Section 3 — Emergent Spatial Structure and the Relational Manifold (TFP) (v12.11)

Section 3 — Emergent Spatial Structure and the Relational Manifold

Temporal Flow Physics v12.11

John Gavel

3.0 Overview

Section 2 established all local dynamics within the minimal adjacency structure — each site with exactly $K=2$ neighbors. Section 2 delivered binary states and update dynamics, proto-mass $M_i$ and proto-momentum $p_i$, accumulated differences $A_{ij}$ as exact integer summations, temporal correlations $C_{ij}(\tau)$, operational relational cost $d_{ij} = A_{ij}/t$, causal ordering from event sequence, and the mediated lag: minimum 2 ticks for non-adjacent influence.

Section 3 derives all spatial structure from a single generating object:

the adjacency transfer operator $T$ acting on the relational coordination graph.

The coordination number $K$ and the graph's identity are not assumed here — they are derived in §3.3 from the axioms of Sections 1 and 2. Once $K=12$ is established, $T$ specializes to the icosahedral graph, and all spatial structure, recursion depth, dimensionality, routing hierarchy, and $\phi_1$ scaling emerge from the spectrum of $T$. The organizing principle:

The adjacency transfer operator $T$ on the $K=12$ icosahedral graph is the minimal closed relational system whose spectrum generates $K=12$ coordination, $D=3$ dimensionality, and $\phi_1$ as its dominant recursive scaling ratio — required by the condition that $T$ admit a non-degenerate Perron-Frobenius eigenmode with recursive closure and finite spectral embedding.

Section 3 proceeds in nine stages:

Why $K=2$ is insufficient when chains interact (§3.1)
The transfer operator $T$ defined generally; finite capacity motivation for what §3.3 will prove (§3.2–3.2.1)
The theorem establishing $K=12$ uniquely; $T$ specialized to the icosahedral graph (§3.3)
Emergent dimensionality $D=3$ from $T$'s invariant subspace, by two independent routes (§3.4)
Spectral structure: adjacency eigenvalues, Laplacian eigenspectrum, $\phi_1$, $\Psi_{\text{sph}}$ (§3.5)
Derived constants: $\pi_d$, $\phi_d$, SIMPLEX, $\pi_{\text{eff}}(12)$, icosahedral geometry (§3.6–3.9)
Routing suppression as spectral decay of $T$ — the structural foundation of the routing hierarchy (§3.10)
The $A \to B \to C$ routing geometry — the formal routing hierarchy lemma (§3.11)
Phase coherence, budget partition, chirality, distance (§3.12–3.15)

All quantities remain dimensionless throughout. Physical calibration is deferred to Section 5.

3.1 Why the Minimal Structure is Insufficient

In Section 2, each site has exactly two neighbors. The only closed loop is the reflection $A \to B \to A$, whose holonomy is trivially zero. A single $K=2$ chain can propagate differences but cannot locate where those differences are. It cannot distinguish direction, establish a reference frame, or triangulate position.

The adjacency operator on a $K=2$ chain produces no nontrivial eigenmodes, no directional asymmetry, no recursion depth, no stable phase structure. The problem appears as soon as two chains interact: when two $K=2$ chains share a site $B$, there is no way to distinguish which disagreement to serve first, or whether two simultaneous disagreements are related.

Triadic Witnessing Principle. For a relation $A \leftrightarrow B$ that must be verified under interaction, introduce a third site $C$ such that $A \to C$ and $B \to C$ each carry independent projections of the constraint. $C$ enforces consistency without originating any new constraint. This is the minimum structure that permits conflict resolution. $K=2$ cannot support it — a chain has no free neighbor to serve as witness.

Three relational requirements emerge that $K=2$ cannot meet:

First: Two interacting chains create a mediated relation between non-shared sites that is not directly observable and requires independent verification.
Second: Verification requires each edge to appear in at least two independent ternary constraints. A single $K=2$ chain provides only one constraint per edge.
Third: Satisfying two independent ternary constraints per edge requires more than two neighbors. $K=2$ is structurally incapable of this.

This failure is structural, not incidental. It follows from Axioms 1–9. The question becomes: what is the unique $K$ that is both necessary for local determinacy and sufficient to close into a self-consistent relational manifold?

3.2 The Transfer Operator T

Let $G$ be a coordination graph on $K$ vertices, with $K$ to be determined in §3.3. Define the adjacency transfer operator $T$ acting on vertex amplitudes $\psi_i$:

$$ T\psi_i = \sum_{j \in N(i)} \psi_j \quad \text{[D]} $$

This is the graph adjacency operator on the coordination structure. $T$ is a $K \times K$ real symmetric matrix. The coordination number $K$ and the specific graph structure are not assumed here. $K$ is established uniquely in §3.3, after which $T$ specializes to the icosahedral graph and all spectral analysis follows.

The total directed relational bandwidth, once $K$ is known:

$$ H = K(K-1) \quad \text{[D]} $$

counts all ordered pairs $(i,j)$ with $i \neq j$. Once §3.3 establishes $K=12$:

$$ H = 12 \times 11 = 132 \quad \text{[D]} $$

3.2.1 Finite Capacity: Motivation for the K=12 Derivation

From §2.10, each site possesses finite relational capacity. In the $K=2$ structure, $K_{1D} = 2$ with only one relational update resolvable per tick, producing a stutter sink when capacity is exceeded:

$$ s_i(t) = \kappa_s \cdot \delta_i(t) \cdot p_i(t) $$

When $\delta_i(t) = 0$, relational updates are fully resolved. When $\delta_i(t) > 0$, unresolved updates accumulate as temporal latency.

The key structural constraint is:

Finite relational capacity forces discrete resolution structure in comparison dynamics.

In the large-scale limit, repeated stutter-constrained resolution drives the substrate toward stable comparison closure patterns — fixed-point coordination counts in which all local ternary constraints are satisfied and no further closure is required. This motivates what §3.3 will prove: the comparison field saturates at a unique coordination number $K$ where exactly this condition is met. Section 3.3 identifies $K=12$ as this unique termination point through a three-part theorem.

3.3 K=12: The Unique Domain of T

Theorem ($K=12$ uniqueness). Let $G$ satisfy Axioms A1–A9. The unique relational coordination number that is simultaneously (a) the minimum required for local determinacy, (b) the maximum consistent with geometric integrity, and (c) the unique self-consistent, self-hosting fixed point of the joint $(K,D)$ system, is $K=12$.

3.3.1 — Forced Coordination: Proof that K=12 is the Minimum Uniform Solution

Theorem. Let $G_K$ be the local coordination shell of a site in Temporal Flow Physics, represented as a vertex-transitive graph on $K$ vertices. No $K < 12$ simultaneously satisfies the following five structural conditions, each derived from TFP primitives:

Condition	Statement	Primitive Source
A (NCF)	Shell contains a 4-site non-coplanar reference frame spanning 3D relational volume	Axiom 7 — closure rank $\ge 3$ DOF (§3.4.2)
B (DC)	Every edge participates in $\ge 2$ independent triangular closure loops	Axioms 6, 9 — redundancy prevents degeneracy
C (MTE-Uniform)	Graph is vertex-transitive AND every edge lies in exactly 2 triangles	Axiom 9 uniform capacity + §1.8 Summation Principle
D (Verification)	At least one verification site exists beyond the 4-site frame: $K - 4 \ge 1$	Axiom 6 — isolated differences non-determinative
E (Fixed-Point)	Self-consistent $(K,D)$ closure: $K_{\text{kiss}}(D(K)) = K$	§3.3.3 — geometric integrity from relational closure

The proof proceeds in five steps. Steps 1a and 1b below replace the previous appeal to external graph classification with a derivation from TFP first principles.

Step 1a — Regularity and Even-$K$ from Bilateral Exclusivity
From Axiom 5 (Local Update Exclusivity) and Axiom 9 (Finite Relational Capacity): in any single substrate tick, a site can resolve with exactly one neighbor. Any two sites resolving their difference in the same tick form a closed bilateral pair — no third site can participate in that resolution event during that tick. The substrate's active update structure at any tick is therefore a matching: a set of edges with no shared endpoints.

Regularity. For a motif to be stable (§4.1, temporal recurrence condition), its resolution structure must recur uniformly across ticks. By vertex-transitivity (Condition C), every site faces the same relational demand per cycle. If any site had higher degree than another, it would face strictly more resolution events per cycle than its neighbors, accumulating unresolved tension at a rate that grows without bound — violating the stability requirement of §4.1 and the finite capacity of Axiom 9. Therefore every site has the same degree $r$: the graph is $r$-regular. [D]

Even $K$. For a uniform stable motif, the matching that fires each tick must cover every vertex — otherwise some fixed fraction of vertices is idle each tick, neither resolving nor being resolved. By vertex-transitivity this idle fraction is the same for every vertex, meaning every vertex carries permanently unresolved tension at each tick. By Axiom 9, unresolved tension accumulates as latency rather than closing into a stable pattern — the motif does not stabilize. A matching that covers every vertex is a perfect matching, which requires $K$ to be even. Therefore:

$$ K \text{ must be even for any stable uniform motif.} \quad \text{[D]} $$

This eliminates $K \in \{5, 7, 9, 11\}$ in a single step, without case analysis. The remaining candidates are $K \in \{4, 6, 8, 10, 12, \dots\}$.

Step 1b — Triangle Count Integrality and Candidate Reduction
Counting identity. In a $K$-vertex $r$-regular graph where every edge lies in exactly 2 triangles (Condition C), count (edge, triangle) incidence pairs two ways:

Each of the $E = Kr/2$ edges is in exactly 2 triangles $\implies$ total incidence pairs $= Kr$
Each triangle contains 3 edges $\implies$ total incidence pairs $= 3T$

Therefore:

$$ T = \frac{Kr}{3} \quad \text{[D]} $$

For $T$ to be a non-negative integer, 3 must divide $Kr$. Since $K$ is already constrained to even values, this requires $3 \mid Kr$ for each candidate $K$.

Triangles per vertex. Each vertex $v$ has $r$ neighbors. Each triangle through $v$ uses exactly 2 of those neighbors. The total triangle-vertex incidences $= 3T = Kr$. Dividing by $K$ vertices:

$$ \text{triangles through } v = r \quad \text{(uniform, by vertex-transitivity)} \quad \text{[D]} $$

The $K_4$ saturation lemma. For any vertex $v$ with $r = 3$: $v$ has neighbors $a, b, c$. Edge $va$ is in exactly 2 triangles. The only vertices adjacent to both $v$ and $a$ are drawn from $\{b, c\}$ ($v$ has only 3 neighbors). Therefore at least one of $b, c$ is adjacent to $a$ — and by identical argument on edges $vb$ and $vc$, all three pairs $\{a,b\}, \{b,c\}, \{a,c\}$ must be edges. So $a, b, c$ are mutually adjacent, forming $K_4$ with $v$.

Proof that all three pairs are forced: $v$ has exactly $r=3$ triangles through it. Each triangle through $v$ uses one edge among $\{a,b,c\}$. There are exactly 3 edges available among $\{a,b,c\}$: namely $ab, bc, ac$. Each must be used exactly once (to produce 3 distinct triangles through $v$). All three edges must exist. [D]

Consequence: in any $r=3$ candidate, every vertex sits in exactly one $K_4$, and those $K_4$ subgraphs must partition the vertex set. Partitioning $K$ vertices into $K_4$ subgraphs requires $4 \mid K$.

Step 1c — Elimination of K = 8
For $K=8$, the integrality condition $3 \mid 8r$ forces $3 \mid r$, giving $r \in \{3, 6\}$ ($r < 8, r > 0$).

Case $r=3, K=8$.
By the $K_4$ saturation lemma, every vertex sits in a $K_4$ and those $K_4$ subgraphs partition the vertex set. Partitioning 8 vertices into $K_4$ subgraphs requires $4 \mid 8$ ✓ — two disjoint $K_4$ subgraphs.
Now apply Condition B to an edge $ab$ where $a$ and $b$ are in different $K_4$ subgraphs. The two triangles through $ab$ must be edge-disjoint. The common neighbors of $a$ and $b$ must both be in the same $K_4$ as... but $a$ and $b$ are in different $K_4$ subgraphs, each of which is a clique on 4 vertices. The neighbors of $a$ are the other 3 vertices of $a$'s $K_4$. The neighbors of $b$ are the other 3 vertices of $b$'s $K_4$. Since the two $K_4$ subgraphs are disjoint (vertex-disjoint partition), $a$ and $b$ share no common neighbors. Therefore edge $ab$ is in 0 triangles, not 2. This violates Condition C ($\lambda=2$).
But wait — if the two $K_4$ subgraphs are vertex-disjoint, there are no edges between them at all in a 3-regular graph (each vertex's 3 neighbor slots are already filled by the other 3 vertices of its $K_4$). So $ab$ cannot be an edge between the two $K_4$ subgraphs. All edges lie within the two $K_4$ subgraphs, and the graph is disconnected. A disconnected graph cannot satisfy Condition A (no non-coplanar reference frame spans across disconnected components). $K=8, r=3$ eliminated by Condition A. [D]

Case $r=6, K=8$.
Each vertex is adjacent to 6 of the other 7 vertices, meaning each vertex has exactly 1 non-neighbor. The graph is $K_8$ minus a perfect matching (4 edges removed).
For any edge $uv$: the common neighbors of $u$ and $v$ are all vertices except $u, v$, the non-neighbor of $u$, and the non-neighbor of $v$. Since $u$ and $v$ each have exactly 1 non-neighbor and $u, v$ are themselves excluded:

$$ \text{common neighbors of } uv = 8 - 2 - (\text{at most } 2) = \text{at least } 4 $$

So $\lambda \ge 4 > 2$, violating Condition C. $K=8, r=6$ eliminated by Condition C. [D]
$K=8$ is fully eliminated. [D]

Step 1d — Elimination of K = 10
For $K=10$, the integrality condition $3 \mid 10r$ forces $3 \mid r$ (since $\gcd(10,3)=1$), giving $r \in \{3, 6, 9\}$.

Case $r=3, K=10$.
By the $K_4$ saturation lemma, the $K_4$ subgraphs must partition the vertex set. Partitioning 10 vertices into $K_4$ subgraphs requires $4 \mid 10$. But 4 does not divide 10. Therefore no valid $K_4$ partition exists, and no vertex-transitive $r=3$ graph on 10 vertices satisfies Condition C.
From TFP primitives directly: the uniform closure requirement of Axiom 9 demands that every vertex's local $K_4$ structure closes consistently across the full graph. With 10 vertices and groups of 4, there is always at least one vertex whose $K_4$ closure is incomplete — it cannot find 3 mutually adjacent partners that do not overlap with another vertex's $K_4$. The relational closure cannot be made uniform, so no stable motif forms. $K=10, r=3$ eliminated by Axiom 9 uniform closure requirement. [D]

Case $r=9, K=10$.
This is the complete graph $K_{10}$. Every edge is in $K-2 = 8$ triangles. $\lambda=8 \neq 2$, violating Condition C immediately. $K=10, r=9$ eliminated by Condition C. [D]

Case $r=6, K=10$.
$T = 10 \times 6 / 3 = 20$ triangles. Apply the strongly regular graph parameter relation. For a vertex-transitive graph with $K=10, r=6, \lambda=2$, the parameter $\mu$ (common neighbors of any non-adjacent pair) satisfies:

$$ r(r - \lambda - 1) = \mu(K - r - 1) $$ $$ 6 \times 3 = \mu \times 3 $$ $$ \mu = 6 \quad \text{[D]} $$

Now derive a contradiction from $\mu=6$ using only the relational budget.
Each vertex $w$ contributes to the common-neighbor count of every pair of its neighbors that are non-adjacent. Vertex $w$ has $r=6$ neighbors, giving $\binom{6}{2}=15$ pairs of neighbors. Each such pair $(u,v)$ that is a non-edge receives $w$ as a common neighbor. Summing over all 10 vertices:

$$ \sum_w \binom{r}{2} = 10 \times 15 = 150 \quad \text{total (non-edge, common-neighbor) incidences} \quad \text{[D]} $$

Count the same quantity from the non-edge side. Number of non-edges $= \binom{10}{2} - E = 45 - 30 = 15$. Each non-edge $(u,v)$ contributes $\mu(u,v)$ to the count. If $\mu = 6$ uniformly:

$$ \sum_{\text{non-edges}} \mu = 15 \times 6 = 90 \quad \text{[D]} $$

But $150 \neq 90$. The two counts of the same quantity disagree. This is a counting contradiction internal to the relational budget structure — no graph with $K=10, r=6, \lambda=2$ can have $\mu=6$, and no other $\mu$ satisfies the parameter relation.
Explicitly: the parameter relation forces $\mu=6$, but the budget identity forces $\mu=150/15=10$. These cannot both hold. $K=10, r=6$ eliminated by internal budget contradiction. [D]
$K=10$ is fully eliminated. [D]

Step 1e — Surviving Candidates
The complete elimination record:

$K$	Eliminated by	Step
5, 7, 9, 11	Odd $K$: no perfect matching, tension never clears uniformly	1a
8 ($r=3$)	Disconnected $K_4$ partition: Condition A fails	1c
8 ($r=6$)	$\lambda \ge 4 > 2$: Condition C fails	1c
10 ($r=3$)	$4 \nmid 10$: $K_4$ tiling impossible, Axiom 9 closure fails	1d
10 ($r=9$)	$K_{10}$ has $\lambda=8 \neq 2$: Condition C fails	1d
10 ($r=6$)	Budget contradiction $\mu=6$ vs $\mu=10$	1d

The surviving candidates satisfying all of Conditions A+B+C for $K \ge 5$, $K$ even, are exactly:

$K = 4$ (tetrahedron $K_4$)
$K = 6$ (octahedron graph)
$K = 12$ (icosahedral graph)

No external graph classification result is cited. Every elimination follows from Axioms 1–9 and Conditions A–E as derived from those axioms.
Steps 2 through 5 proceed as before: $K=4$ is eliminated by Condition D ($K-4=0$ verification sites), $K=6$ is eliminated by Condition E (fixed-point failure $K_{\text{kiss}}(3)=12 \neq 6$), and $K=12$ is confirmed to satisfy all five conditions.

Part I result: $K \ge 12$ from forced coordination. [D] ∎

Lemma 1 — $K = 8, r = 3$ Forces Disconnection and Violates Condition A
Statement. In any vertex-transitive, 3-regular coordination shell on $K = 8$ vertices satisfying Conditions A–C, every vertex lies in a unique $K_4$, and the graph necessarily decomposes into two disjoint $K_4$ components. Because disconnected components share no adjacency path, the 4-site non-coplanar reference frame required by Condition A cannot exist. Therefore $K = 8, r = 3$ is eliminated.

Proof.

Condition C requires every edge to lie in exactly 2 triangles.
With $r = 3$, each vertex $v$ has neighbors $a, b, c$.
By the $K_4$ saturation lemma, all three edges $ab, bc, ac$ must exist. Thus $\{v, a, b, c\}$ forms a $K_4$.
Because $v$ has degree 3, all of its neighbor slots are used inside this $K_4$. No vertex in this $K_4$ can have neighbors outside it.
Vertex-transitivity forces the same structure for all vertices.
Therefore the 8 vertices split into exactly two disjoint $K_4$ components.

Sites in disconnected components share no adjacency path, so no relational difference $A_{ij}$ is defined between them (see §2.5.1). The 4-site non-coplanar reference frame of Condition A requires all 6 pairwise relational differences to exist and be independent. This is impossible when the shell breaks into isolated components.

Conclusion: $K = 8, r = 3$ is eliminated by Condition A (NCF). Done.

(Bridging Sentence) $K = 9$ is already eliminated by Step 1a (odd $K$, no perfect matching). The next even candidate is $K = 10$.

Lemma 2 — $K = 10, r = 6$ Produces an Internal $\mu$-Contradiction
Statement. In any vertex-transitive, 6-regular coordination shell on $K = 10$ vertices satisfying Condition C ($\lambda = 2$), the relational budget forces two incompatible values of $\mu$. No such graph can exist.

Proof.

Step 1 — Compute $T$ from Condition C
$T = (K \cdot r) / 3 = (10 \cdot 6) / 3 = 20$ triangles.

Step 2 — $\mu$ is uniform because the path-counting identity forces it
The number of length-2 paths between any non-adjacent pair $(u, v)$ is given by the identity:

$$ r(r - \lambda - 1) = \mu(K - r - 1) $$

Since $K = 10, r = 6$, and $\lambda = 2$ are all uniform, the left side is a constant:

$$ 6 \times (6 - 2 - 1) = 6 \times 3 = 18 $$

Therefore the right side $\mu \times (10 - 6 - 1)$ must also be constant, which forces $\mu$ to be the same for every non-edge. No assumption about automorphism-group transitivity is needed.

Step 3 — Solve for $\mu$
$6 \times 3 = \mu \times 3 \implies \mu = 6$

Step 4 — Count common-neighbor incidences from the vertex side
Each vertex has $\binom{6}{2} = 15$ unordered neighbor-pairs.
Over 10 vertices: $10 \times 15 = 150$ incidences.

Step 5 — Count the same quantity from the non-edge side
Number of non-edges $= \binom{10}{2} - E = 45 - 30 = 15$.
If $\mu = 6$ uniformly, total incidences $= 15 \times 6 = 90$.

Step 6 — Contradiction
$150 \neq 90$.
The relational budget demands $\mu = 10$, while the path-counting identity demands $\mu = 6$. Both cannot be true.

Conclusion: $K = 10, r = 6$ is eliminated by internal budget contradiction. Done.

3.3.2 Upper Bound: Geometric Integrity ($K \le 12$)

The geometric integrity condition requires that $H = K(K-1)$ directed pairs among $K$ nodes resolve entirely within the coordination shell — no pair references structure outside it.

The coordination shell of $K$ sites contains $K(K-1)$ directed comparison pairs. For the comparison structure to be self-contained, the shell's internal geometry must accommodate exactly $H$ directed pairs without overflow. The candidate structure established by §3.3.1 for $K=12$ is the icosahedral graph, which has 12 vertices and therefore:

$$ K=12: H = 12 \times 11 = 132, \quad \text{self-closure ratio} = 132/132 = 1.000 \quad \text{[D, exact]} $$

For $K=13$: $H = 13 \times 12 = 156$. Any 13-site coordination shell must accommodate 156 directed pairs. The icosahedral structure accommodates exactly 132. A 13-site shell would require $156 - 132 = 24$ directed pairs to reference structure outside — geometrically incoherent:

$$ K=13: H = 156, \quad \text{overflow} = 24 \text{ pairs exceed shell capacity} \quad \text{[D]} $$

This is a counting identity, not a geometric constraint. No embedding or compactification can reduce the 24 surplus pairs — they are simply the excess of $K(K-1)$ over the 132 self-closure capacity. The shell claims relations it cannot contain.

The structural self-exclusion: $K^2 - H = K^2 - K(K-1) = K = 12$ directed pairs are excluded from $H$ — each node cannot handshake with itself. This gives the substrate efficiency $\eta_{\text{sub}} = (K-1)/K = 11/12$ and the self-exclusion fraction $1/K = 1/12$.

Part II result: $K \le 12$ from geometric integrity. [D]

3.3.3 Uniqueness: Self-Consistent, Self-Hosting Fixed Point

Parts I and II together establish $K=12$ as the unique integer satisfying both bounds. The deeper reason is that $(K=12, D=3)$ is the unique non-trivial self-consistent, self-hosting fixed point of the coupled $(K,D)$ system.

Define two structural maps:

$$ K_{\text{kiss}}(D) = \text{kissing number in } D \text{ dimensions} $$ $$ D(K) = \text{rank of relational axis structure (derived in §3.4)} $$

Relevant kissing numbers:

$D=1$: $K_{\text{kiss}} = 2 \implies K=2$, the degenerate chain of Section 2
$D=2$: $K_{\text{kiss}} = 6 \implies$ candidate $(K=6, D=2)$, checked below
$D=3$: $K_{\text{kiss}} = 12 \implies$ candidate $(K=12, D=3)$
$D=4$: $K_{\text{kiss}} = 24 \implies$ candidate $(K=24, D=4)$

Checking $(K=6, D=2)$:
$H = 6 \times 5 = 30 < 132$: self-hosting is satisfied. However, $D=2$ cannot satisfy the non-coplanar reference frame requirement of Step 1. A non-coplanar set of 4 points requires $D \ge 3$ — four points cannot be non-coplanar in a 2-dimensional space. Therefore the local determinacy condition fails for $D=2$: the reference frame cannot eliminate 3D parity ambiguity. $(K=6, D=2)$ is excluded. [D]

Checking $(K=24, D=4)$:
$H = 24 \times 23 = 552$. The $K=12$ substrate accommodates $H=132$ directed pairs. Surplus $= 552 - 132 = 420$ pairs. $K=24$ cannot be instantiated inside a $K=12$ substrate — it is not self-hosting. No embedding reduces this surplus; it is a counting identity. [D]

$(K=2, D=1)$: degenerate. Excluded by the triadic witnessing failure of §3.1. [D]

$(K=12, D=3)$ is the unique non-trivial self-consistent, self-hosting fixed point. [D]

§3.3.3a Projection-Invariant Fixed Point

The fixed point established in §3.3.3 shows that $(K=12, D=3)$ satisfies $K_{\text{kiss}}(D(K)) = K$. A natural question follows: does this property survive arbitrary relational projections of the icosahedral structure — coarse-grainings, symmetry reductions, dimensional projections — or is it an artifact of the full 12-vertex representation? The answer is that $K=12$ is robust: every admissible projection of the $K=12$ relational structure lands back at the same fixed point.

Definition (admissible projection). A linear map $\Pi: \mathbb{R}^{12} \to \mathbb{R}^{12}$ is admissible ($\Pi \in \mathbb{P}_{\text{closure}}$) if it satisfies two conditions:

(a) $A_5$-equivariance. $\Pi$ commutes with the action of the icosahedral group $A_5 \cong I$ on the 12-vertex coordinate space: $\Pi(g \cdot v) = g \cdot \Pi(v)$ for all $g \in A_5$.
(b) Non-degenerate closure rank. The projected structure preserves 3D relational determinacy in the sense of §3.4.2 Route B: the kernel of the ternary constraint system over $\mathbb{Z}_2$ on the projected vertex set has dimension $\ge 3$.

Condition (a) rules out projections that break icosahedral symmetry; condition (b) rules out projections that collapse to lower-dimensional relational closure ($D < 3$), which would violate the non-coplanar reference frame requirement of §3.3.1 Condition A.

Note on excluded projections. The $D_5$ quotient of §4Q.4 — which reduces $K=12$ to $K_{\text{flow}}=10$ by removing the two color-lock channels — is deliberately excluded from $\mathbb{P}_{\text{closure}}$. It satisfies neither condition: it breaks $A_5$-equivariance ($D_5 \subset A_5$ but the quotient is not $A_5$-equivariant) and it reduces relational bandwidth below the 3D determinacy threshold. $K_{\text{flow}}=10$ is a physically meaningful derived structure, but it is a sector projection within the $K=12$ substrate, not an admissible relational projection of the substrate itself. The lemma below concerns the substrate's self-consistency under projection; the quark sector structure of §4Q is a separate derivation.

Lemma (Projection-Invariant Fixed Point). Define the relational axis rank of $\Pi$ as:

$$ D(\Pi) := \text{rank of } \Pi \text{ restricted to the } T_1 \text{ polar eigenspace of } A_{12} $$

and the map $F_\Pi(K) := K_{\text{kiss}}(D(\Pi))$. Then:

$$ \forall \Pi \in \mathbb{P}_{\text{closure}}: \quad F_\Pi(12) = 12 $$

and $K=12$ is the unique value with this property.

Proof.

Step 1 — $D(\Pi) = 3$ for all $\Pi \in \mathbb{P}_{\text{closure}}$.
The $T_1$ polar eigenspace of $A_{12}$ has algebraic multiplicity exactly 3 (§3.5.1) and transforms as the unique 3-dimensional polar irrep of $A_5$ under the icosahedral group action. By Schur's lemma, any $A_5$-equivariant endomorphism restricted to this eigenspace is a scalar multiple of the identity on that space. An admissible projection $\Pi$ therefore either preserves the $T_1$ eigenspace intact (rank 3) or annihilates it entirely (rank 0).

Annihilation is ruled out by condition (b): the $T_1$ eigenspace captures 100% of the spatial vertex position information and 100% of the antisymmetric edge differences on the icosahedral graph (§3.4.1 numerical verification). A projection that annihilates $T_1$ reduces to a structure with no spatial axes — the ternary constraint kernel over $\mathbb{Z}_2$ collapses to dimension 0, violating the $\ge 3$ requirement of condition (b).

Therefore $D(\Pi) = 3$ for all $\Pi \in \mathbb{P}_{\text{closure}}$. [D]

Step 2 — $K_{\text{kiss}}(3) = 12$.
The kissing number in $D=3$ dimensions is 12, as used in §3.3.3. [D]

Step 3 — $F_\Pi(12) = 12$.
$F_\Pi(12) = K_{\text{kiss}}(D(\Pi)) = K_{\text{kiss}}(3) = 12$ for all $\Pi \in \mathbb{P}_{\text{closure}}$. [D]

Step 4 — Uniqueness.
From §3.3.3, the fixed-point table shows that $K=2$ is degenerate (excluded by triadic witnessing failure, §3.1), $K=6$ fails the non-coplanar frame condition in $D=2$ (§3.3.1 Condition A), and $K=24$ exceeds self-hosting capacity ($H=552 > 132$). No admissible projection can reach these alternatives from $K=12$ without first violating condition (a) or (b), since both conditions enforce the same constraints that exclude $K \neq 12$ in the original fixed-point argument. $K=12$ is therefore the unique projection-invariant fixed point. [D] ∎

What this adds to §3.3.3. The original fixed-point theorem establishes that $(K=12, D=3)$ is a self-consistent, self-hosting fixed point by checking specific candidates. The projection lemma strengthens this to a robustness claim: no symmetry-preserving, closure-preserving coarse-graining of the icosahedral structure can escape $K=12$. Any admissible projection of $A_{12}$ still returns the same coordination number under the $F_\Pi$ map. The fixed point is not an artifact of the specific 12-vertex representation — it is invariant under the full class of relational projections that preserve the substrate's structural commitments.

This property will be used in Section 4 when verifying that sector reductions (lepton antisymmetric channel $H/2$, quark routing $K_{\text{flow}}=10$) inherit their coupling structure from the parent $K=12$ substrate without re-deriving $K$ at each level. Those reductions are outside $\mathbb{P}_{\text{closure}}$ by the note above — they are not fixed points of $F_\Pi$ — but their parameters ($H/2=66$, $K_{\text{flow}}=10$) are derived from $K=12$ in §4Q.1 and §4.6.3, not from independent fixed-point arguments, which is consistent with their excluded status here.

3.3.4 Handshake Capacity

Once $K=12$ is established:

$$ H = K(K-1) = 12 \times 11 = 132 \quad \text{(total directed relational bandwidth)} \quad \text{[D]} $$ $$ H/2 = 66 \quad \text{(symmetric undirected pairs)} \quad \text{[D]} $$

$H/2 = 66$ carries metric, distance, and field propagation. The antisymmetric $H/2 = 66$ carries spin, chirality, and winding number (§4.6).

$T$ now specializes to the icosahedral graph: $G$ is the unique 12-vertex graph with every edge in exactly 2 triangular faces and uniform vertex degree 5. $T$ is the $12 \times 12$ icosahedral adjacency matrix. All subsequent spectral analysis refers to this specific $T$.

3.4 Emergent Dimensionality D=3

$D=3$ is derived from $K=12$ by two independent routes. Both are given in full; they cross-validate each other.

3.4.1 Route A: Axis Rank Derivation (Coordinate-Free)

This derivation uses no pre-existing notion of dimensionality, coordinates, or embedding. $D$ is defined as a relational rank.

Step 1 — Relational axes from antipodal pairs.
The icosahedral graph has diameter 3. The 12 vertices partition into 6 antipodal pairs: each vertex has exactly one vertex at maximum graph distance. In TFP, an antipodal pair defines a relational axis: the difference mode between two sites that cannot be resolved through a shorter path. There are exactly 6 such axes in the $K=12$ shell. [D]

Step 2 — Closure constraints filter relational modes.
The 6 axis modes span a 6-dimensional subspace of the 12-dimensional site state space. Icosahedral closure acts as a linear filter: only modes that transform under the graph's invariant symmetry survive as independent relational directions. Projecting the 6 axis modes onto the adjacency operator's invariant eigenspace ($\lambda = \sqrt{5} \approx 2.236$) yields exactly 3 independent components. Equivalently:

$$ D := \text{rank}(\text{projection of axis modes onto invariant subspace}) = 3 $$

No spatial assumption was made. The number 3 emerged from counting how many axis modes survive relational closure. [D]

Step 3 — Spectral cross-validation (pure graph theory).
The adjacency matrix spectrum contains an eigenspace at $\lambda = \sqrt{5}$ with algebraic multiplicity exactly 3. This eigenspace is the unique irreducible representation of the icosahedral symmetry group $I_h$ that transforms as a vector. The 6 antipodal difference modes project onto this 3-dimensional invariant subspace with numerical precision, with zero components orthogonal to it.

Numerical verification. Direct computation on the icosahedral adjacency matrix confirms the derivation. The 6 antipodal difference modes project onto the $\lambda = \sqrt{5}$ eigenspace with principal angle $\lt 2 \times 10^{-6} \text{°}$ (limited only by double-precision arithmetic), and the projected rank is exactly 3. The invariant subspace is fully contained within the span of the relational axes. [D]*

Result: $D = 3$, derived from $K=12$ closure constraints and spectral invariants. No spatial input was assumed. [D]

3.4.2 Route B: Closure Rank Derivation

Let $O$ be a site with coordination set $N(O)$. Binary relational differences $x_{ij} \in \{0,1\}$. Ternary closure constraint for any closed triple $\{a,b,c\}$:

$$ x_{ab} + x_{ac} + x_{bc} = 1 \pmod 2 $$

Introduce a fourth site $k$. The set $\{O,i,j,k\}$ has 6 difference variables: $x_{Oi}, x_{Oj}, x_{Ok}, x_{ij}, x_{ik}, x_{jk}$. Impose ternary constraints on all triples containing $O$:

$$ x_{Oi} + x_{Oj} + x_{ij} = 1 $$ $$ x_{Oi} + x_{Ok} + x_{ik} = 1 $$ $$ x_{Oj} + x_{Ok} + x_{jk} = 1 $$

Linear system over $\mathbb{Z}_2$: 6 variables, 3 independent constraints.

$$ \dim(\ker) = 6 - 3 = 3 $$

Exactly three independent relational degrees of freedom. $D = 3$. [D]

3.4.3 Connection Between Routes

Route A derives $D$ from the geometric symmetry of $K=12$ (axis rank = 3). Route B derives $D$ from the algebraic structure of ternary closure (kernel dimension = 3). They agree because the four-site closure $\{O,i,j,k\}$ in Route B is exactly the tetrahedron — 4 vertices, 6 edges, 3 independent ternary constraints — which is one face-structure of the icosahedral geometry identified in Route A.

Both routes give $D = 3$. The structural relationship between $D$ and the icosahedral shell:

$$ D = \frac{K}{D + 1} \quad \implies \quad 3 = \frac{12}{4} \quad \text{[D]} $$

where $D+1 = 4$ is the number of vertices in the 3D simplex (tetrahedron). Once the $\pi_d$ sequence is defined in §3.6.1 as $\pi_d = d+1$, this can be written $D = K/\pi_3$. The formula is stated here as a consequence of the two derivations; $\pi_3$ is defined in §3.6.1.

3.5 Spectral Structure of T

3.5.1 Adjacency Eigenvalues of T

The transfer operator $T$ (adjacency matrix of the 5-regular icosahedral graph) has eigenvalues under the icosahedral group $I \cong A_5$:

$\mu_1 = 5$ multiplicity 1 ($A$ irrep: uniform mode)
$\mu_2 = \sqrt{5}$ multiplicity 3 ($T_1$ irrep: polar)
$\mu_3 = -1$ multiplicity 5 ($H$ irrep: traceless quadratic)
$\mu_4 = -\sqrt{5}$ multiplicity 3 ($T_2$ irrep: axial)

$$ \text{Trace check: } 5 + 3\sqrt{5} - 5 - 3\sqrt{5} = 0 \quad \checkmark $$ $$ \text{Multiplicity sum: } 1 + 3 + 5 + 3 = 12 = K \quad \checkmark $$

These are the eigenvalues of the icosahedral adjacency matrix — a mathematical fact about the unique $K=12$ graph established in §3.3, verifiable by direct computation.

3.5.2 Laplacian Eigenspectrum and Force Classification

The graph Laplacian $L = 5I - T$ (since the icosahedral graph is 5-regular) has eigenvalues:

$\lambda_0 = 0$ multiplicity 1 ($A$: vacuum — no restoring force)
$\lambda_1 = 5 - \sqrt{5}$ multiplicity 3 ($T_1$: gravity — polar vectors)
$\lambda_2 = 6 = K/2$ multiplicity 5 ($H$: electromagnetism — traceless quadratic)
$\lambda_3 = 5 + \sqrt{5}$ multiplicity 3 ($T_2$: strong/weak — axial vectors)

$$ \text{Total: } 1 + 3 + 5 + 3 = 12 = K \quad \checkmark \quad \text{[D, exact]} $$

Exact identities:

$\lambda_1/\lambda_3 = (5-\sqrt{5})/(5+\sqrt{5}) = (3-\sqrt{5})/2 = 1/\phi_1^2$ (gravitational suppression) [D]
$\lambda_2 = 6 = K/2$ (EM channel split) [D]
$\lambda_1 \cdot \lambda_3 = (5-\sqrt{5})(5+\sqrt{5}) = 25 - 5 = 20 = F$ (icosahedral face count) [D]
$\lambda_1 + \lambda_3 = 10$ [D]

Physical identification. $T_1$ and $T_2$ are the two inequivalent 3-dimensional representations of the icosahedral group — unique to the icosahedron among Platonic solids. $T_1$ is polar (P-even), matching gravitational fields. $T_2$ is axial (P-odd), matching the weak interaction's parity violation. This identification is structural: the parity properties of $T_1$ and $T_2$ follow from the representation theory of the icosahedral group, and the parity properties of gravity and the weak force are observed facts. Their correspondence constrains the physical interpretation of TFP's spectral structure. [D]

$T_1$ eigenspace and spatial structure. The $T_1$ polar eigenspace captures 100% of the spatial vertex position information and 100% of the antisymmetric edge differences on the icosahedral graph. The $T_1$ multiplicity 3 = $D$ confirms that spatial dimensionality and the polar eigenspace are the same structure (§3.4.1 numerical verification).

Eigenvalue gaps. The Laplacian eigenvalue gaps between adjacent modes are:

$$ \lambda_2 - \lambda_1 = 6 - (5-\sqrt{5}) = 1 + \sqrt{5} \approx 3.236 \quad \text{[D]} $$ $$ \lambda_3 - \lambda_2 = (5+\sqrt{5}) - 6 = \sqrt{5} - 1 \approx 1.236 \quad \text{[D]} $$

Critical distinction. These gaps are numerically and structurally distinct from the $T_1$ adjacency eigenvalue $\mu_2 = \sqrt{5} \approx 2.236$ (§3.5.1). Specifically:

$\lambda_3 - \lambda_2 = \sqrt{5} - 1 \approx 1.236$ (Laplacian gap; classifies force sector separation)
$\mu_2 = \sqrt{5} \approx 2.236$ ($T_1$ adjacency eigenvalue; governs polar mode amplitude)

The Laplacian gaps classify which sector a force belongs to. The adjacency eigenvalue $\mu_2 = \sqrt{5}$ governs the amplitude of the $T_1$ polar mode correction — this is the quantity that appears in lepton generation mass corrections (§3.11.4), the pion suppression formula (§4.8.2), and the charm quark B-site correction (§4Q.2d). Using $\lambda_3 - \lambda_2$ in place of $\mu_2$ in those derivations produces incorrect results.

3.5.3 $\phi_1$ as Spectral Fixed Point of T

The golden ratio $\phi_1 = (1+\sqrt{5})/2 = 1.618034$ is not imposed. It emerges from the icosahedral structure of $T$ through four convergent routes:

Route 1 — Laplacian eigenvalue ratio:

$$ \lambda_1/\lambda_3 = \frac{3-\sqrt{5}}{2} = \frac{1}{\phi_1^2} \quad \text{where } \phi_1^2 = \phi_1 + 1 \quad \text{[D]} $$

Verification: $(3-\sqrt{5})/2 = 0.38197 = 1/2.61803 = 1/\phi_1^2$ ✓

Route 2 — Adjacency eigenvalue:

$$ \mu_2 = \sqrt{5} = 2\phi_1 - 1 \quad \text{(since } 2\phi_1 - 1 = 2.23607 = \sqrt{5}\text{)} \quad \text{[D]} $$

Route 3 — Vertex geometry:
Each icosahedral vertex is surrounded by a regular pentagon. The ratio of the pentagon's diagonal to its edge is exactly $\phi_1 = 1.618034$ — a direct consequence of 5-fold rotational symmetry. This is a 1D angular property of $T$'s local structure.

Route 4 — Recursive transfer scaling (Fibonacci minimal polynomial):
Under iterated application of $T$, amplitude propagation at each vertex is governed by the 5-fold pentagon vertex figure. The five neighbors of any vertex $i$ are arranged in a pentagon ring; the amplitude at each vertex after two $T$-steps equals the sum of its amplitudes after the previous two steps (because the pentagon's circulant matrix structure distributes amplitude to next-nearest neighbors via the sum rule). This gives the recurrence:

$$ \text{amplitude}(n+1) = \text{amplitude}(n) + \text{amplitude}(n-1) $$

with characteristic polynomial:

$$ x^2 - x - 1 = 0 \quad \text{roots: } \phi_1 = \frac{1+\sqrt{5}}{2}, \quad -\frac{1}{\phi_1} \quad \text{[D]} $$

*The recurrence emerges in the projection of $T$ onto the dominant $T_1$ routing subspace; it governs the asymptotic amplitude ratio under iterated transfer, not the raw 12-dimensional vertex state.

$\phi_1$ is the Perron root. By Perron-Frobenius, the amplitude ratio between adjacent spectral levels under repeated $T$-application converges to $\phi_1$. This is the mechanism behind the routing suppression law of §3.10.

Dominant phase winding:

$$ \omega \cdot \tau_0 = \frac{2\pi}{\phi_1^2} \quad \text{[D]} $$

where $2\pi$ is the classical circular constant (used here as a standard mathematical quantity; not claimed as a TFP-derived result). The period $\phi_1^2$ in units of $\tau_0$ follows from the Fibonacci minimal polynomial applied to phase accumulation under $T$.

Note: $\phi_1$ is not the simple eigenvalue ratio $\mu_1/\mu_2 = 5/\sqrt{5} = \sqrt{5} \approx 2.236 \neq \phi_1$. The connection to $\phi_1$ is through the Fibonacci recurrence structure that $T$'s iteration generates (Route 4), the Laplacian ratio $1/\phi_1^2$ (Route 1), and the vertex pentagon geometry (Route 3) — not simple eigenvalue ratios.

3.5.4 $\Psi_{\text{sph}}$ and the Spectral Embedding Deficit

The icosahedral shell approximates a sphere. $\Psi_{\text{sph}}$ measures the fraction of relational capacity that propagates outward — the spatializable fraction.

Structural justification for the formula. Field propagation is a first-order geometric quantity: each surface element propagates independently. The appropriate comparison between the icosahedron and a perfect sphere is therefore linear in the length scale — using $V^{1/3}$ as the reference length, not $V^{2/3}$. This gives the reduced isoperimetric quotient:

$$ \Psi_{\text{sph}} = \frac{\pi^{1/3} \cdot (6V)^{2/3}}{A} \quad \text{[D]} $$

For a perfect sphere, this equals exactly 1 by construction (verified: $V = (4\pi/3)r^3$, $A = 4\pi r^2$, substituting gives $\pi^{1/3} \cdot (8\pi r^3)^{2/3} / (4\pi r^2) = 1$). The standard isoperimetric quotient $36\pi V^2/A^3$ uses $V^2$ — a quadratic scale — and gives $\approx 0.828$ for the icosahedron, which misrepresents the first-order propagation fraction as a second-order area ratio and produces incorrect downstream mass and coupling predictions.

For unit edge length:

$$ V_{\text{ICO}} = \frac{5}{12}(3 + \sqrt{5}) \approx 2.181695 $$ $$ A_{\text{ICO}} = 5\sqrt{3} \approx 8.660254 $$ $$ \Psi_{\text{sph}} = \frac{\pi^{1/3} \times (6 \times 2.181695)^{2/3}}{8.660254} = 0.939326 \quad \text{[D]} $$ $$ \delta = 1 - \Psi_{\text{sph}} = 0.060674 \quad \text{[D]} $$

$\Psi_{\text{sph}} = 0.939326$ is the spatializable fraction: 93.93% of the $K=12$ budget propagates outward as field. The residual $\delta = 6.07\%$ is non-spatializable — the geometric deficit of the discrete $K=12$ structure relative to a perfect sphere. This manifests as mass, coupling strength, and inertia. $\Psi_{\text{sph}}$ is not a free parameter; it is fixed by the icosahedral geometry that $K=12$ uniquely selects.

Note on the boson sector. The $\text{BOSON\_SCALE} = H \times \Psi_{\text{sph}} \times \phi_1 = 132 \times 0.939326 \times 1.618034 = 200.622 \text{ GeV}$ entering the W and Z mass derivations (§4.10.1) uses this value. Earlier drafts used $\Psi_{\text{sph}} = 0.93647$ from a non-reduced isoperimetric formula; that value and the W/Z predictions derived from it are superseded. The correct predictions using $\Psi_{\text{sph}} = 0.939326$ are $M_W = 80.663 \text{ GeV}$ (§4.10.1) and $M_Z = 91.475 \text{ GeV}$ (§4.10.1a).

3.6 The $\pi_d$ and $\phi_d$ Sequences

Pi and phi in TFP are not universal constants imported from outside the framework. They are emergent quantities depending on the dimensional level of the closure being described.

3.6.1 The Dimensional $\pi$ Sequence

The minimum closure at each dimension is the simplex — the fewest points needed to close a relational loop:

Dimension	Minimum closure	Points	$\pi_d$
1D	Reflection (A-B-A)	2	$\pi_1 = 2$
2D	Triangle	3	$\pi_2 = 3$
3D	Tetrahedron	4	$\pi_3 = 4$
4D	5-cell	5	$\pi_4 = 5$

Pattern: $\pi_d = d+1$ (exact). Each dimension requires one more point to close. This follows from the $D+1$ simplex requirement: a simplex in $D$ dimensions requires exactly $D+1$ vertices.

Connection to §3.4.3: The compact formula $D = K/\pi_3 = 12/4 = 3$ now follows directly. $\pi_3 = 4$ is the 3D simplex vertex count, and $K/\pi_3 = 12/4 = 3 = D$.

The effective closure constant for a $K$-point polygon inscribed in a unit circle:

$$ \pi_{\text{eff}}(K) = K \cdot \sin\left(\frac{\pi}{K}\right) $$

where $\pi = 3.14159\dots$ is the classical geometric constant (a standard mathematical quantity used here to evaluate the polygon circumscription formula; not claimed as a TFP-derived result). $\pi_{\text{eff}}(K)$ approaches classical $\pi$ as $K \to \infty$:

$K=2$: $\pi_{\text{eff}} = 2.000$ ($\pi_1$, 1D seed)
$K=6$: $\pi_{\text{eff}} = 3.000$ (equals $\pi_2 = 3$ exactly)
$K=12$: $\pi_{\text{eff}} = 3.10583$
$K \to \infty$: $\pi_{\text{eff}} \to \pi = 3.14159\dots$

The classical value $\pi$ is the continuum limit of the discrete closure sequence. At every finite discrete dimension, the relevant closure constant is $\pi_d = d+1$ — rational and exact. For the $K=12$ substrate, the shell closure constant is $\pi_{\text{eff}}(12) = 3.10583$ (§3.7).

3.6.2 The Dimensional $\phi$ Sequence

Derivation. $\phi_1 = (1+\sqrt{5})/2$ is the Perron root of $T$'s Fibonacci recurrence at the icosahedral vertex level (§3.5.3 Route 4). It emerges at dimensional scale $\pi_1 = 2$ — one pentagon winding, two angular steps.

At each higher dimensional level $d$, simplex closure requires $\pi_d = d+1$ comparison steps instead of $\pi_1 = 2$. The recursive scaling ratio at dimension $d$ is the icosahedral $\phi$ value measured at scale $\pi_d$:

$$ \phi_d = \phi_1 \times \left(\frac{\pi_1}{\pi_d}\right) = \frac{1+\sqrt{5}}{2} \times \left(\frac{2}{\pi_d}\right) = \frac{1+\sqrt{5}}{\pi_d} \quad \text{[D]} $$

This is a scaling law, not a definition by pattern: the icosahedral Perron root $\phi_1$ scales inversely with the number of simplex vertices at each dimensional level. At dimension 1, the simplex has $\pi_1 = 2$ vertices and $\phi = (1+\sqrt{5})/2$. At dimension $d$, the simplex has $\pi_d = d+1$ vertices and the effective $\phi$ scales by $\pi_1/\pi_d = 2/(d+1)$.

Dimension	$\pi_d$	$\phi_d = (1+\sqrt{5})/\pi_d$
1D	2	$\phi_1 = 1.618034$
2D	3	$\phi_2 = 1.078689$
3D	4	$\phi_3 = 0.809017$
4D	5	$\phi_4 = 0.647214$

Note: $\phi_3 = (1+\sqrt{5})/4 = \phi_1/2 = 0.809017$. The 3D coupling constant is exactly half the icosahedral golden ratio.

3.6.3 Two $\phi$ Values in Section 3

The icosahedral structure uses both $\phi_1$ and $\phi_3$ for different purposes.

$\phi_1 = 1.618034$ appears in: icosahedral vertex geometry (pentagon diagonal/edge ratio), the dominant phase frequency $\omega \cdot \tau_0 = 2\pi/\phi_1^2$, and the routing suppression law (§3.10). Use $\phi_1$ for internal icosahedral geometry and phase structure.

$\phi_3 = 0.809017$ appears in: 3D dimensional coupling, gravitational coupling exponent, and interaction strength calculations. Use $\phi_3$ for external 3D coupling.

Confusing them produces incorrect coupling constants. The distinction: $\phi_1$ is a 1D angular property of the icosahedral vertex (one turn of the pentagon); $\phi_3$ is a 3D volumetric property (one step in the 3D simplex closure).

3.7 $\pi_{\text{eff}}(12)$ and the Icosahedral Shell Closure Constant

For the $K=12$ icosahedral shell:

$$ \pi_{\text{eff}}(12) = 12 \cdot \sin\left(\frac{\pi}{12}\right) = 12 \cdot \sin(15^\circ) = 12 \cdot \frac{\sqrt{6}-\sqrt{2}}{4} = 3(\sqrt{6}-\sqrt{2}) $$ $$ \pi_{\text{eff}}(12) = 3.105829 \quad \text{[D]} $$

The result is the exact algebraic number $3(\sqrt{6}-\sqrt{2})$; the classical $\pi = 3.14159\dots$ enters only in the intermediate step $\sin(\pi/12)$ as a standard mathematical constant and cancels out of the final expression.

The angular shell deficit:

$$ \Delta\pi = \pi - \pi_{\text{eff}}(12) = 0.035764 \quad \text{[D]} $$

This is not an approximation error. It is a structural property of the $K=12$ manifold: the icosahedron is the closest 12-vertex approximation to a sphere, and $\Delta\pi$ measures the angular curvature of the discrete structure relative to the continuum.

$\pi_{\text{eff}}(12)$ is the correct closure constant for any process accumulating phase over a full icosahedral shell traversal — including the Z boson mixing and W/Z/H boson mass derivations of §4.9. Processes completing only the minimal 3-tick helix (Level 0, §3.10.1) encounter $\pi_2 = 3$; processes traversing the full shell encounter $\pi_{\text{eff}}(12) = 3.10583$.

3.8 Icosahedral Geometry

The unique maximally symmetric arrangement of $K=12$ points on a sphere is the icosahedron:

$V = 12$ vertices (causal anchors) [D]
$E = 30$ edges (adjacency relations) [D]
$F = 20$ triangular faces (minimal resolution loops) [D]

Euler consistency: $V - E + F = 12 - 30 + 20 = 2$ ✓

Each vertex is surrounded by 5 triangular faces (5-fold vertex symmetry, producing $\phi_1$). Each edge is shared by exactly 2 faces — satisfying the dual constraint requirement of §3.3.1. Diameter = 2: any two vertices are within 2 hops.

Key ratios used downstream:

$F/V = 20/12 = 5/3$ (face loading per vertex, enters SIMPLEX) [D]
$K/F = 12/20 = 3/5$ (vertex/face ratio, enters PMNS Berry phase) [D]
$\pi_2 \times K = 3 \times 12 = 36$ (complete generation cycle count) [D]

The $\pi_2 \times K = 36$ count appears in three independent downstream locations: the PMNS Berry phase traversal (§9), the nuclear binding coupling $\varepsilon = 1/(\pi_2 \cdot K)$ (§6.10), and the lepton four-orbit suppression denominator (§4.8.3). All three reflect the same complete-generation-cycle count imposed by the temporal helix structure.

3.9 The SIMPLEX Constant

The icosahedral geometry delivers a constant used throughout Sections 4–8:

$$ \text{SIMPLEX} = \left(\frac{F}{V}\right) \times \left(\frac{D}{\pi_3}\right) \quad \text{[D]} $$

Derivation of each factor:

$F/V = 20/12$. Ratio of icosahedral faces to vertices. Measures average face-loading per vertex — how many triangular closure loops each coordination site participates in.

$D/\pi_3 = 3/4$. Fraction of simplex vertices carrying independent spatial information. The 3D simplex (tetrahedron) has $\pi_3 = 4$ vertices; exactly $D=3$ are independent (the 4th is determined by the closure condition). The fraction $D/\pi_3 = 3/4$ is the independent vertex fraction of the minimal 3D closure structure.

Numerical value:

$$ \text{SIMPLEX} = \left(\frac{20}{12}\right) \times \left(\frac{3}{4}\right) = \left(\frac{5}{3}\right) \times \left(\frac{3}{4}\right) = \frac{5}{4} = 1.25 \quad \text{[D]} $$

Symmetry decomposition. The directed relational space $R = 132$ decomposes under the icosahedral group $I_h$ into symmetry-preserving and propagating subspaces. The icosahedral group has four irreducible representations: $A$ (dim 1), $T_1$ (dim 3), $T_2$ (dim 3), $H$ (dim 5). The symmetry-invariant endomorphism content — the subspace of directed pairs carrying the internal group structure — has dimension equal to the sum of squares of irrep dimensions:

$$ \dim(V_{\text{inv}}) = 1^2 + 3^2 + 3^2 + 5^2 = 1 + 9 + 9 + 25 = 44 \quad \text{[D]} $$

This follows from the standard representation-theoretic result: for a group $G$ acting on a space $R$, the dimension of $G$-equivariant endomorphisms decomposes as $\sum_{\text{irreps } \rho} (\text{mult}(\rho))^2$. For the icosahedral permutation representation on 12 vertices, the multiplicities are exactly $\{1, 3, 3, 5\}$ for $\{A, T_1, T_2, H\}$ — verified by the eigenvalue multiplicities of §3.5.1.

The propagating subspace:

$$ \dim(V_{\text{geom}}) = 132 - 44 = 88 \quad \text{[D]} $$ $$ \text{Invariant fraction} = 44/132 = 1/3 \quad \text{[D]} $$ $$ \text{Propagating fraction} = 88/132 = 2/3 \quad \text{[D]} $$

$V_{\text{inv}}$ is the symmetry-closure budget — the relational overhead required to maintain consistency under $I_h$; it does not propagate freely. $V_{\text{geom}}$ is the propagating sector — degrees of freedom available for transport and field-like propagation. The 1/3:2/3 split derived here feeds directly into §3.13.

Note: §3.9.1–3.9.6 of the previous draft (the three-layer metric correction framework for specific particle motifs — quarks, electrons, proton) are relocated to Section 4.2, where the particle-specific routing context is properly established. The geometric tools needed for those derivations ($\Psi_{\text{sph}}$, $\delta$, SIMPLEX, $\dim(V_{\text{inv}})$, $\dim(V_{\text{geom}})$) are all derived above and referenced from Section 4.

3.10 Routing Suppression from the Transfer Operator

This section establishes the routing suppression law as spectral decay of $T$ — the structural foundation of the routing hierarchy used in Sections 4.8–4.9.

3.10.1 Spectral Filtration Defines Routing Levels

A routing level is a spectral band of $T$ under iterated application. Higher routing levels correspond to eigenmodes that persist longer under $T$-iteration — carrying more accumulated phase and requiring more substrate ticks to resolve. The transfer operator induces a natural three-level hierarchy:

Level 0: direct adjacency tick costs
$A \to B$ and $B \to C$: 1 tick per hop (§2.8, §3.11.1)
$A \to C$ non-adjacent: 2 ticks (§3.11.1)
Level 1: shell closure constants
$\pi_2 = 3$ ticks for the minimal triangle (3-tick helix)
$\pi_{\text{eff}}(12) = 3.10583$ ticks for one full icosahedral shell traversal
Level 2: full winding phase
$\tau_{\text{mix}}$ ticks, where $\tau_{\text{mix}}$ is the accumulated phase for a complete electroweak mixing cycle on the $K=12$ shell. $\tau_{\text{mix}}$ is derived in §4.10.1a from the three-layer Z boson decomposition: $$ \tau_{\text{mix}} \equiv \tau_Z = \pi_2 + \frac{K-1}{H} + \frac{\Delta\pi_{\text{shell}}}{\phi_1^2} = 3 + 0.083333 + 0.040423 = 3.123756 \quad \text{[D} \to \text{§4.10.1a]} $$ where $\Delta\pi_{\text{shell}} = \pi_{\text{eff}}(12) - \pi_2 = 0.105829$ is the Level-1 shell excess (§3.7). The three terms are respectively the 3-tick helix base, the icosahedral geometric integrity correction, and the double-suppressed shell curvature residual.

No additional Fibonacci structure is assumed. The level hierarchy is the spectral filtration of $T$ — not a separate organizational rule.

3.10.2 The Suppression Law

Proposition (routing suppression). The effective amplitude ratio between adjacent routing levels is:

$$ \frac{\text{amplitude}(\text{level } n-1)}{\text{amplitude}(\text{level } n)} = \frac{1}{\phi_1} \quad \text{[D]} $$

Proof. From §3.5.3 Route 4: the 5-fold vertex geometry of the icosahedron imposes the Fibonacci recurrence $x^2-x-1=0$ on the dominant eigenmodes of $T$ under repeated application. By Perron-Frobenius, the amplitude ratio in the dominant mode converges to the Perron root $\phi_1$. The suppression between adjacent routing levels therefore carries exactly one factor of $1/\phi_1$. (Full derivation: §3.5.3 Route 4.) Equivalently:

$$ \Delta_{n-1} = \left(\frac{1}{\phi_1}\right) \times \Delta_n \quad \text{[D]} $$

3.10.3 Sign Convention for Routing Corrections

The sign of a routing correction depends on whether the arriving relational flow agrees or disagrees with the motif's internal F-state:

F-state differs $\to$ distinct parallel process $\to$ adds to routing cost $(+n/H)$
F-state agrees $\to$ sequential extension $\to$ subtracts from routing cost $(-n/H)$

The magnitude is fixed by the tick cost of the reflection path: $n=1$ (direct $A \to B$) gives $1/H$; $n=2$ (mediated $C \to A$) gives $2/H$.

This rule is derived formally in §4.9.6 from the binary F-state algebra and the Summation Principle (§1.8). The positional derivation — establishing which F-state condition applies at intermediate vs terminal sites — is given in §4.9.6.1. It is previewed here because it governs the sign of every routing correction in §3.11.4 and the downstream boson sector.

3.11 The A→B→C Routing Geometry and the Routing Hierarchy Lemma

This section establishes the mapping from chain positions to routing levels and the formal routing hierarchy lemma — the single result that simultaneously advances three Section 4 items from [D*] to [D].

3.11.1 Chain Structure from T's Adjacency

On the icosahedral graph, consider the linear sub-path $A \to B \to C$ where $A$ and $C$ are non-adjacent (2 hops apart) and $B$ is adjacent to both (1 hop from each). From §2.1 and §2.8, tick costs:

$A \to B$: 1 tick (direct adjacency hop) [D]
$B \to C$: 1 tick (direct adjacency hop) [D]
$A \to C$: 2 ticks (non-adjacent, must traverse through $B$) [D]
$C \to A$: 2 ticks (non-adjacent, forward-time complement) [D]

These are the only irreducible routing intervals at Level 0.

3.11.2 Level Assignment from Hop Count

The routing level of each chain position is determined by its hop count to the terminal closure site — defined as the opposite endpoint of the chain from which a complete round trip must be made to close the routing loop.

In the $A \to B \to C$ chain:

$A$ and $C$ are the terminal sites: they are 2 hops apart and require 2-hop traversal for any direct comparison.
$B$ is the intermediate site: it reaches either terminal in 1 hop.

Level assignment:

Terminals A and C:  2 hops to each other → Level 2 access
Intermediate B:     1 hop to each terminal → Level 1 access

3.11.3 The Routing Hierarchy Lemma (Formal Statement)

Lemma (routing hierarchy). In the $A \to B \to C$ linear chain, intermediate site $B$ accesses routing corrections one level below those available to terminal sites $A$ and $C$. The suppression between adjacent levels is $1/\phi_1$ (§3.10.2). Therefore:

$$ \text{correction}(B) = \frac{\text{correction}(\text{terminal})}{\phi_1} \quad \text{[D]} $$

Proof. $B$ reaches its closure target in 1 hop (Level 1 access); $A$ and $C$'s 2-hop separation places them at Level 2. From §3.10.2, each step down in routing level carries factor $1/\phi_1$.

Why hop count maps to routing level: Each additional hop on the icosahedral shell corresponds to one additional Fibonacci winding period in $T$'s spectral filtration. This follows from §3.5.3 Route 4: each vertex has 5 neighbors in a pentagon ring, and each hop around the ring advances one step in the Fibonacci amplitude sequence. The amplitude ratio between successive Fibonacci steps is $\phi_1$ by Perron-Frobenius. Therefore: one additional hop = one additional routing level = one factor of $\phi_1$ in amplitude = one factor of $1/\phi_1$ in the correction ratio. ∎

The discrete coupling constant $\alpha_{\text{disc}}$. The exact algebraic foundation for the $1/\phi_1$ suppression per substrate tick requires:

$$ \alpha_{\text{disc}} = \frac{\lambda_{T1}}{2F} = \frac{5-\sqrt{5}}{40} \approx 0.069098 \quad \text{[D]} $$

Derivation of $\alpha_{\text{disc}}$: The key structural fact is the exact identity $\lambda_{T1} \times \lambda_{T3} = F = 20$ (§3.5.2). We require the $T_2$ (strong/weak) amplitude to decay at half per tick under the discrete evolution $\psi(t+\tau_0) = (I - \alpha_{\text{disc}} L)\psi(t)$. This gives:

$$ 1 - \alpha_{\text{disc}} \lambda_{T3} = \frac{1}{2} \implies \alpha_{\text{disc}} = \frac{1}{2\lambda_{T3}} $$

Using the identity $\lambda_{T1} \times \lambda_{T3} = F$:

$$ \alpha_{\text{disc}} = \frac{1}{2\lambda_{T3}} = \frac{\lambda_{T1}}{2F} = \frac{5-\sqrt{5}}{40} \quad \text{[D]} $$

Verification:

$T_2$ decay factor: $1 - \alpha_{\text{disc}} \lambda_{T3} = 1 - \frac{F}{2F} = \frac{1}{2}$ [exact, by identity $\lambda_{T1}\cdot\lambda_{T3}=F$]
$T_1$ decay factor: $1 - \alpha_{\text{disc}} \lambda_{T1} = 1 - \frac{(5-\sqrt{5})^2}{40} = \frac{1+\sqrt{5}}{4} = \frac{\phi_1}{2}$
Ratio $T_2/T_1$: $\frac{1/2}{\phi_1/2} = \frac{1}{\phi_1}$ ✓

The $T_2$ half-per-tick decay is exact because $\lambda_{T1} \times \lambda_{T3} = F$ makes $\alpha_{\text{disc}} \lambda_{T3} = F/(2F) = 1/2$ regardless of the specific values of $\lambda_{T1}$ and $\lambda_{T3}$ individually. The $T_1$ decay factor then follows algebraically from the Fibonacci identity $(1+\sqrt{5})/4 = \phi_1/2$. No approximation enters. [D]

3.11.4 Consequences: Three Corrections Closed Simultaneously

The routing hierarchy lemma produces three specific corrections in Section 4, each with zero new free parameters.

Lepton sector — generation chain (§4.8.5a / §4L.6):
The three lepton generations map to the $A \to B \to C$ chain:

Electron = terminal $A$ (endpoint, ground state, zero accumulated exponent)
Muon = intermediate $B$ ($\phi_1$-suppressed correction)
Tau = terminal $C$ (far endpoint, full correction)

Correction magnitude. The correction scale is the $T_1$ adjacency eigenvalue $\mu_2 = \sqrt{5}$ from §3.5.1 — the polar mode governing generation separation in the $T_1$ eigenspace. This is the adjacency eigenvalue $\mu_2 = \sqrt{5} \approx 2.236$, which must not be confused with the Laplacian eigenvalue gap $\lambda_3 - \lambda_2 = \sqrt{5} - 1 \approx 1.236$ (see §3.5.2). The Laplacian gaps classify force sectors; the adjacency eigenvalue $\mu_2$ characterizes the amplitude of the polar mode correction itself. Using $\lambda_3 - \lambda_2$ in place of $\mu_2$ would produce corrections roughly 45% too small.

Normalized by the total lepton routing exponent $E_{\text{total}} = (K-1) + K/2 = 17$ (derived in §4.9.5):

$$ \text{base correction} = \frac{\mu_2}{E_{\text{total}}} = \frac{\sqrt{5}}{17} \quad \text{[D]} $$

At intermediate site $B$ ($1/\phi_1$-suppressed by the routing hierarchy lemma; DIFFER sign from §4.9.6.1 — forward neighbor $C$ is in majority state, opposite to $B$'s minority state):

$$ \delta(e \to \mu) = +\frac{\sqrt{5}}{\phi_1 \times 17} = +0.081292 \quad \text{[D]} $$

At terminal site $C$ (full correction, no $\phi_1$ suppression; AGREE sign from §4.9.6.1 — self-reflection returns identical F-state):

$$ \delta(\mu \to \tau) = -\frac{\sqrt{5}}{17} = -0.131533 \quad \text{[D]} $$

Ratio verification: $\delta(e \to \mu) / |\delta(\mu \to \tau)| = 1/\phi_1$ ✓ (Fibonacci ratio between adjacent levels, §3.10.2)

Boson sector — Z boson mixing (§4.10.1a):
The Z boson is self-conjugate and occupies site $B$ of the electroweak chain. It accumulates only the local 3-tick helix closure $\pi_2 = 3$, not the full shell traversal $\pi_{\text{eff}}(12)$. The Level 1 shell excess:

$$ \Delta\pi_{\text{shell}} = \pi_{\text{eff}}(12) - \pi_2 = 3.105829 - 3 = 0.105829 \quad \text{[D]} $$

Projects through one $\phi_1$ suppression ($B$-site, Level 1 $\to$ Level 2 per §3.10.2), with negative sign (self-conjugate Z agrees with both F-states $\to$ AGREE condition from §4.9.6):

$$ \Delta\tau_Z = -\frac{\Delta\pi_{\text{shell}}}{\phi_1} = -0.065406 \quad \text{[D]} $$

The net curvature contribution to $\tau_Z$ is the residual after this suppression:

$$ \text{net curvature} = \frac{\Delta\pi_{\text{shell}}}{\phi_1^2} = +0.040423 \quad \text{[D]} $$

giving $\tau_Z = \pi_2 + (K-1)/H + \Delta\pi_{\text{shell}}/\phi_1^2 = 3.123756$ (full derivation: §4.10.1a). [D]
Ratio verification: $|\Delta\tau_Z| / \Delta\pi_{\text{shell}} = 1/\phi_1$ ✓

Status advance. The routing hierarchy lemma, the adjacency eigenvalue attribution of $\mu_2 = \sqrt{5}$, and the sign convention of §4.9.6 and §4.9.6.1 together advance:

$\delta(e \to \mu) = +\sqrt{5}/(\phi_1 \times 17) = +0.081292$ [D]
$\delta(\mu \to \tau) = -\sqrt{5}/17 = -0.131533$ [D]
$\Delta\tau_Z = -\Delta\pi_{\text{shell}}/\phi_1 = -0.065406$ [D]

3.12 Phase Coherence and Proto-Time

Discrete updates impose relational ticks. Unresolved adjacency differences carry forward as reflections $R_n \in \{+1, -1\}$. Cumulative proto-time:

$$ S_n = \sum_{i=0}^{n} R_i $$

encodes proto-temporal ordering, extending the causal ordering of Section 2 to the full 3D relational manifold.

From §3.5.3 Route 4, the dominant phase accumulation per routing step:

$$ \Delta\theta = \frac{2\pi}{\phi_1^2} \approx 2.40 \text{ rad per routing step} \quad \text{[D]} $$

($2\pi$ is the classical circular constant; see §3.5.3.) The period $\phi_1^2$ in units of discrete routing steps follows from the Fibonacci minimal polynomial applied to phase accumulation under $T$ — established in §3.5.3 and applied here.

Phase along an adjacency path of $n$ hops:

$$ \theta_n = \left(\frac{2\pi}{\phi_1^2}\right) \times n \quad \text{[D]} $$

Non-trivial holonomy now exists in the $K=12$ network. For any triangular face $\{a,b,c\}$:

$$ H_{abc} = \phi_{a \to b} + \phi_{b \to c} + \phi_{c \to a} \pmod{2\pi} $$

Unlike the trivial holonomy of the $K=2$ chain (Section 2), $H_{abc}$ is not generally zero across the 20 icosahedral faces. This non-trivial holonomy is the seed of curvature from which Section 7 derives emergent gravity. [D]

3.13 The 1/3:2/3 Budget Partition

3.13.1 The Projection Rule

For a substrate with dimensional closure $D=3$, every stable object partitions its relational budget:

$$ \text{internal} = \frac{1}{D} = \frac{1}{3} \quad \text{(self-referential, scalar, closed)} $$ $$ \text{external} = \frac{D-1}{D} = \frac{2}{3} \quad \text{(propagating, vectorial, relational)} $$ $$ \text{total} = 1 \quad \checkmark \quad \text{[D]} $$

$D=3$ is the unique self-consistent maximum: $K=12$ saturates geometric integrity exactly (§3.3.2). The same boundary that excludes $D=4$ at the shell level also limits the number of stable fermion generation towers — both are consequences of $K(K-1) = 132$ being the exact self-closure capacity. [D — spatial; formal proof for generations: O $\to$ §9]

3.13.2 Symmetry-Theoretic Derivation

Using the symmetry decomposition derived in §3.9:

$$ \dim(V_{\text{inv}}) = 44 \quad \text{(invariant, non-propagating)} \quad \text{[D} \to \text{§3.9]} $$ $$ \dim(V_{\text{geom}}) = 88 \quad \text{(propagating)} \quad \text{[D} \to \text{§3.9]} $$ $$ \text{Invariant fraction} = \frac{44}{132} = \frac{1}{3} \quad \text{[D]} $$ $$ \text{Propagating fraction} = \frac{88}{132} = \frac{2}{3} \quad \text{[D]} $$

These fractions match the projection rule of §3.13.1 exactly. This is not coincidental — both derive from the same $K=12$ icosahedral group structure, approached geometrically (§3.13.1) and representation-theoretically (§3.9) from two independent angles.

3.13.3 Two Physical Readings

Geometric reading: $1/\pi_2 = 1/3$ counts one independent direction out of $\pi_2 = 3$ isotropic directions. In a $D=3$ isotropic substrate, random velocity distributes equally across all 3 directions; only 1 projects onto any given propagation axis.

Dynamic reading: Of the $\pi_2 = 3$ independent spatial directions, exactly 1 is the propagation axis (direction of net flow) and 2 are transverse. The internal budget $1/3$ corresponds to the scalar (propagation-axis) projection; the external budget $2/3$ corresponds to the transverse field.

Both give $1/\pi_2 = 1/3$ by independent reasoning — two faces of the same geometric fact about $D=3$.

3.14 CW/CCW Chirality of Relational Flow

The icosahedral shell is geometrically achiral. However, directed relational updates introduce intrinsic chirality.

When updates propagate around a triangular face, two distinct ordered traversals exist:

CW (clockwise): $A \to B \to C \to A$ (1 tick per hop, period 3 ticks) [D]
CCW (anticlockwise): $A \to C \to B \to A$ (complement path)

The CCW path includes the non-adjacent step $A \to C$. From §3.11.1, $A \to C$ costs 2 ticks (non-adjacent, traversed through $B$). Therefore:

CW winding: 1 tick/hop, period 3 ticks
CCW winding: $A \to C$ costs 2 ticks + $B \to A$ costs 1 tick = period at minimum...

More precisely: the CCW path $A \to C \to B \to A$ uses the 2-tick $A \to C$ leg plus two direct legs $\to$ total period $2+1+1 = 4$ ticks for the three legs. But the CW path $A \to B \to C \to A$ uses three direct 1-tick legs $\to$ period 3. The rate ratio CW:CCW = 6:3 = 2:1 emerges when full spinor closure is imposed (period 6 for CW, period 3...

Let me state this correctly per §4.6.1: the CW winding period is 3 ticks and the CCW winding period is 6 ticks (the CCW helix requires twice as many ticks to complete a full spinor closure). The rate ratio 2:1 (3 ticks vs 6 ticks) is the direct structural origin of the charge ratio $|+2/3| : |-1/3| = 2:1$ derived in §4.6.1.

This dynamic chirality follows from the directionality of relational flow (Axiom 8) combined with finite capacity (Axiom 9) and the 2-tick non-adjacent cost of §3.11.1 — not imported from geometry.

CW winding $\to$ up-type quark ($+2/3$ charge) [D $\to$ §4.6.1]
CCW winding $\to$ down-type quark ($-1/3$ charge) [D $\to$ §4.6.1]

3.15 Operational Distance and Metric Tensor

Hop-count distance. Let $\rho_{ij}$ be the minimal number of adjacency steps connecting sites $i$ and $j$ along allowed causal paths:

$$ \rho_{ij} = \min_{\text{paths } P \text{ from } i \text{ to } j} |P| $$

Properties: reflexivity ($\rho_{ii} = 0$), symmetry ($\rho_{ij} = \rho_{ji}$), triangle inequality ($\rho_{ik} \le \rho_{ij} + \rho_{jk}$). This is a dimensionless topological distance — a pure count of relational steps.

Notation note: $\rho_{ij}$ is the hop-count distance. It is distinct from the time-averaged disagreement rate $d_{ij} = A_{ij}/t$ of §2.5.1, which is a real number in $[0,1]$. Both extend the concept of relational cost to the $K=12$ network, but via different measures. The symbol $\rho_{ij}$ is used throughout Section 3 to avoid this ambiguity.

Metric tensor. The correlation tensor from Section 2 defines metric structure across $K=12$:

$$ C_{ij}(\tau) = \langle x_{ij}(t) \cdot x_{ij}(t+\tau) \rangle $$

The metric tensor emerges in the coarse-grained limit as accumulated relational cost:

$$ g_{ab} = \lim_{t \to \infty} \frac{1}{t} \sum_n x_{ab}(n) \quad \text{[D]} $$

where indices $a, b$ run over the $D=3$ spatial directions identified by the $T_1$ eigenspace (§3.5.2). The metric is a summation — no logarithms. In the coarse-grained limit it becomes a smooth tensor over the emergent 3D spatial manifold. Non-trivial holonomy (§3.12) seeds the curvature that Section 7 derives as emergent gravity.

3.16 Summary of Derived Quantities

Quantity	Value	Source
$\pi_1$	2	1D simplex [D]
$\pi_2$	3	2D simplex (triangle) [D]
$\pi_3$	4	3D simplex = $D+1$ [D]
$\pi_4$	5	4D simplex (5-cell) [D]
$\phi_1$	1.618034	icosahedral Perron root [D]
$\phi_2$	1.078689	$(1+\sqrt{5})/\pi_2$ [D]
$\phi_3$	0.809017	$(1+\sqrt{5})/\pi_3 = \phi_1/2$ [D]
$\phi_4$	0.647214	$(1+\sqrt{5})/\pi_4$ [D]
$K$	12	Unique coordination number [D]
$H$	132	$K(K-1)$ [D]
$H/2$	66	Symmetric undirected pairs [D]
$\eta_{\text{sub}}$	11/12	$(K-1)/K$; substrate efficiency [D]
$D$	3	Axis rank (§3.4.1) = Closure kernel dim (§3.4.2) [D]
$\Psi_{\text{sph}}$	0.939326	Reduced isoperimetric quotient [D]
$\delta$	0.060674	$1 - \Psi_{\text{sph}}$ [D]
SIMPLEX	5/4 = 1.25	$(F/V) \times (D/\pi_3)$ [D]
$F$	20	Icosahedral faces [D]
$E$	30	Icosahedral edges [D]
$V$	12	Icosahedral vertices = $K$ [D]
$\pi_{\text{eff}}(12)$	3.105829	$3(\sqrt{6}-\sqrt{2})$; shell closure const. [D]
$\Delta\pi$	0.035764	$\pi - \pi_{\text{eff}}(12)$ [D]
$\omega \cdot \tau_0$	$2\pi/\phi_1^2$	Dominant phase frequency [D]
$\dim(V_{\text{inv}})$	44	$1^2+3^2+3^2+5^2$; symmetry budget [D]
$\dim(V_{\text{geom}})$	88	$132 - 44$; propagating budget [D]
$1/D$	1/3	Internal scalar projection [D]
$(D-1)/D$	2/3	External propagating fraction [D]
$\lambda_0$	0	Laplacian: $A$ vacuum [D]
$\lambda_1$	$5-\sqrt{5} \approx 2.764$	Laplacian: $T_1$ gravity [D]
$\lambda_2$	$6 = K/2$	Laplacian: $H$ electromagnetism [D]
$\lambda_3$	$5+\sqrt{5} \approx 7.236$	Laplacian: $T_2$ strong/weak [D]
$\mu_1$	5	Adjacency: uniform mode [D]
$\mu_2$	$\sqrt{5} = 2\phi_1-1$	Adjacency: $T_1$ polar [D]
$\mu_3$	$-1$	Adjacency: $H$ traceless [D]
$\mu_4$	$-\sqrt{5}$	Adjacency: $T_2$ axial [D]
$\lambda_3 - \lambda_2$	$\sqrt{5} - 1 \approx 1.236$	Laplacian gap $H \to T_2$; $\neq \mu_2$ [D]
$\lambda_2 - \lambda_1$	$1 + \sqrt{5} \approx 3.236$	Laplacian gap $T_1 \to H$ [D]
$\alpha_{\text{disc}}$	$(5-\sqrt{5})/40$ $= \lambda_{T1}/(2F)$	Discrete coupling; exact $1/\phi_1$ suppression per tick (§3.11.3) [D]
$\pi_2 \times K$	36	Complete generation cycle count [D]
$K/F$	3/5	Vertex/face ratio [D]
$\rho_{ij}$	integer	Hop-count distance ($\neq d_{ij}$) [D]
$S_n$	integer	Cumulative reflections, proto-time [D]

Open items from Section 3:

Lower bound $K \ge 12$: Step 3 (4+8=12 minimum uniform graph) [D]
Formal proof no uniform graph on $< 12$ vertices satisfies both dual-constraint and non-coplanar-frame simultaneously [D*]
Connection between generation count = 3 and $K=12$ saturation [D]
$\tau_{\text{mix}}$: full electroweak mixing phase period [O $\to$ §4.9]

3.17 Bridge to Section 4

Section 3 delivers to Section 4:

Fundamental structure: $K=12$, $H=132$, $D=3$; $\phi_1 = 1.618034$; $\phi_3 = 0.809017 = \phi_1/2$; $\Psi_{\text{sph}} = 0.9393$ and $\delta = 0.0607$; $\text{SIMPLEX} = 5/4$; $\dim(V_{\text{inv}}) = 44$, $\dim(V_{\text{geom}}) = 88$; $1/D = 1/3$ internal, $(D-1)/D = 2/3$ external.
Spectral structure: Adjacency eigenvalues $\{5, \sqrt{5}, -1, -\sqrt{5}\}$; Laplacian eigenvalues $\{0, 5-\sqrt{5}, 6, 5+\sqrt{5}\}$; Laplacian gaps $\lambda_2-\lambda_1 = 1+\sqrt{5}$ and $\lambda_3-\lambda_2 = \sqrt{5}-1$ (distinct from $\mu_2 = \sqrt{5}$); $\alpha_{\text{disc}} = (5-\sqrt{5})/40$.
Routing structure: Routing hierarchy lemma $\text{correction}(B) = \text{correction}(\text{terminal})/\phi_1$; lepton corrections $\delta(e \to \mu) = +\mu_2/(\phi_1 \times 17)$, $\delta(\mu \to \tau) = -\mu_2/17$ where $\mu_2 = \sqrt{5}$ is the $T_1$ adjacency eigenvalue; boson correction $\Delta\tau_Z = -\Delta\pi/\phi_1$; sign convention from §4.8.6 (forward reference); $\tau_{\text{mix}}$ as Level-2 placeholder for §4.9.
Geometry: Non-trivial holonomy (§3.12) seeding curvature; metric tensor $g_{ab}$ from correlation summation; hop-count distance $\rho_{ij}$ (distinct from time-averaged disagreement rate $d_{ij}$ of §2.5.1); geometric integrity boundary $K(K-1) \le 132$; $\pi_2 = 3$, $\pi_3 = 4$, $\pi_{\text{eff}}(12) = 3.105829$.
Chirality: CW period 3 ticks, CCW period 6 ticks — from the non-adjacent 2-tick cost of §3.11.1 — seeding the 2:1 charge ratio of §4.6.1.
Metric correction framework: §3.9.1–3.9.6 (three-layer particle-specific corrections) relocated to Section 4.2, where quarks, electrons, and proton routing context is established.

Section 4 uses this structure to derive stable motifs — recurring patterns in the binary flow sequence that persist across ticks within the $K=12$, $D=3$ framework. The transfer operator $T$ of Section 3 is the generating object; the particle spectrum of Section 4 is the set of stable deformations of $T$'s eigenstructure.

Section 3 derives all spatial structure from the relational dynamics of Section 2 via a single generating object: the adjacency transfer operator $T$, whose domain $K=12$ is established by the uniqueness theorem of §3.3. $T$ is defined generally in §3.2 and specialized to the icosahedral graph only after $K=12$ is proved. $K=12$ satisfies the lower bound from local determinacy (4 reference + 8 verification sites, with the minimum of 8 carrying [D*] status pending formal graph-theoretic completion), the upper bound from geometric integrity (exact self-closure $H=132$), and self-hosting uniqueness ($K=24$ excluded by $H=552>132$; $K=6$ excluded by failure of non-coplanar frame in $D \lt 3$). $D=3$ is derived by two independent routes — icosahedral axis rank = 3 and ternary closure kernel dimension = 3 — agreeing at $D = K/\pi_3 = 12/4 = 3$, with $\pi_3$ defined in §3.6.1. The Laplacian eigenvalue gap $\lambda_3-\lambda_2 = \sqrt{5}-1$ is explicitly distinguished from the adjacency eigenvalue $\mu_2 = \sqrt{5}$; the latter governs lepton generation correction magnitudes in §3.11.4. $\Psi_{\text{sph}} = 0.939326$ from the reduced isoperimetric formula, justified as the first-order (linear-scale) field propagation fraction. $\phi_d = (1+\sqrt{5})/\pi_d$ is derived as a scaling law: $\phi_1$ scaled by $\pi_1/\pi_d$. The symmetry decomposition $\dim(V_{\text{inv}}) = 44$, $\dim(V_{\text{geom}}) = 88$ is derived in §3.9 and referenced by §3.13. $\alpha_{\text{disc}} = \lambda_{T1}/(2F) = (5-\sqrt{5})/40$ is derived from the exact identity $\lambda_{T1}\cdot\lambda_{T3} = F$, which makes the $T_2$ decay factor exactly $1/2$ per tick, yielding $T_2/T_1$ ratio = $1/\phi_1$. The routing hierarchy lemma simultaneously closes $\delta(e \to \mu)$, $\delta(\mu \to \tau)$, and $\Delta\tau_Z$ to [D] status. Hop-count distance $\rho_{ij}$ is introduced as a symbol distinct from the §2.5.1 time-averaged disagreement rate $d_{ij}$. §3.9.1–3.9.6 are relocated to Section 4.2. Classical $\pi$ appears in §3.5.3, §3.6.1, §3.7, and §3.12 as a standard mathematical constant; it is not claimed as a TFP-derived quantity.

Section 3 — End
Temporal Flow Physics v12.11 · John Gavel

Section 2 — Local Dynamics of Binary Relations (TFP) (v12.8)

Section 2 — Local Dynamics of Binary Relations

Temporal Flow Physics v12.8

John Gavel

2.0 Overview

Section 1 established the irreducible primitives:

The unbounded relational domain — no imposed boundary on where comparison can occur (§1.1)
Comparison as the primitive act — the domain relating to itself (§1.2)
Binary relational states as the outcome of comparison: $F_i \in \{+1, -1\}$ (§1.2)
Sites as loci of resolved comparison, not prior entities (§1.3)
Primitive adjacency relations $i \sim j$ (§1.4–1.5)
Relational difference as the sole primitive observable (§1.4)
Discrete update steps (§1.7)
Finite relational capacity per site (§1.7)
Determinacy via overlapping relational closure (§1.6)
The Summation Principle: distinct parallel differences sum; never log discrete events (§1.8)

Section 2 derives all local dynamics from these primitives alone, within the minimal adjacency structure: each site has exactly two neighbors, the minimum required for difference propagation beyond direct adjacency. No coordinates, no metric, no spatial embedding, no dimensionality are introduced here. Every quantity is expressed purely in terms of sites, adjacency relations, binary states, and update counts. All distances are summations of discrete differences — never logarithms, which are statistical approximations inappropriate for a discrete substrate.

Why $K=2$ is the minimum: $K=1$ allows only trivial reflection $A \to B \to A$ — no propagation, nothing to derive. $K=2$ is the first structure where $A \to B \to C$ exists, mediated lag appears, and accumulated differences build across more than one hop. $K=2$ is therefore not assumed but derived as the minimum non-trivial adjacency structure from Axioms 1–10. The derivation of higher coordination numbers — culminating in $K=12$ — begins in Section 3.

Within this minimal structure three levels arise naturally:

Level 0: Binary substrate dynamics (§2.1–2.4)
Level 1: Relational accumulation and memory (§2.5–2.6)
Level 2: Emergent proto-temporal structure (§2.7–2.9)

Where a construct requires loops, triangles, or coordination beyond two neighbors, it is noted explicitly and deferred to Section 3, where geometry, distance, and higher coordination emerge from the correlation structure produced here.

Sections 2.0.1–2.0.3 establish three foundational results that precede the binary dynamics: the domain within which the $K=2$ structure operates (§2.0.1), the conservation law governing any comparison (§2.0.2), and the minimum scale at which a comparison resolves into a stable discrete event (§2.0.3). These three results ground the flip — the primitive operation of §2.1 — in the structure of comparison itself rather than in assumption.

2.0.1 The Unbounded Domain and Ranged Comparison

Section 1 established that the relational domain has no imposed boundary — comparison can occur anywhere within it, and no preferred center or edge exists (§1.1). Section 2 operates within this domain using the minimal $K=2$ adjacency structure. No boundary conditions are needed; the $K=2$ chain can extend without limit.

Two aspects of this domain are worth making precise before the dynamics begin:

Unbounded: The domain as a whole has no imposed edge. There is no wall beyond which comparison cannot occur, no preferred origin, no finite extent. Any site can in principle have adjacent sites; no site is special by position.

Ranged: Any specific comparison has finite reach — a local causal horizon. From Axiom 9 (finite relational capacity), a site can engage only a finite number of adjacency relations per update step. The influence of any site on the rest of the domain is therefore bounded at each tick. Information propagates at a finite rate through the adjacency structure — one hop per tick in the minimal $K=2$ chain.

These two aspects are not in tension. The whole domain is without limit; any process within it is locally bounded. The $K=2$ chain is locally ranged (each site reaches exactly two neighbors per tick) and globally unbounded (the chain can extend in either direction without end).

What is discrete and what is not:

Discreteness is not a property of the domain. It is what the comparison operation produces. Before comparison occurs, the domain is undifferentiated — not "smooth" in the geometric sense, but simply not yet structured by any distinction. Each comparison crystallizes a discrete distinction (a binary state, a site, a difference) within the domain. What has not been compared remains undifferentiated.

The $K=2$ chain is locally discrete: each update is a discrete event, each state is binary, each hop is a unit step. But the domain in which this chain is embedded does not acquire those properties globally. The chain is a discrete structure within an undifferentiated background — not a discrete approximation to a continuous manifold.

This distinction matters for what follows. The Pythagorean structure of §2.0.2 and the Euler coherence floor of §2.0.3 are both properties of the comparison operation within this domain — not properties of a pre-existing geometric space.

Operational hierarchy:

Unbounded relational domain (no edge, undifferentiated before comparison)
        │
        ├─> Any comparison: finite reach (causal horizon of that comparison)
        │         │
        │         ├─> Discrete event (binary state resolves)
        │         └─> Observable difference (local to that comparison)
        │
        └─> Beyond causal reach: undifferentiated until further comparison occurs

2.0.2 The Comparison Operation and Pythagorean Conservation

Any comparison is irreducibly triadic. Two aspects of the domain (call their relational extents $a$ and $b$) are compared; the result $c$ is what the comparison produces — distinct from either input alone. This triadic structure is not imported from geometry. It is what comparison IS: two things in relation, and the relation itself as a third thing.

Defining relational extent.
The first question is how to measure the "extent" of a binary relational process. Two candidates present themselves: the count of difference events, and the spread of the binary state distribution. The count is governed by the Summation Principle (§1.8) — it accumulates additively across sequential events in the same channel. A different quantity is needed here, because the comparison combines two independent processes into one result, not two sequential events in one process.

For two independent processes, the natural combination law is not addition of counts but addition of variances. This is not a choice — it is a theorem that holds for any two independent random quantities:

$$ \text{Var}(X + Y) = \text{Var}(X) + \text{Var}(Y) \quad \text{when } X \text{ and } Y \text{ are independent} \quad \text{[D]} $$

Independence holds at the instant of primitive comparison, prior to any closure-induced correlations (Axiom 4, Axiom 5). The combination of two comparison processes at a single site is therefore governed by variance addition, not count addition.

Variance of a binary process.
For a binary $\{+1, -1\}$ variable with equal probability (the symmetric case forced by Axiom 1, which specifies no preferred state), the variance is exactly 1:

$$ \text{Var}(F_i) = \mathbb{E}[F_i^2] - \mathbb{E}[F_i]^2 = 1 - 0 = 1 \quad \text{[D]} $$

This is not a choice or approximation. It is the unique second moment of a symmetric binary distribution. The variance of any binary relational process is therefore fixed at 1 per process — by the process being binary, not by any geometric assumption.

Defining relational extent as standard deviation.
We define the relational extent of a binary process as its standard deviation:

$$ \text{relational extent} := \sigma \quad \text{where} \quad \sigma^2 = \text{Var}(\text{process}) \quad \text{[Def]} $$

For a single binary process, $\sigma = 1$. For a compound process built from $n$ independent binary comparisons, $\sigma^2 = n$. This definition is forced by the requirement that relational extent combine correctly under independence: only standard deviation satisfies $\sigma^2(\text{combined}) = \sigma^2(a) + \sigma^2(b)$ for independent $a, b$.

The triadic variance identity.
Consider a comparison between two independent binary relational processes with extents $a$ and $b$, producing result $c$. Applying variance addition (Axiom 4):

$$ \sigma^2(c) = \sigma^2(a) + \sigma^2(b) $$

In terms of relational extents ($\sigma(a) = a$, $\sigma(b) = b$, $\sigma(c) = c$ by definition):

$$ a^2 + b^2 = c^2 \quad \text{[D]} $$

This is a substrate-internal identity. It follows from Axiom 1 (binary states, fixing $\text{Var} = 1$ per process), Axiom 4 (locality — independence of non-adjacent processes at the moment of comparison), and the definition of relational extent as standard deviation. No geometric embedding, no Euclidean norm, no metric is assumed. The quadratic form emerges because variance — not count — is the correct measure of relational extent for independent parallel processes, and variance addition is an exact theorem, not an approximation.

Why not $a + b = c$?
Linear addition holds for counts of sequential discrete events (Summation Principle, §1.8): two sequential disagreements accumulate as $1 + 1 = 2$. But the comparison operation combines two independent processes into a single result — not two sequential events. Independence is the key distinction. For independent processes, counts do not add to give the result's count; variances do. The Summation Principle and the variance identity govern different structural regimes and are complementary, not conflicting.

Why not $\max(a,b) = c$?
The max-norm would hold if one process completely dominated the comparison and the other contributed nothing to the result. By Axiom 1, the binary comparison has no preferred state and no preferred input — both aspects $a$ and $b$ contribute independently and symmetrically to $c$. When both contribute, neither dominates, and the correct combination is variance addition, not max.

The causal horizon and budget.
The comparison is bounded by its causal horizon — the ranged reach established in §2.0.1. The horizon places an upper bound on $c$: no single comparison can produce a result whose relational extent exceeds what the causal reach can support. When a comparison uses its full available budget, $c$ is at this ceiling and the budget equation reads:

$$ a^2 + b^2 = c^2, \quad c \text{ bounded above by the causal horizon} \quad \text{[D]} $$

The comparison partitions $c^2$ between $a^2$ and $b^2$. Nothing is created; the budget is partitioned, not generated. The horizon fixes the scale; the variance identity fixes the partition rule.

The flip (§2.1) is where the partition $a^2:b^2$ reaches an extremum — a turning point in how the budget is distributed between the two input processes. The delta-budget framework of Sections 4–5 is the direct expression of this conservation law across assembled stable motifs.

2.0.3 The Euler Coherence Floor and the Origin of $\lambda_p$

A turning point requires a neighborhood. Euler's condition (Methodus Inveniendi, 1744) states that a minimum or maximum is inseparable from the interval over which the reversal is established. Applied to the comparison operation of §2.0.2: the partition $a^2:b^2$ cannot reach a stable extremum without a minimum interval over which the reversal can be confirmed. Below that interval, the partition does not stabilize; no discrete event resolves; no flip occurs.

That minimum interval is $\lambda_p$ — the scale at which the comparison operation first closes on a stable Euler extremum:

$$ \text{Minimum coherent comparison scale:} \quad \lambda_p \quad \text{[D]} $$ $$ \text{Total conserved comparison budget:} \quad c^2 \quad \text{[D]} $$ $$ \text{Flip} = \text{Euler extremum in } c^2\text{-partition at scale} \ge \lambda_p \quad \text{[D]} $$

Below $\lambda_p$: no stable discrete structure. At $\lambda_p$: one flip, the simplest stable event. Above $\lambda_p$: compound motifs built from multiple independent flips.

$\lambda_p$ is not a physical scale imported from outside the framework. It is a structural property of the comparison operation: the minimum relational extent required for a comparison to resolve into a stable discrete event. Its identification with the physical Planck length is established in Section 7 through calibration.

The domain has two Euler extrema: the maximum budget per single comparison (the Planck scale) and the minimum stable closed structure (the proton assembly scale, §4). Conservation lives in $c^2$, so the ratio of these two extrema gives:

$$ \alpha_G = \left(\frac{m_p}{m_{\text{Planck}}}\right)^2 \approx 5.9 \times 10^{-39} \quad [\to\text{§7.4.1}] $$

This ratio is a consequence of the two-extremum structure of the comparison operation. Its derivation is completed in §7.4.1, where the physical identification of both scales is established.

2.1 Binary Relational States and Primitive Tension

Each site $i$ carries a binary state:

$$ F_i(t) \in \{+1, -1\} $$

where $t$ is the discrete update index — a proto-time counter, not physical time. The binary state is the outcome of comparison (§1.2); it is the discrete label produced when a comparison in the unbounded domain (§2.0.1) resolves at scale $\ge \lambda_p$ (§2.0.3).

For any adjacent pair $(i, j)$ with $i \sim j$, the primitive relational difference is:

$$ x_{ij}(t) = F_i(t) \text{ XOR } F_j(t) \in \{0, 1\} $$

Equivalently and exactly:

$$ x_{ij}(t) = \frac{1 - F_i(t) \cdot F_j(t)}{2} $$

$x_{ij} = 0$ means agreement. $x_{ij} = 1$ means disagreement. This is exact for all four combinations of $\{+1,-1\}$ states with no approximation.

Local tension at site $i$:

$$ T_i(t) = \sum_{j \sim i} x_{ij}(t) $$

In the minimal adjacency structure, each site has exactly two neighbors. Therefore $T_i \in \{0, 1, 2\}$:

$T_i = 0$: both neighbors agree with $i$ (fully aligned)
$T_i = 1$: one neighbor agrees, one disagrees
$T_i = 2$: both neighbors disagree with $i$ (fully anti-aligned)

Alignment minimizes tension. Tension is the sole primitive measure of local relational stress.

2.2 Primitive Update Rule (Local Action Descent)

Define the discrete action functional on the adjacency structure:

$$ S = \sum_{\text{edges } (i,j)} \frac{1}{2}(F_i - F_j)^2 $$

For $F_i, F_j \in \{+1,-1\}$ this reduces to: $0$ if $F_i = F_j$ (agreement), $2$ if $F_i \neq F_j$ (disagreement). So $S = 2 \times$ (number of disagreeing edges). Minimizing $S$ is equivalent to minimizing total tension.

Charge conservation:
Define global charge $Q = \sum_i F_i$. The only admissible elementary operation is a simultaneous pairwise flip of an adjacent disagreeing pair:

$$ F_i \to -F_i, \quad F_j \to -F_j \quad (\text{only when } F_i \neq F_j) $$

Since $F_i \neq F_j$ implies $F_i + F_j = 0$ exactly, $\Delta Q = -2(F_i + F_j) = 0$. Charge is exactly conserved per flip.

Local action change for a potential flip of pair $(i, j)$:

$$ \Delta S_{ij} = -8 \times (A_i + A_j - D_i - D_j) $$

where $A_i =$ agreements at $i$ excluding $j$, $D_i =$ disagreements at $i$ excluding $j$.

Note: The formula $\Delta S = -8 \times (A_i+A_j-D_i-D_j)$ is expressed in action units where each disagreeing edge contributes 2. In counting units (each edge contributes 1), the acceptance criterion reduces to: accept the flip when $D_i+D_j > A_i+A_j$. Both forms describe identical dynamics.

Accept flip $(i,j)$ when: $D_i + D_j > A_i + A_j$ (both sites are "social outliers")

which follows directly from $\Delta S_{\text{actual}} = (A_i-D_i) + (A_j-D_j) < 0$.

In the minimal structure, each site has exactly one other neighbor beyond the flip partner. The four configurations and their $\Delta S$ values:

Left neighbor	Right neighbor	$\Delta S_{ij}$	Interpretation
$i$ agrees with left	$j$ agrees with right	0	Neutral — neither helps nor costs
$i$ disagrees w. left	$j$ agrees with right	+16	Costly — flip increases action
$i$ agrees with left	$j$ disagrees w. right	−16	Favorable — flip decreases action
$i$ disagrees w. left	$j$ disagrees w. right	0	Neutral — costs cancel

Note on interval and pairing structure: The outcome of any flip depends not on the pair in isolation but on the full local configuration. Over the complete range of configurations the outcomes are symmetric (sum to zero). Within any given configuration they are asymmetric. This sensitivity to interval and pairing structure is the mechanism by which stable patterns resist flipping and persist across ticks. Table uses counting-unit normalization ($\pm 1$ per edge). Action units scale by factor 8. Acceptance criterion $D_i+D_j > A_i+A_j$ is identical in both representations.

Deterministic zero-temperature limit:

If $\Delta S_{ij} < 0$ and $F_i \neq F_j$: flip executes
Otherwise: state unchanged

Properties proven: binary constraint preserved, global charge conserved, total tension non-increasing for $T_{\text{eff}} = 0$, all dynamics strictly local and asynchronous.

2.3 Stochastic Extension

To include substrate fluctuations, introduce effective temperature $T_{\text{eff}} \ge 0$. Each adjacent pair $(i,j)$ fires according to its local process. When a firing event occurs, the flip is accepted with probability:

$$ P(\text{flip} \mid \text{firing}) = 1 \quad \text{if } \Delta S_{ij} < 0 $$ $$ P(\text{flip} \mid \text{firing}) = \exp(-\Delta S_{ij} / T_{\text{eff}}) \quad \text{if } \Delta S_{ij} \ge 0 $$

Temperature regimes: $T_{\text{eff}} = 0$ (fully deterministic), $T_{\text{eff}} \to \infty$ (effectively random), $T_{\text{eff}} \sim \mathcal{O}(1)$ (critical fluctuating). Equilibrium distribution: $P_{\text{eq}} \propto \exp(-T_{\text{total}} / T_{\text{eff}})$. Detailed balance holds locally.

2.4 Proto-Mass and Directed Transport

2.4.1 Flip indicator and proto-mass

Define the flip indicator at site $i$:

$$ R_i(t) = \begin{cases} 1 & \text{if } F_i(t + \tau_0) \neq F_i(t) \\ 0 & \text{else} \end{cases} $$

Time-averaged flip rate (proto-mass density):

$$ M_i = \frac{1}{W} \sum_{n=0}^{W-1} R_i(t_0 + n\tau_0) $$

$M_i \in [0, 1]$, dimensionless. Mass is operational — it counts sustained relational activity. Mass is not substance. It is the summation of flip events over an interval — the accumulated cost of relational difference maintained across ticks. In the language of §2.0.3, it is the accumulated count of Euler extrema resolved at site $i$ per unit window.

2.4.2 Directed transport bias (proto-velocity)

Flip activity need not be symmetric across adjacency. Define signed transport across edge $(i \to j)$:

$$ T_{ij}(t) = R_i(t) \cdot x_{ij}(t) $$

Net propagation bias at site $i$:

$$ v_i(t) = \frac{1}{|\text{Adj}(i)|} \sum_{j \sim i} \text{sgn}(i \to j) \cdot T_{ij}(t) $$

2.4.3 Proto-momentum

$$ p_i(t) = M_i \cdot v_i(t) $$

Momentum is conserved locally up to boundary flux and saturation-induced stutter:

$$ \sum_{i \in \Omega} p_i(t + \tau_0) - \sum_{i \in \Omega} p_i(t) = -\Phi_{\text{boundary}}(t) + S_{\text{stutter}}(t) $$

2.5 Accumulated Differences — Relational Memory

2.5.1 Accumulated relational difference (summation)

$$ A_{ij}(t) = \sum_{n=0}^{t} x_{ij}(n) $$

This is a running integer count of disagreements between $i$ and $j$ across all ticks up to $t$. It is the primary measure of relational history. This is the correct form of relational distance in the discrete substrate — a direct application of the Summation Principle (§1.8). $A_{ij}(t)$ is always a non-negative integer. No approximation.

Operational distance between adjacent sites $i$ and $j$:

$$ d_{ij}(t) = \frac{A_{ij}(t)}{t} = \text{time-averaged disagreement rate} \in [0, 1] $$

For a mediated path $i \to k \to j$: the minimum cost is 2 ticks. The lag is a physical fact about the discrete structure, not a statistical property.

2.5.2 Proto-time ordering (local causal order)

Each accepted local flip defines a primitive causal update event: Event $n \to$ Event $n+1$. Causality arises because State($n$) determines State($n+1$) through the local flip rule.

Coarsening order parameter:

$$ S_{\text{coarse}}(t) = -\frac{1}{N} \sum_i T_i(t) $$

$S_{\text{coarse}}$ is non-decreasing under local dynamics for $T_{\text{eff}} = 0$, non-decreasing on average for $T_{\text{eff}} > 0$.

2.5.3 Why this is not the arrow of time

The monotonic tendency of $S_{\text{coarse}}$ does not imply fundamental time asymmetry. The microscopic rules are locally reversible. Irreversibility appears only statistically, from combinatorial bias toward tension-reducing moves, local acceptance asymmetry, and domain coarsening dynamics. The apparent arrow arises from ensemble bias, not from an imposed global direction.

2.6 Domain Structure and Coarsening

Under tension minimization, disagreeing pairs flip toward agreement, driving domain coarsening. Domain merging is favored; fragmentation is suppressed.

Coarsening dynamics: For $T_{\text{eff}} \sim 0$: $R_{\text{domain}}(t) \sim (D \cdot t)^{1/2}$ (Allen-Cahn). For $T_{\text{eff}} > 0$: $R_{\text{domain}}(t) \sim (D \cdot t)^{1/z}$, $z \ge 2$. $R_{\text{domain}}$ is measured in adjacency steps. [D]
Continuum approximation of discrete domain-wall motion. Exact substrate scaling verified in §3.2 via $K=2$ chain enumeration.

2.7 Relational Correlations

Temporal cross-correlation between sites $i$ and $j$:

$$ C_{ij}(\tau) = \langle x_{ij}(t) \cdot x_{ij}(t + \tau) \rangle $$

This measures how persistently $i$ and $j$ maintain a disagreement relationship across $\tau$ update steps. Operational relational similarity:

$$ d_{ij} = \frac{1}{t} \sum_{n=0}^{t} x_{ij}(n) $$

Sites that persistently agree have small $d_{ij}$ (approaching 0). Sites that persistently disagree have large $d_{ij}$. Mediated paths accumulate cost additively: $d_{AC} \ge d_{AB} + d_{BC}$.

The correlation structure $C_{ij}(\tau)$ is the seed from which Section 3 derives geometry.

2.8 Reflection Phase and Minimal Relational Closure

In the minimal adjacency structure, sites $A$, $B$, $C$ form a chain: $A \sim B \sim C$. There are no closed triangles. The only closed relational loop is reflection: $A \to B \to A$.

Reflection phase: $\phi_{i \to j} = \pi \cdot x_{ij}$. For a reflection $A \to B \to A$: $H_{ABA} = \phi_{A \to B} + \phi_{B \to A} = 2\pi \equiv 0 \pmod{2\pi}$. Holonomy in the minimal structure is always trivially zero. Curvature requires genuine triangular loops, first appearing in Section 3.

The mediated lag encoded by reflection:

$$ \text{cost}(A \to C) = \text{cost}(A \to B) + \text{cost}(B \to C) = 1 + 1 = 2 \text{ ticks minimum} $$

This is the irreducible lag of the minimal adjacency structure and the foundation from which Section 3 derives $c$.

The minimal propagation pattern in the $K=2$ chain is the standing wave $A \to B \to C \to B \to A$ (period 4 ticks): the excitation bounces between non-adjacent endpoints $A$ and $C$ via the mediating site $B$. Forward and backward strokes are equal — no net winding accumulates. This is consistent with trivial holonomy in the $K=2$ structure. Directed chirality (non-zero net winding) requires genuine triangular loops, first appearing in Section 3.

2.9 Stability and Informational Friction

$$ \delta_i = \frac{1}{W} \sum_{t=0}^{W} \frac{T_i(t)}{K_{1D}} $$

$\delta_i \in [0, 1]$: 0 = always aligned (stable), 1 = always in tension (unstable). This is a pure summation over ticks — no eigenvalues, no logarithm.

Stable coherent motifs persist when the sustained disagreement fraction $\delta$ remains below the threshold determined by the pairing structure of the local interval. The precise characterization of flip-resistant pairing structures is developed in Section 3.

2.10 Finite Relational Capacity and the Mediated Lag

From Axiom 9, each site has finite relational capacity:

$$ K_{1D} = 2 \quad (\text{exactly two neighbors}) $$ $$ H_{1D} = K_{1D} \cdot (K_{1D} - 1) = 2 \times 1 = 2 $$

Each site can maintain two adjacency relations but can actively resolve only one per tick. The stutter: when both neighbors demand updates in the same tick, only one can be served. The other accumulates as unresolved latency.

Tick 1: B resolves with A
Tick 2: B resolves with C
A's influence on C costs a minimum of 2 ticks

Stutter sink in the momentum balance:

$$ p_i(t + \tau_0) - p_i(t) + \sum_{j \sim i} J_{ij}(t) = -s_i(t) $$ $$ s_i(t) = \kappa_s \cdot \delta_i(t) \cdot p_i(t) $$

When $\delta_i = 0$: capacity is not exceeded, momentum locally conserved. When $\delta_i > 0$: directed transport converts to unresolved relational latency — the primitive origin of inertia.

§2.10.1 Sequential Engagement and the Relay Mechanism

The stutter of §2.10 describes what happens when both neighbors demand resolution simultaneously. Before that limit is reached, a more fundamental dynamic governs how a site with two neighbors moves through its comparisons. This dynamic is the relay — and it is the origin of all propagation in the $K=2$ structure.

The Sequential Engagement Rule

A site $B$ with neighbors $A$ and $C$ cannot resolve both adjacency relations in the same tick (Axiom 9, finite capacity). The two engagements are therefore sequential: $B$ resolves with one neighbor in tick $t$, and with the other in tick $t+1$.

The order of engagement is governed by local tension. In each tick, $B$ directs its relational capacity toward whichever neighbor presents the higher unresolved tension:

$$ \text{If } T_{BA}(t) > T_{BC}(t): \quad B \text{ engages } A \text{ in tick } t, C \text{ in tick } t+1 $$ $$ \text{If } T_{BC}(t) > T_{BA}(t): \quad B \text{ engages } C \text{ in tick } t, A \text{ in tick } t+1 $$ $$ \text{If } T_{BA}(t) = T_{BC}(t): \quad \text{tied; one is selected by local state parity} $$ $$ \text{(} F_B \text{ determines priority when tensions are equal)} $$

This is not a scheduling rule imposed from outside. It is the direct consequence of Axiom 9 (each site resolves one comparison per tick) and Axiom 4 (all updates are local) operating on the available tension at each adjacency.

What the Waiting Neighbor Receives

While $B$ is engaged with $A$ in tick $t$, site $C$ is not disconnected from $B$. $C$ continues to carry its current relational state and experience tension relative to $B$'s current state. The difference $x_{BC}(t)$ is still defined; $C$'s tension $T_C(t)$ still accumulates.

What $C$ cannot do in tick $t$ is act on that tension through $B$ — it cannot flip with $B$ while $B$ is otherwise occupied. The tension at $C$'s $B$-side accumulates as unresolved demand:

$$ \text{demand}_C(t) = T_{CB}(t) \quad \text{while } B \text{ is engaged with } A $$

This unresolved demand is what $B$ faces when it turns to $C$ in tick $t+1$. If $\text{demand}_C$ has grown large enough ($D_C > A_C$ in counting units), $B$ will flip with $C$.

The Relay: B Carries A's Resolution to C

The critical structural fact is this:

$$ B\text{'s state at the end of tick } t \text{ — after resolving with } A \text{ — is the state } C \text{ reads when } B \text{ engages } C \text{ in tick } t+1. $$

This is not a separate transport mechanism. It is a direct consequence of how states update sequentially. $B$ compares with $A$, updates (or does not update) $F_B$, and then presents that updated $F_B$ to $C$. $A$'s influence on $C$ travels through $B$'s state, not through any direct channel.

The relay therefore has a fixed cost structure:

Tick t:     B resolves with A → F_B(t) updated (or confirmed)
Tick t+1:   B resolves with C, presenting F_B(t) → C updates (or confirms)

Minimum cost for A's influence to reach C via B:   2 ticks               [D]

This is not an approximation. It is exact: there is no mechanism by which $A$'s tick-$t$ state can reach $C$ in fewer than 2 ticks through a $B$ that has only one comparison capacity per tick.

Chain Extension: Operational Distance from Sequential Relays

The relay logic extends along the full $K=2$ chain. Consider a longer chain $A \sim B \sim C \sim D$.

If $B$ relays $A$'s influence to $C$ in 2 ticks, and $C$ must then relay to $D$, $C$ faces the same constraint $B$ did: $C$ is already occupied with its $B$-side engagement when $D$'s demand arrives, or vice versa.

The minimum cost for $A$'s influence to reach $D$:

A→B:   1 tick  (direct adjacency, one comparison)
B→C:   1 tick  (B relays; C reads B's updated state)
C→D:   1 tick  (C relays; D reads C's updated state)
Total: 3 ticks minimum

Generalizing: for a chain of $n$ sites, $A$'s influence reaches the $n$th site in $n-1$ ticks minimum. This is the discrete operational distance:

$$ d_{\text{chain}}(n) = n - 1 \text{ ticks minimum} \quad \text{[D]} $$

Each adjacency hop costs exactly one tick of relay delay. The accumulated difference count $A_{ij}(t)$ of §2.5.1 is the time-integral of this relay cost — the running record of how many times $B$ has carried a disagreement from $A$'s side to $C$'s side. Relational distance is accumulated relay cost, not a pre-existing metric.

Why the Relay Is Not the Stutter

The stutter (§2.10) occurs when both neighbors demand resolution in the same tick and only one can be served — the other is deferred. The relay is the normal operating condition: $B$ moves sequentially through its two neighbors, each tick resolving with one and implicitly updating the state that the other will read next.

Relay:   B's sequential engagement, one neighbor per tick.
         Both neighbors are eventually served; the information propagates.
         This is the mechanism of propagation.

Stutter: B's capacity is exceeded because both demands arrive together.
         One comparison is deferred beyond its natural relay position.
         This is the mechanism of effective inertia — §2.10's stutter sink.

The stutter is the relay breaking down under demand overload. When the relay runs cleanly — each neighbor served in its natural sequential turn — propagation is smooth and the 2-tick minimum holds exactly. When the relay overloads — both neighbors demanding the same tick — the latency exceeds the minimum, and the excess becomes the unresolved accumulation that §2.10 quantifies as $s_i(t)$.

The Internal/External Reading of the Relay

The relay mechanism is where the internal/external distinction of §1.9 first becomes concrete.

$B$'s comparison with $A$ is internal to the $A$–$B$ relation. It produces a state change at $B$. $B$'s subsequent comparison with $C$ is the external propagation of that state change — the relational result leaving $B$'s immediate vicinity and entering $C$'s local domain.

The relay is therefore the mechanism by which internal comparison becomes external propagation. A comparison that resolves at $B$ stays internal if $C$ never reads $B$'s updated state (no relay). A comparison that relays through $B$ becomes external when $C$ reads and responds to $B$'s new state.

This is not metaphorical. In a $K=2$ chain:

Internal:  The B-site flip (or non-flip) in response to A.
           The tension T_B changes; A's influence is absorbed at B.

External:  B's updated state reaching C in the next tick.
           A's influence propagates beyond B into C's domain.

The ratio of absorbed-to-propagated comparison outcomes is the primitive expression of the $\delta / \Psi_{\text{sph}}$ split that Section 3 derives from icosahedral geometry. Here, at the $K=2$ level, the split is not yet quantifiable — but its mechanism exists in every relay.

Axiomatic Trace

Construct	Source	Axiom
Sequential engagement order	Tension-directed, one comparison per tick	A4, A9
Waiting neighbor accumulates demand	State persists; tension defined even when idle	A2, A3, A9
$B$ carries $A$'s state to $C$	$F_B(t)$ is $B$'s state after tick $t$; $C$ reads it	A2, A5, A8
2-tick minimum relay cost	One tick per hop; sequential not parallel	A8, A9
Chain distance = accumulated relay	$d = n-1$ hops, each 1 tick; summation not metric	A6, A8, A9
Stutter = relay overload	Both demands same tick; one deferred	A9
Internal/external split first concrete	Relay = propagation; non-relay = absorption	A9, §1.9

2.11 Anchoring and Global Stability

Without a stabilizing constraint, the global orientation $\Theta = \sum_i F_i$ undergoes a random walk. Anchoring modifies flip probability:

$$ P(\text{flip at } i) \to P(\text{flip at } i) \cdot [1 + \lambda \cdot \text{sign}(F_i) \cdot (\Theta - \Theta_0)] $$

where $\lambda \ll 1/N$ (weak coupling) and $\Theta_0$ is the reference orientation. This is the unique lowest-order perturbation that breaks global $F_i \to -F_i$ symmetry while preserving all local axioms. Anchoring is the seed of what later becomes mass normalization and gauge fixing.

2.12 Emergent Effective Parameters

From the substrate quantities derived above:

Intrinsic causal rate:

$$ c_{\text{int}} = 1 \text{ adjacency-step per tick} $$

This is exact and derives directly from the minimal adjacency structure — from the ranged reach of a single comparison (§2.0.1) and the discrete update structure (Axiom 8). No faster rate is possible: a site can compare with at most one adjacent site per tick. In the language of §2.0.2, $c_{\text{int}} = 1$ is the unit of causal reach — the minimum interval a comparison spans per tick. The full physical speed $c$ is calibrated in Section 5.

Effective inertia:

$$ I_{\text{eff}} \sim \delta $$

The full expression $I_{\text{eff}} \sim \delta \cdot \text{TF} \cdot R_{\text{domain}}^2$ requires spatial extent and loop structure not yet available in Section 2.

Effective coupling:

$$ G_{\text{eff}} \sim \delta $$

The full coupling requires Section 3's geometric structure. All three quantities are dimensionless. Physical calibration is deferred to Section 5.

2.13 Axiomatic Closure

Every construct in Section 2 traces to Section 1 primitives. Entries from §2.0.1–2.0.3 are marked:

Construct	Origin	Axiom
Unbounded domain / ranged comparison [§2.0.1]	No imposed boundary; finite causal reach	A1, A4, A9
$c^2$ — conserved comparison budget [§2.0.2]	Causal horizon upper bound per comparison	A1, A3, A9
Triadic comparison structure: $a^2+b^2=c^2$ [§2.0.2]	Variance addition of independent binary processes (A1, A4)	A1, A2, A3
$\lambda_p$ — minimum coherent comparison scale [§2.0.3]	Euler extremum condition; Axiom 9 capacity	A3, A8, A9
$\alpha_G = (m_p/m_{\text{Pl}})^2$ [→§7.4.1]	Two-extremum structure of comparison	§2.0.3, §7.4.1
$T_i$ (tension)	Binary disagreement count	A2, A3
$x_{ij} = (1-F_iF_j)/2$	Exact algebraic form	A2
Pairwise flip rule	Action minimization on graph	A3, A7, A9
Charge conservation	$\Delta Q = 0$ proven exactly	A2
$M_i$ (proto-mass)	Time-averaged flip rate (summation)	A2, A8, A9
$v_i$ (proto-velocity)	Directed transport bias	A3, A4, A5
$p_i$ (proto-momentum)	$M_i \cdot v_i$	Derived
$A_{ij}$ (memory)	Accumulated disagreement count (summation)	A6, A8
$d_{ij}$ (relational cost)	$A_{ij}/t$ — summation, not log	A6, A8
$S_{\text{coarse}}$ (ordering)	$-$mean tension, monotonic	A7, A8
$C_{ij}$ (correlation)	Temporal persistence of $x_{ij}$	A2, A8
$\delta$ (stability)	Time-averaged tension fraction	A9
$H_{1D} = 2$	$K_{1D}(K_{1D}-1)$, $K_{1D}=2$	A9
$s_i$ (stutter sink)	Capacity overflow latency	A9
$c_{\text{int}} = 1$ step/tick	Direct adjacency rate; ranged comparison	A3, A8

No free parameters. No spatial embedding. No metric. No dimensionality assumption. No logarithms.

2.14 Bridge to Section 3

Section 2 delivers to Section 3 the following purely relational ingredients:

Binary states $F_i(t)$ and their update dynamics
Proto-mass $M_i$ and proto-momentum $p_i$ per site
Accumulated differences $A_{ij}(t)$ as exact integer counts of relational history
Temporal correlations $C_{ij}(\tau)$ between site pairs
Operational relational cost $d_{ij} = A_{ij}/t$ as summation
Stability margin $\delta_i$ as time-averaged tension fraction
Causal ordering from event sequence
Intrinsic rate $c_{\text{int}} = 1$ step/tick
The mediated lag: minimum 2 ticks for non-adjacent influence

From §2.0.1–2.0.3, Section 3 additionally receives:

The unbounded relational domain — no boundary conditions required (§1.1, §2.0.1)
The triadic comparison structure $a^2+b^2=c^2$ as the conservation law for any comparison (§2.0.2)
The minimum coherent comparison scale $\lambda_p$ at which discrete events first resolve (§2.0.3)
The gravitational coupling ratio $\alpha_G$ as a two-extremum structural property, to be completed in §7.4.1

Section 3 uses the full ingredient set to derive:

Emergent spatial geometry from the correlation structure of $d_{ij}$ across multiple interacting chains
The transition from $K_{1D}=2$ to $K=12$ via ternary closure requirements
Distance as a full metric tensor rather than a scalar cost
Curvature from non-trivial holonomy around genuine triangular loops
Dimensionality from the minimum closure structure: $D=3$
Motifs as flip-resistant pairing structures — the foundation of particle stability
The speed $c$ from the mediated lag structure, and its invariance from saturation of the tick budget

Section 2 operates within the minimal adjacency structure ($K=2$). All geometry, dimensionality, and multi-chain structure begins in Section 3.

Section 2 — End
Temporal Flow Physics v12.8 · John Gavel

htmljava