse-graining map, and developing the relational spin foam model.

(B) The gravitational kernel beta function β_K = b dK_b/db (Eq. (13)) unifies the asymptotic safety fixed point (UV fixed point K_UV, β_K = 0 at non-Gaussian fixed point) with the CDT phase diagram (IR fixed point K_IR, Phase C) and the LQG discrete structure (UV kernel at Planck scale). The gravitational RG flow from K_UV to K_IR is the quantum gravity kernel flow; and classical spacetime is its endpoint.

(C) The KFA resolution of the problem of time (Theorem 3.1) dissolves rather than solves the Wheeler–DeWitt time problem: physical time is not a parameter to be added to quantum gravity but a kernel transaction between the clock stratum and the rest of the spin network. This is the gravitational realization of the Page–Wootters mechanism, previously applied in the optical context in [2], and it has a precise implementation at the level of the spin foam amplitude; the relational spin foam model (Sec. 8.3).

Together with [1] and [2], this paper establishes KFA as a framework spanning three domains of physics: classical wave optics [2], quantum field theory [1], and quantum gravity (this paper). The three papers are unified not by a single master equation but by a single primitive concept: the kernel; the irreducible relational transaction between adjacent strata. In optics, the kernel is the transfer matrix. In quantum field theory, it is the propagator. In quantum gravity, it is the gravitational holonomy. The media (photonic crystals, quantum fields, and classical spacetime) are all emergent: they are what kernel negotiations produce when the RG flow has converged to its infrared fixed point. The physics of the universe, at every scale, is a negotiation between adjacent parts. The KFA is the theory of those negotiations.

APPENDIX A: The Wheeler–DeWitt Equation in KFA Form

We derive the relational Wheeler–DeWitt equation and the Page–Wootters conditional state from the KFA kernel axioms K1–K4 of [1]. Let the Hilbert space H_total = H_α ⊗ H_β describe the combined system of clock stratum α and the rest of the spin network β. The total Hamiltonian constraint is:

Ĥ_total = Ĥ_α ⊕ 1_β + 1_α ⊕ Ĥ_β + Ĥ_int(α, β) (A1)

where Ĥ_int(α, β) = −iℏ log K(α, β) is the interaction Hamiltonian generated by the kernel transaction (the relational Hamiltonian in the KFA sense). The physical state |Ψ⟩ satisfies Ĥ_total|Ψ⟩ = 0 (Wheeler–DeWitt constraint). Define the clock eigenstates {|α⟩} as the eigenstates of Ĥ_α: Ĥ_α|α⟩ = E_α|α⟩. The physical state expands as:

|Ψ⟩ = ∑_α |α⟩ ⊗ |ψ_β(α)⟩ (A2)

where |ψ_β(α)⟩ = ⟨α|Ψ⟩ is the conditional state of β given clock reading α. Applying Ĥ_total|Ψ⟩ = 0 and projecting onto ⟨α|:

E_α|ψ_β(α)⟩ + Ĥ_β|ψ_β(α)⟩ + Ĥ_int(α, β)|ψ_β(α)⟩ = 0 (A3)

Identifying E_α = iℏ ∂_α (the clock energy as the generator of kernel-time translation) and rearranging:

iℏ ∂_α|ψ_β(α)⟩ = −(Ĥ_β + Ĥ_int(α, β))|ψ_β(α)⟩ ≡ H_rel(α|β)|ψ_β(α)⟩ (A4)

This is the relational Schrödinger equation (7) of the main text. The relational Hamiltonian H_rel(α|β) = −(Ĥ_β + Ĥ_int(α,β)) is entirely determined by the kernel K(α, β) through Ĥ_int(α,β) = −iℏ log K(α,β). No external time parameter appears. Time evolution is generated by the kernel transaction. This completes the derivation of Theorem 3.1. The derivation follows the standard Page–Wootters argument [22,23] but with the crucial step of identifying the interaction Hamiltonian as the kernel logarithm; a step available in KFA from axiom K3 (composability implies the existence of a kernel Lie algebra, and thus a kernel generator).

APPENDIX B: Spin Foam Amplitude as a Kernel Partition Function

Definition B.1 (Kernel Partition Function).

In KFA [1, Sec. 8], the kernel partition function Z[K] is defined as the sum over all kernel configurations {K(α, β)}_{adjacent (α,β)} weighted by the product of kernel amplitudes:

Z[K] = ∑_{{kernel configs}} ∏_{adjacent (α,β)} K(α, β)

The kernel partition function is the generating functional for all kernel correlation functions; the fundamental object of the KFA statistical mechanics of kernels.

We now show that the EPRL/FK spin foam amplitude is a kernel partition function in the sense of Definition B.1. The EPRL amplitude for a closed spin foam complex σ (with no boundary) is:

Z_EPRL(σ) = ∑_{{jf} on f ∈ σ} ∑_{{ie} on e ∈ σ} ∏_{f ∈ σ} A_f(j_f) ⋅ ∏_{e ∈ σ} A_e(i_e) (B1)

Identifying: (i) the kernel configuration = the spin labeling {j_f, i_e} of all faces and edges; (ii) the kernel amplitude for adjacent 4-simplices α, β = A_f(j_f) · A_e(i_e) where f and e are the face and edge shared by α and β; (iii) the sum over kernel configurations = the sum over {j_f, i_e}: then (B1) is exactly Z[K] of Definition B.1 for the gravitational kernel K(α, β) = A_f(j_f, i_e). The EPRL spin foam amplitude is a kernel partition function summing over all kernel configurations (spin labelings) weighted by kernel transaction amplitudes (face and edge amplitudes).

The group field theory (GFT) formulation [10] extends this to a sum over all spin foam complexes σ:

Z_GFT = ∑_σ λ^V(σ) / sym(σ) · Z_EPRL(σ) (B2)

where V(σ) is the number of vertices in σ, sym(σ) is the symmetry factor, and λ is the GFT coupling. This is the kernel grand canonical partition function: a sum over all kernel graph topologies (σ) and all kernel configurations ({j_f, i_e}), weighted by the kernel amplitude and a topological fugacity λ^V. GFT is therefore the field theory of gravitational KFA kernels: it implements the full kernel partition function, including the sum over spacetime topologies. This is why GFT is the most complete existing implementation of the KFA meta-architecture in quantum gravity; it encompasses both the kernel content (EPRL amplitude) and the topological sum (sum over σ), approximating the full KFA kernel RG flow.

APPENDIX C: Comparison of Optical and Gravitational KFA Instantiations

Appendix Table C. Full Side-by-Side Comparison of the Optical [2] and Gravitational (LQG/Spin Foams/CDT) KFA Instantiations.

KFA Element	Optical Instantiation [2]	Gravitational Instantiation
Primitive entity	Dielectric interface / thin film layer	Spin-network node (LQG) / 4-simplex (SF, CDT)
Kernel formula K(α,β)	M(α,β) = [cosδ, (i/p)sinδ; ip sinδ, cosδ] ∈ SL(2,ℂ)	he = P exp(∫e A) ∈ SU(2); Af(jf, ie); exp(iSRegge/ℏ)
Kernel group	SL(2,ℂ) (complex symplectic)	SU(2) (compact); Spin(3,1) (Lorentz group in SF)
Seed of relation formula	Hopt(α|β) = −iℏc log M(α,β) [2, Eq. 12]	Hrel(α|β) = −(Ĥβ + Ĥint(α,β)) [Eq. A4]
Relational generator eigenvalues	Optical phase δ = nkd (continuous)	Area A = 8πγlP2√(j(j+1)) (discrete, quantized)
Coarse-graining map Rb	Product of n transfer matrices: Meff = Mn…M1	Composition of n holonomies: hblock = hn…h1
IR fixed point	Homogeneous effective medium: MIR = [1, 0; 0, 1] (trivial)	Smooth spacetime manifold: KIR = exp(iSEH/ℏ) · gμν
UV fixed point	Photonic bandgap center: MUV = [−1,0;0,−1] (Bloch resonance)	Quantum foam: KUV = non-Gaussian gravitational fixed point
Connective tissue formula	Berry phase φBerry = ∫ A · dk (geometric phase in k-space)	Wilson loop Wγ = Tr P exp(∮γ A) ∈ ℝ
Connective tissue group	U(1) (Abelian: Berry phase is a real number)	SU(2) (non-Abelian: holonomy is a matrix)
Domain definition	Photonic crystal domain: region of periodic M with bandgap	Classical spacetime domain: region of spin network at KIR
Domain wall	Photonic bandgap edge; interface between photonic crystal and free space	Black hole horizon; Big Bang singularity; inflationary domain wall
Observer / clock	Reference dielectric layer α (optical clock); layer index as time	Clock spin-network node α (Page–Wootters clock); kernel index as time
Emergent medium	Effective optical medium: refractive index profile neff(x)	Classical spacetime: Lorentzian metric gμν(x)
Key equation	Relational optical Hamiltonian [2, Eq. 15]	Relational Wheeler–DeWitt (Eq. A4)
Problem of time analog	Propagation direction ambiguity (resolved by clock layer [2])	Wheeler–DeWitt time problem (resolved by clock stratum, Theorem 3.1)
Experimental signature	Gyroscope holonomy phase; transmission spectra; photonic bandgap [2]	CMB power-law correlations (Eq. 23); GW holonomy background (Eq. 24)
RG beta function	βn = b dneff/db (effective index running)	βK = b dKb/db (Eq. 13)

REFERENCES

1. D. Costello et al., “Kernel-First Architecture: A First-Principles Theory of Relational Structure, Emergent Media, and Cross-Stratum Coherence,” Foundations of Physics 56, 012 (2026). [KFA foundational paper]
2. D. Costello et al., “Relational Optical Hamiltonians, Gyroscope Precession Mapping, and Renormalization Group Coarse-Graining in Stratified Media: A Kernel-First Architecture,” Optics Express 34, 031405 (2026). [KFA optical paper]
3. M. H. Goroff and A. Sagnotti, “The ultraviolet behavior of Einstein gravity,” Nuclear Physics B 266, 709–736 (1986).
4. S. Deser and P. van Nieuwenhuizen, “One-loop divergences of quantized Einstein-Maxwell fields,” Physical Review D 10, 401 (1974).
5. J. Polchinski, String Theory, Vols. I and II (Cambridge University Press, Cambridge, 1998).
6. A. Ashtekar, “New variables for classical and quantum gravity,” Physical Review Letters 57, 2244 (1986).
7. A. Ashtekar, “New Hamiltonian formulation of general relativity,” Physical Review D 36, 1587 (1987).
8. C. Rovelli and L. Smolin, “Loop space representation of quantum general relativity,” Nuclear Physics B 331, 80–152 (1990).
9. T. Thiemann, Modern Canonical Quantum General Relativity (Cambridge University Press, Cambridge, 2007).
10. A. Perez, “The spin foam approach to quantum gravity,” Living Reviews in Relativity 16, 3 (2013).
11. J. C. Baez, “An introduction to spin foam models of quantum gravity and BF theory,” Lecture Notes in Physics 543, 25–93 (2000).
12. M. P. Reisenberger and C. Rovelli, “‘Sum over surfaces’ form of loop quantum gravity,” Physical Review D 56, 3490 (1997).
13. J. Ambjorn, J. Jurkiewicz, and R. Loll, “Emergence of a 4D world from causal quantum gravity,” Physical Review Letters 93, 131301 (2004).
14. J. Ambjorn, J. Jurkiewicz, and R. Loll, “Reconstructing the universe,” Physical Review D 72, 064014 (2005).
15. J. Ambjorn, A. Goerlich, J. Jurkiewicz, and R. Loll, “Nonperturbative quantum gravity,” Physics Reports 519, 127–210 (2012).
16. C. Rovelli and L. Smolin, “Discreteness of area and volume in quantum gravity,” Nuclear Physics B 442, 593–619 (1995).
17. A. Ashtekar and J. Lewandowski, “Quantum theory of geometry. I: Area operators,” Classical and Quantum Gravity 14, A55 (1997).
18. J. Engle, E. Pereira, C. Rovelli, and E. Livine, “LQG vertex with finite Immirzi parameter,” Nuclear Physics B 799, 136–149 (2008). [EPRL model]
19. L. Freidel and K. Krasnov, “A new spin foam model for 4D gravity,” Classical and Quantum Gravity 25, 125018 (2008). [FK model]
20. K. V. Kuchař, “Time and interpretations of quantum gravity,” in Proceedings of the 4th Canadian Conference on General Relativity, eds. G. Kunstatter, D. Vincent, J. Williams (World Scientific, Singapore, 1992), pp. 211–314.
21. B. Dittrich, “Diffeomorphism symmetry in quantum gravity models,” Advanced Science Letters 2, 151–163 (2009); B. Dittrich et al., “Coarse graining methods for spin net and spin foam models,” New Journal of Physics 14, 035008 (2012).
22. D. N. Page and W. K. Wootters, “Evolution without evolution: Dynamics described by stationary observables,” Physical Review D 27, 2885 (1983).
23. V. Giovannetti, S. Lloyd, and L. Maccone, “Quantum time,” Physical Review D 92, 045033 (2015).
24. A. Ashtekar, “New variables for classical and quantum gravity,” Physical Review Letters 57, 2244 (1986). [see [6] above]
25. J. F. Barbero, “Real Ashtekar variables for Lorentzian signature space-times,” Physical Review D 51, 5507 (1995).
26. G. Immirzi, “Real and complex connections for canonical gravity,” Classical and Quantum Gravity 14, L177 (1997).
27. G. Immirzi, “Quantum gravity and Regge calculus,” Nuclear Physics B (Proceedings Supplement) 57, 65–72 (1997). [Barbero–Immirzi parameter]
28. C. Rovelli, “Relational quantum mechanics,” International Journal of Theoretical Physics 35, 1637–1678 (1996).
29. B. S. DeWitt, “Quantum theory of gravity. I. The canonical theory,” Physical Review 160, 1113 (1967).
30. J. A. Wheeler, “Superspace and the nature of quantum geometrodynamics,” in Batelle Rencontres, eds. C. DeWitt and J. A. Wheeler (Benjamin, New York, 1968), pp. 242–307.
31. R. Penrose, “Angular momentum: An approach to combinatorial spacetime,” in Quantum Theory and Beyond, ed. T. Bastin (Cambridge University Press, Cambridge, 1971).
32. C. Rovelli, Quantum Gravity (Cambridge University Press, Cambridge, 2004).
33. A. Connes and C. Rovelli, “Von Neumann algebra automorphisms and time-thermodynamics relation in generally covariant quantum theories,” Classical and Quantum Gravity 11, 2899 (1994).
34. R. Gambini and R. A. Porto, “Relational time in generally covariant quantum systems: Four models,” Physical Review D 63, 105014 (2001).
35. M. Reuter, “Nonperturbative evolution equation for quantum gravity,” Physical Review D 57, 971 (1998).
36. M. Reuter and F. Saueressig, “Quantum Einstein gravity,” New Journal of Physics 14, 055022 (2012).
37. E. Verlinde, “On the origin of gravity and the laws of Newton,” Journal of High Energy Physics 2011, 029 (2011).
38. N. Aghanim et al. (Planck Collaboration), “Planck 2018 results. VI. Cosmological parameters,” Astronomy and Astrophysics 641, A6 (2020).
39. P. Amaro-Seoane et al. (LISA Collaboration), “Laser Interferometer Space Antenna,” arXiv:1702.00786 [astro-ph.IM] (2017).
40. G. Vidal, “Entanglement renormalization,” Physical Review Letters 99, 220405 (2007). [MERA]
41. B. Swingle, “Entanglement renormalization and holography,” Physical Review D 86, 065007 (2012).
42. J. M. Maldacena, “The large N limit of superconformal field theories and supergravity,” International Journal of Theoretical Physics 38, 1113 (1999). [AdS/CFT]
43. B. Czech, J. L. Karndel, S. McCandlish, and J. Sully, “Integral geometry and holography,” Journal of High Energy Physics 2015, 175 (2015). [holographic reconstruction]
44. M. V. Berry, “Quantal phase factors accompanying adiabatic changes,” Proceedings of the Royal Society of London A 392, 45–57 (1984).
45. B. Simon, “Holonomy, the quantum adiabatic theorem, and Berry’s phase,” Physical Review Letters 51, 2167 (1983).
46. M. F. Atiyah, “Topological quantum field theory,” Publications Mathématiques de l’IHÉS 68, 175–186 (1988).

^* Correspondence: Daryl Costello, Rosendale, NY 12493, USA. Manuscript prepared: October 18, 2026. This work received no specific grant funding from any public, commercial, or not-for-profit agency. The authors declare no conflicts of interest. Data sharing: this is a theoretical paper; no data were generated.

Paper III in the KFA Series. Papers [1] and [2] are companion works cited throughout. Cross-references to [1] and [2] are essential for full context; this paper is self-contained but substantially enriched by the preceding papers in the series.