Time is not a line. You already knew this — every year you are older than you were, and returning to a place you lived as a child is not the same as being there. The question is why. What is the mathematical structure underneath the feeling that the return is always different?
The first four operators of the chain — C, K, F, U — describe a single generative cycle: from compression through threshold, fold, and unfolding. One complete application of G = U ∘ F ∘ K ∘ C is one circle. But what happens after the circle closes? Where does the system go?
In a simple loop, the system returns to x₀. Back to the start. The clock resets. But in the dm³ framework on a contact manifold, the return carries a record — encoded in the action variable z — of everything that happened during the circuit. The source is enriched. The system returns to x₀′, not x₀. And x₀′ ≠ x₀.
This is why a practitioner who has completed 33 cycles is not the same as a beginner who has completed none. The operators are identical. The framework is the same. But the action variable z is different. The source has been enriched by 33 complete circuits, and the 34th cycle begins from a fundamentally different place.
The full chain is:
T does not replace the other operators. It is what happens when one complete G-cycle closes and the system asks: where do I begin again? The answer is not the same place. It is a new place, enriched by the accumulated thermodynamic cost of everything just completed. That cost is not lost. It is carried in z.
The g-series names the qualitative behavioral regimes of a dm³ system. The indices are motivated labels, not derived constants — ordinal markers for distinct stability thresholds.
| Label | Index | Regime | Criterion |
|---|---|---|---|
| g⁰ | 0 | Seed | Single operator application. No closure condition. The beginning. |
| g² | 2 | Compositional | Two-operator closure. Minimal self-reference. F is active for the first time. |
| g⁶ | 6 | Cycle-scale | Full G-cycle closure. The limit cycle Γ is entered. The system orbits. |
| g³³ | 33 | Soft equilibrium | Heuristic stability: ⌈log₂(3!)·4⌉×3 = 33. Three invariants achieve simultaneous robust closure. Conjectured, not proved — open verification target in AXLE (Lean 4). |
| g⁶⁴ | 64 | Phase threshold | Complete possibility-space coverage. 2⁶ = 64 operator-state combinations. x₆₄ = G⁶⁴(x₀). The full map is drawn. |
The computation: 3! = 6 orderings of three invariants that must close simultaneously. log₂(6) × 4 ≈ 10.34, so ⌈log₂(3!) × 4⌉ = 11. Multiply by 3 for triple-confirmation robustness: 3 × 11 = 33. This is heuristic, not derived from the contact geometry. The index 33 is a conjecture — stated precisely as an open problem for AXLE. Below 33 cycles the system is conjectured to be fragile. At or above 33, the organizing invariant is conjectured to persist across disruption.
The contact form α = dz − λ is the key. On a symplectic manifold (no z variable), the system could in principle return exactly to its starting point — energy is conserved, the phase space is closed. On a contact manifold, the action variable z accumulates dissipation along every orbit. The system on Γ = {ρ = 1} satisfies:
The z variable is not entropy in the thermodynamic sense — it is the action variable of the contact geometry, the record of all accumulated dissipation. But its behavior mirrors entropy: it never decreases along an orbit. Every circuit costs something. That cost is carried forward. This is why the return is a spiral and not a loop.
The dm³ Lyapunov function V(ρ) = ½(ρ − 1)² guarantees that small perturbations to the operator chain do not destroy the spiral return. The exact stability radius:
F1. Absence of spiral enrichment (x₀′ = x₀, no accumulated z) in dm³ contact simulations above the g³³ stability index would refute Theorem T1.
F2. Violation of dm³ structural stability (ε₀ = 1/3) under operator perturbations within the stated bounds would refute T1.
F3. A formal proof in Lean 4 (AXLE repository) that the Gronwall bound does not imply transverse contraction toward Γ would refute T1.
The contact-geometric structure of GTCT has structural analogies — not proved connections — to three results in quantum foundations. These are open research problems:
Two-State Vector Formalism (Aharonov & Vaidman, 2001): the dm³ operator chain G = U ∘ F ∘ K ∘ C has structural resemblance to the bidirectional picture — F introduces self-reference at mid-circuit, U projects outward. Whether a functor exists from the dm³ operator category to pre- and post-selected quantum states, preserving ε₀ = 1/3, is an open problem.
Delayed-Choice Quantum Eraser (Kim et al., 2000): the Fold operator F — which introduces the first self-reference — is structurally analogous to the beam-splitter choice: the point at which the system curves back onto its own prior state. No retrocausal claim is made.
Entanglement Swapping (Ma et al., 2012): photons that never coexisted show correlations. The spiral return x₀ → x₆₄ → x₀′ with x₀′ ≠ x₀ is formally analogous: the source state is enriched by one completed circuit without returning to its original configuration.
Navrátil Convergence (2026): Navrátil independently derives a geometric Hilbert space with inner product ⟨j|k⟩ = η⁻ᵏδ and geometric Born rule from the SL(3,ℤ) Tribonacci algebra, departing from a discrete algebraic lattice rather than a contact manifold. The convergence on η⁻ᵏ as a structural weight from independent starting points is an open structural question.
A student who completes all six has verified the theorem independently. Each exercise targets a different skill level.
Verify numerically that log₂(3!) × 4 = 10.3398… and therefore ⌈10.3398⌉ = 11. Then verify 3 × 11 = 33. Explain in your own words why multiplying by 3 rather than 1 is necessary for the triple-confirmation robustness criterion.
Prove the Gronwall inequality in differential form using the integrating factor μ(t) = exp(−∫₀ᵗ k(s) ds). Apply it to the dm³ transverse deviation ξ(t) with k = |μ_max| = 2 and show ε₀ = 1/3 is exact.
Solve the dm³ normal form equations numerically (Euler method, dt = 0.01) for (ρ₀, θ₀, z₀) = (1.2, 0, 0) with μ_max = −2, β = 1, ω = 1. Integrate for 10 periods. Plot ρ(t) and show convergence to Γ = {ρ = 1}. Compute λ_⊥ = exp(−4π) and compare to the numerical rate.
Show that the dm³ contact form α = dz − λ is non-degenerate: α ∧ (dα)ⁿ ≠ 0 everywhere. Explain why symplectic geometry (dλ only, no z variable) cannot support attractors on compact manifolds but contact geometry can.
Read Kim et al. (2000, PRL 84:1). Identify which of C, K, F, U corresponds to each stage of the delayed-choice quantum eraser experiment. Which operator is the beam-splitter choice? Which is the Fold that makes the future measurement retroactively constrain the past path? Write a 400-word analysis.
The g³³ stability index conjectures that spiral enrichment requires at least 33 operator cycles. Design an experimental protocol (quantum optics or NMR) to test this conjecture. Specify: (1) system and rationale; (2) operationalization of 33 cycles; (3) observable distinguishing spiral from linear return; (4) result that would falsify the conjecture. Submit as a 1-page research proposal on Zenodo.