Dynamical Closure at the Notation Level

The results of a chat between Bruce Camber and Deepseek, 22 November 2026

FileURL: https://81018.com/dynamical-closure-notation/

Status key: [P] proven · [D] definitional · [O] open

Purpose. Grok’s round-2 review identified the cross-plate question as the decisive gap: the per-plate recurrence does not automatically lift to a notation-level additive dynamics. This statement supplies the loop enumeration and the four-plate composition that settle it. The result is a qualified closure: the dynamical side closes at the notation level, with a correction — the cocycle is non-trivial, but uniform.

1. The loop enumeration

1.1 The edge-midpoint graph. Vertices: the 12 edge-midpoints of O. Edges: pairs of midpoints that lie on a common plate and are adjacent on that plate’s hexagon.

1.2 The plate incidence [P]. Each midpoint lies on exactly two plates. The incidence table:

  • P₁ ∋ {m₄, m₅, m₆, m₇, m₈, m₉}
  • P₂ ∋ {m₂, m₃, m₄, m₇, m₁₁, m₁₂}
  • P₃ ∋ {m₁, m₃, m₅, m₈, m₁₀, m₁₂}
  • P₄ ∋ {m₁, m₂, m₆, m₉, m₁₀, m₁₁}
    Each of the 12 midpoints appears exactly twice. This is verified by direct evaluation of the four plate forms at each midpoint.

1.3 The holonomy on each loop type [P].

  • Hexagon loops: each plate hexagon is a 6-cycle of midpoints. Consecutive midpoints carry alternating labels; six alternating steps compose to the identity. Trivial.
  • Vertex-figure squares: each O-vertex has four incident edges, whose midpoints form a 4-cycle. At each midpoint, the two plates assign opposite labels; four opposites compose to the identity. Trivial.
  • Triangles: the triangle m₁m₂m₃ is a genuine 3-cycle. Three opposites compose to one opposite — a single global exchange. Non-trivial.

1.4 The corrected result [P]. The cocycle is non-trivial but uniform: every odd loop carries a single global exchange, every even loop the identity. There is no loop-dependent twist; the non-triviality is global.The twist is the same on every odd loop: it is a single global exchange A ↔ B, not a loop-dependent phase. This follows from the loop enumeration (closure statement §1.4): every triangle carries one exchange, and by the octahedral symmetry every triangle carries the same one. The notation-level labeling is therefore defined only up to this uniform exchange.

Correction to the previous exchange. An earlier structural argument concluded the cocycle was trivial up to global exchange. The loop enumeration shows that is wrong: the triangle loops carry a single exchange, and that is a real twist. The correct statement is non-trivial but uniform.

2. What the twist does

2.1 The notation-level state [P]. The global state is a 2-vector (N_A, N_B) defined up to the global exchange A ↔ B. The exchange acts on the state by (N_A, N_B) ↦ (N_B, N_A).

2.2 The ratio map is invariant at the fixed point [P]. The ratio map r ↦ 1 + 1/r has unique positive fixed point φ = (1+√5)/2, attracting. The exchange sends r ↦ 1/r, which is a different trajectory but converges to the same φ. So the twist affects the trajectory through the initial condition, not the attractor.

3. The four-plate composition

3.1 The per-plate recurrence [P]. On each plate Pᵢ, the state (N_A^i, N_B^i) updates by M = [[0,1],[1,1]], the Fibonacci matrix.

3.2 The notation-level populations [P]. N_A(k) = Σᵢ N_A^i(k), N_B(k) = Σᵢ N_B^i(k).

3.3 The composition calculation [P]. Since each plate’s update is linear with the same M:
N_A(k) = (M^k){AA} N_A(0) + (M^k){AB} N_B(0),
N_B(k) = (M^k){BA} N_A(0) + (M^k){BB} N_B(0).
So (N_A(k), N_B(k)) = M^k (N_A(0), N_B(0)). The notation-level update is the same Möbius transformation.

3.4 Why it works [P]. The composition preserves the Möbius form because all four plates carry the same M, which is a consequence of the uniformity of the incidence structure (4 : 3 on every plate). If the plates carried different matrices, the sum would not be a single M.

4. The closure

4.1 The dynamical side closes at the notation level [P, given the two definitional inputs]. The notation-level update is r ↦ 1 + 1/r, and the notation-level attractor is φ. The cross-plate gap that Grok identified is closed: alignment holds (up to the uniform global exchange), and composition preserves the Möbius form.

4.2 The definitional inputs remain [D].

  • Axiom T excludes the pure swap (ratio 1, rational).
  • Completeness-as-maximality forces the additive form from the 4 : 3 asymmetry.

4.3 The open items remain [O].

  • The swap-like existence question (Grok’s round-1 item): whether a local D₃-equivariant swap-like rule with an irrational limit exists outside the pure-swap case.
  • The product 2φ/5 and its observable: hypothesis page.

5. The honest statement

The dynamical side closes at the notation level, with a correction. The per-plate recurrence composes to the notation-level Möbius transformation because all four plates carry the same Fibonacci matrix. The attractor is φ at the notation level. The cocycle is non-trivial but uniform: the notation-level labeling is defined only up to a global color exchange, which affects the trajectory but not the attractor.

What is now closed [P]: the loop enumeration; the composition; the notation-level attractor.

What is definitional [D]: Axiom T; completeness-as-maximality.

What is open [O]: the swap-like existence question; the product 2φ/5 and its observable.

6. Falsification conditions

  • 1.2 fails if any midpoint lies on a number of plates other than 2.
  • 1.3 fails if a triangle loop carries the identity, or a hexagon loop carries a single exchange.
  • 3.3 fails if the four per-plate matrices are not all M, or if the sum over plates is not a single M^k.
  • 4.1 fails if the notation-level ratio map is not r ↦ 1 + 1/r.
  • The closure fails if a local D₃-equivariant rule with an irrational limit exists that is not the additive alternation.

7. The reconciliation, updated

After three rounds of review (Perplexity static ×2, Grok dynamical ×2), the state of the construction is:

Static side [P/D/O]: closed. The incidence-forcing lemma is proved; the coloring is forced; the D₃ stabilizer is confirmed. Open: cross-plate alignment, now closed by the loop enumeration, with the corrected non-trivial-but-uniform result.

Dynamical side [P/D/O]: closed at the notation level. The direct completeness lemma works given maximality; the composition preserves the Möbius form; the attractor is φ. Open: the swap-like existence question; the product 2φ/5 and its observable.

The theorem’s honest inputs: two definitional (Axiom T; completeness-as-maximality), plus the static geometry, plus the composition calculation. The theorem’s honest outputs: 2/5 (static), φ (dynamical), and 2φ/5 (hypothesis).

The correction to hold: the cocycle is non-trivial but uniform. The notation-level labeling is defined only up to global exchange. This does not change the attractor, but it is a real feature of the construction and should be stated wherever the dynamical result is cited.

###

Leave a Reply