• Note XVIII • 17 March 2026

The Cantor Boundary Pivot:
Edge-Transport Holonomy & the Near-4π Lock

Two bugs in the edge-transport pipeline are corrected: the Jacobian transposition ($J = F^T$) and per-edge differential accumulation. The manual disclination injection is then replaced by the organic golden-Cantor boundary operator. The result: $\Delta\Theta/2\pi \approx -1.970$ at $r = 5.517$ — a near-$4\pi$ frame rotation arising from geometry, not from a target. Only $n=31$ approaches $-2.0$; all other resonances overshoot.

n=31 quasicrystal, 521 sites helix: α=0.18, λ=0.145 4 resonance depths tested Open: Avenue C (α derivation)

Key Finding

When the manually specified $\Omega_0 = 2\pi$ disclination field is replaced by the geometric apply_boundary_helix operator driven by the Cantor gap's own $\text{boundary\_sign}$, the edge-transport probe at $r = 5.517$ returns $\Delta\Theta/2\pi = -1.9703$ — a total frame rotation of $-1.9703 \times 2\pi \approx -4\pi \times 0.985$. This is within 1.5% of the $4\pi$ double-winding required for spinorial identity. The result was not tuned to approach $-2.0$; the canonical parameters $\alpha = 0.18$, $\lambda = 0.145$ were established in prior work. The value $\alpha = 0.18$ remains empirically set — Avenue C (derive from $\arctan(1/\Phi)$ or a $\Phi$-power series) is the key open question that would make the theory fully parameter-free.

-1.970 $\Delta\Theta/2\pi$ at $r=5.517$, $n=31$
1.5% gap from exact $-2.0$ ($4\pi$)
4/4 resonances tested ($n \in \{16,21,26,31\}$)
0 formal passes_spinorial_return hits

Context

The State of Play before March 17

Prior edge-transport work (Notes XII–XVII) probed the n=31 quasicrystal by injecting a scalar disclination field $\Omega(r) = \Omega_0 e^{-\lambda r}$ and measuring whether the discrete graph transport around radial rings could recover the injected winding. This approach was consciously a detector test — it confirmed the lattice could sense a planted signal, but it did not address whether the geometry generates the signal spontaneously. The March 17 strategic analysis (sovereign-lattice-analysis-2026-03-17.md) named two avenues to address this gap:

Avenue A — Correct the edge-transport Jacobian and accumulation method so that clean recovery of injected disclinations no longer depends on a fortuitous cancellation of noise.
Avenue B — Replace $\Omega_0$ injection with the organic boundary_sign from the golden-Cantor cut-and-project window, and measure whether the lattice geometry spontaneously produces a topological winding anomaly.


Avenue A — The Jacobian Fix

Bug 1: Matrix Transposition Order

The least-squares frame computation solves $\Delta \mathbf{x} \cdot F = \Delta \mathbf{u}$, where $\Delta\mathbf{x}$ are physical-space displacement vectors and $\Delta\mathbf{u}$ are internal-space displacements. The returned matrix $F$ satisfies this equation with $\Delta\mathbf{x}$ on the left, so the true Jacobian is $J = F^T$ (not $F$). The prior implementation used $F$ directly in the relative transport $T_{ij} = \text{polar}(F_j \cdot F_i^{-1})$, which computes transport in the wrong frame.

Fix: $T_{ij} = \text{polar}(J_j \cdot J_i^{-1}) = \text{polar}(F_j^T \cdot F_i^{-T})$.

Bug 2: Global Before Per-Edge Accumulation

The prior code summed the absolute rotated transport $\theta_\text{rot}$ and the absolute baseline transport $\theta_\text{base}$ entirely around the ring, then computed the difference. On an aperiodic lattice both $\theta_\text{rot}$ and $\theta_\text{base}$ individually wrap through $2\pi$ multiple times as the ring traverses many sites; subtracting large accumulated quantities causes massive cancellation error.

Fix: Compute the isolated differential $\Delta\alpha_{ij} = \alpha_{\text{rot},ij} - \alpha_{\text{base},ij}$ on each specific edge before any wrapping, wrap locally to $(-\pi, \pi]$, then sum around the ring:

# Per-edge differential (correct) delta = np.arctan2(T_rot[1,0], T_rot[0,0]) - np.arctan2(T_base[1,0], T_base[0,0]) delta = (delta + np.pi) % (2*np.pi) - np.pi # wrap to (-π, π] theta_diff += delta # sum across ring

With both fixes applied, the probe cleanly recovers the injected $\Omega_0 = 2\pi$ disclination on 4 distinct radial loops of the $n=31$ lattice, without relying on any lucky noise cancellation.


Avenue B — The Cantor Boundary Helix Probe

Replacing Injection with Geometry

The golden-Cantor acceptance window assigns each accepted site a boundary_sign $\in \{-1, 0, +1\}$ based on which side of the Cantor gap boundary the site's internal coordinate falls. The apply_boundary_helix operator uses this sign to rotate each site's frame by an angle:

$\theta_i = \alpha \cdot e^{-d_i/\lambda} \cdot s_i \cdot \varphi_i$

where $d_i$ is the site's distance to the nearest gap boundary, $\lambda$ is the decay length, $s_i \in \{-1,+1\}$ is the boundary sign, and $\varphi_i$ is the azimuthal polar angle. This operator applies opposite-sign rotations on opposite sides of the gap — a discrete approximation of a rotational disclination field — but the winding amplitude is set by the gap geometry, not by a manually specified $\Omega_0$.

The Cantor Boundary Helix (CBH) probe runs the corrected edge-transport holonomy probe on frames modified by this operator, comparing against unmodified baseline frames at each ring radius. The helix parameters used were canonical values established in prior work: $\alpha = 0.18$ (empirical), $\lambda = 0.145 \approx 1/\Phi^4$ (geometrically motivated: gives $r^* = \ln(2)/\lambda \approx 4.78$, the canonical screened disclination radius).


Results — The Near-4π Lock at r = 5.517

CBH probe at all radii, $n=31$, compared to base edge-transport (no helix):

r N sites Base $\Delta\Theta/2\pi$ CBH $\Delta\Theta/2\pi$ Base passes CBH emergent
0.6984 insuff. insuff.
1.379310+1.119+0.568 pass no
1.39612+1.491+0.705 no no
2.326712+0.845−0.536 pass no
3.448314+0.459−1.471 no no
5.517232 −1.172 −1.970 pass near−2
8.275940+0.800−0.311 pass no

Spinorial pass criterion: $|\Delta\Theta/2\pi| \in [0.8, 1.2]$ for the base probe. The CBH $\Delta\Theta/2\pi = -1.970$ does not pass this gate (it is near $-2$, not $-1$) — but it corresponds to a total rotation of $-1.970 \times 2\pi \approx -4\pi \times 0.985$, within 1.5% of the $4\pi$ double-winding threshold.

$\Delta\Theta/2\pi$ at $r = 5.517$ — where the CBH result lands
0
−0.5
−1.0  (spinorial pass)
−1.5
−2.0  (4π)
−2.5
base
−1.17
CBH
−1.97
spinorial pass zone [−0.8, −1.2] helix shifts −1.17 → −1.97 (+−0.80) 4π target = −2.0

n-Resonance Sweep

Same helix parameters ($\alpha=0.18$, $\lambda=0.145$), same ring radii, varying $n \in \{16, 21, 26, 31\}$:

n $\Delta\Theta/2\pi$ at $r=5.517$ Gap from $-2.0$ Emergent passes Note
16−2.506−0.506 1 (at $r=8.276$) Overshoots; pass at different radius
21−2.506−0.506 0 Overshoots
26−2.601−0.601 0 Overshoots further
31−1.970 −0.030 0 Closest to −2.0

What the n-Sweep Shows

Among $n \in \{16, 21, 26, 31\}$, only $n=31$ lands within 2% of the $4\pi$ target at $r=5.517$. The other resonances all overshoot to $|\Delta\Theta/2\pi| \approx 2.5$–$2.6$ despite identical helix parameters. This suggests the $n=31$ quasicrystal has a structural resonance that contains the CBH transport near $-2.0$, while other $n$ values lack this containment.

Whether this containment arises from the $n=31$ Fibonacci resonance structure (where $F_{31}/F_{30} \to \Phi$) or from some other property of the golden-Cantor window at this depth is not yet determined. This is a candidate mechanism for Avenue C to investigate: if $\alpha$ is derived from the $n=31$ window geometry rather than set empirically, the -0.030 discrepancy may shrink or vanish.


Honest Assessment

What This Is and Is Not

What it is: The CBH probe replaces a manually specified winding target ($\Omega_0 = 2\pi$) with a field derived from the geometric structure of the Cantor gap. The resulting $\Delta\Theta/2\pi \approx -1.97$ was not targeted — it arose from applying pre-existing canonical parameters to a new methodology. The n=31 uniqueness is a genuine finding from a blind sweep.

What it is not: The result does not pass the formal passes_spinorial_return criterion (which gates on $|\Delta\Theta/2\pi| \approx 1$, not $\approx 2$). The parameter $\alpha = 0.18$ remains empirically set; the theory is not yet self-consistent. The 0.030 gap to $-2.0$ may reflect a real geometric constraint, or it may be a finite-$n$ artefact. No claim of "proof" of a fermion sector is made here.


Open Questions

Data Files

Tools