Self-as-an-End
SAE Black Hole Series · Paper 4

Angular Transport in Discrete Black-Hole Interiors: Process Accounts, Spin-Order Lattices, and the Rigid-Rotation Spectrum
离散黑洞内区的角向输运:过程账、旋向格与刚转谱

Han Qin (秦汉) · Independent Researcher · 2026
DOI: 10.5281/zenodo.21569768 · Full PDF on Zenodo · CC BY 4.0
Abstract

Within the pure-occupancy class of discrete black-hole interior models, this paper establishes a process account for angular transport. Three accounts agree tick by tick; the bare occupancy state does not determine the instantaneous flow; the cumulative account is in general not exact along closed histories, and any state extension that integrates the unbounded real-valued account exactly must be infinite; equality of reachable groups does not imply equality of process cost. The single-tick spectrum of a saturated shell is rigidly quantized to $\{0,\pm C_j\}$; the amplitude of any legal transfer obeys a universal bound; closed-chain interface accounts are pinned to an integer subgroup whose algebraic lattice spacing is legislated by capacities, increments, and parent offsets, and a degenerate core, in the fully saturated closed-chain library with at least one non-degenerate shell, supplies the unit generator that collapses the global spacing to one. Observation proceeds through a four-layer compression whose zero fibers are nested, a non-vanishing coarse layer forcing non-vanishing at every finer upstream layer but not conversely; net flow is the shallowest visible layer, not the only one. Two negative results close the paper: a prograde clamp on tangential moves leaves the unique smooth core unchanged, and an annular configuration can be maintained indefinitely by scheduling while nothing in the model lets the angular account enter radial legality or the radial budget. This paper establishes a combinatorial angular-transport account only; the shell weights, time scale, and energy interface required to pass from this account to physical angular momentum are all left unestablished.

1 Introduction, Scope, and Dependency Matrix

This series studies discrete black-hole interiors within a pure-occupancy model class: earlier papers established the radial narrative of collapse [P1], the interior information and timing conventions [P2], and potential exclusion with stratified monotonicity [P3]. In these interiors the only degree of freedom is occupancy and the only dynamics is the legality of moves; the radial voice is in place, the angular voice is not. This paper supplies it. Within this pure-occupancy model class, angular net flow is first of all a transfer-level quantity; its cumulative real-valued account is in general not exact along closed histories. Finite state spaces can still carry orientation, phase classes, and discrete instantaneous-flow tiers; but a state extension that retains an unbounded rigid-rotation history while integrating the cumulative account exactly must be infinite. What this paper establishes is the definition of that account, the proof of its non-exactness, the rigidity of its single-tick spectrum and the conditions that break it, the layered structure of its observable compressions, and two negative results: the angular account does not alter the shape of the smooth core, and it does not supply radial support in this model.

Scope first. Every account in this paper is combinatorial: steps are counted by moves, layers by shells, and sums are taken with declared weights. Scope sentence: this paper establishes a combinatorial angular-transport account; the shell weights, time scale, and energy interface required to pass from this account to physical angular momentum are all left unestablished. Accordingly, four non-claims hold throughout. First, this paper does not claim that physical angular momentum fails to be a state quantity; what is proved is only the non-exactness of the combinatorial cumulative account within this model class. Second, this paper does not identify the unweighted edge-step account with Kerr angular momentum; between the two stand the three interfaces of shell weight, time, and energy. Third, this paper does not claim that all entropy accounts and release orders are blind to the angular sector; the blind zones established are named one by one and are finite in number. Fourth, this paper does not claim that after rigidity breaking the spectrum must fill the interval, and it never speaks of a continuous spectrum; every spectrum here is a discrete combinatorial one. Questions touching inner horizons, closed timelike curves, or negative-coordinate extensions lie outside the scope and are not adjudicated.

The paper delivers three positive results and one negative. Positive, first: the process account, comprising the agreement of three accounts, the underdetermination of instantaneous flow by occupancy, and the non-exactness of the cumulative account (Section 3). Second, spectral legislation: single-shell rigid quantization, the universal amplitude bound, the interface subgroup theorem with its degenerate shortcut, together the integer legislation of spectral lattice spacing (Sections 4 and 5). Third, observational compression: the four-layer observation maps and the nesting of their zero fibers (Section 6). Negative: invariance of the core shape and absence of radial support from annular configurations (Section 7). Section 8 corroborates the theorems with differential-rotation statistics under declared measures, Section 9 lists open problems and interface debts, and Section 10 concludes and restates the firewall.

Minimal dependency matrix. All principal results are conditional; their hypotheses are tabulated below and restated in full at each item; nothing is cited in place of writing. Tier notation: D marks definitions and model inputs; T1 marks conditional theorems with explicit hypotheses; T2 marks readings assembled from conditions; T3 marks speculative interface remarks that carry no load; C1 marks exact enumerations, with C1-R meaning independently recomputed and C1-P meaning pending independent rerun. All numerical values cited in this paper are C1-R.

Result Timing semantics Occupancy Library Weights Tier
Three-account agreement (3.1)synchronousarbitrarylegal transfersper-shell vectorT1
Non-exactness (3.3)synchronoussaturated corecontains a prograde rotationany positive weightsT1
Two budgets (3.5)botharbitraryper branchexplicit $\alpha$T1
Directed word length (3.6)synchronoussaturated core$\mathsf A_1$edge-stepT1
Rigid quantization (4.1)synchronoussaturatedoccupancy-identityper-shellT1
Amplitude bound (4.3)synchronousarbitrarylegal transfersedge-stepT1
Interface subgroup (5.2)synchronoussaturated coreoccupancy-identityedge-stepT1
Spectral lattice (5.4)synchronoussaturated corefull occupancy-identityedge-stepT1
Observation layers (6.1)synchronousarbitrarylegal transfersdeclared $\alpha$T1
Core and annulus (7.1 to 7.3)single-hop and synchronousnon-smooth statesper clausenoneT1
Differential-rotation statistics (8)synchronoussaturated coredeclared measures$\alpha=\mathbf 1$C1-R and T2

Guide: Section 2 fixes the model, timing, and conventions; Section 3 establishes the process account and its non-exactness, closing the distinction between process and state with the directed word length; Section 4 gives the single-shell spectrum and its sign symmetry; Section 5 the interface subgroup and rigidity breaking; Section 6 observational compression; Section 7 the negative results; Section 8 the statistics; Sections 9 and 10 close.

2 Model, Timing, and Conventions

Every item in this section is a definition or model input, tier D; only Lemma 2.4 is a truth-apt claim, tier T1.

Definition 2.1 (onion graph and occupancy). An onion graph $G$ consists of shells $j=0,\dots,R$, shell $j$ having capacity $C_j\ge1$ with cells $(j,i)$, $i\in\mathbb Z_{C_j}$. Tangential edges join ring neighbours within a shell: a cycle for $C_j\ge3$, a single edge for $C_j=2$, none for $C_j=1$. Radial edges join each cell $(j,i)$ with $j\ge1$ to its inward parents $\mathrm{par}(j,i)\subseteq\mathbb Z_{C_{j-1}}$, at least one. Occupancy is $n\in\{0,1\}$ per cell ($n_{\max}=1$); a configuration $x$ is an occupancy function with token number $N=\sum x$. Shells with $C_j\le2$ are called degenerate. The default parent construction, inherited from [P3], is primary-plus-successor; every result states its own hypotheses and none relies on the default.

Definition 2.2 (timing and legal transfers). The model carries two timing semantics. (a) Sequential single hop: one token moves along one incident edge; the move is legal iff the source cell is occupied and the target cell empty. (b) Synchronous composite: one tick applies a move set $\theta=\{a\to b\}$, legal iff four clauses hold: every move runs along an incident edge; each source emits at most one move; each target receives at most one move; every moved token starts from an occupied cell and the final occupancy of every cell remains in $\{0,1\}$. The move set of a legal synchronous transfer is a partial permutation on the graph, and its support decomposes into vertex-disjoint directed paths and closed chains; the occupancy-identity transfers are exactly the unions of closed chains (occupancy identity leaves a partial permutation no open path, so the support is purely closed chains). (c) Naming: a transfer with empty move set is the empty transfer; transfers that are occupancy-identity yet have nonempty move sets leave the state unchanged while the account may be nonzero, and their existence is the forerunner of the non-exactness of Section 3. Which semantics each result uses is recorded in the dependency matrix. This clause restates, self-containedly, the timing conventions of the earlier papers ([P3] Definition 3.0).

Definition 2.3 (sequential structure). A primary-parent selector picks one inward parent $p_j(i)\in\mathrm{par}(j,i)$ per cell. Interface $j$ is sequential if there is a non-decreasing lift $\tilde p_j:\mathbb Z\to\mathbb Z$ with $\tilde p_j(i+C_j)=\tilde p_j(i)+C_{j-1}$. Increments are $g_i=\tilde p_j(i{+}1)-\tilde p_j(i)\ge0$ with period sum $\sum_{i

Lemma 2.4 (transport of orientation). Under sequential interfaces, pick either ring direction on the innermost non-degenerate shell and recursively choose coordinates outward so that each $\tilde p_j$ is non-decreasing; the whole tail acquires an orientation, and reversing the initial choice reverses everything, so the orientation is unique up to a global flip. Proof: order preservation of the lift transports the direction interface by interface; the two initial choices are the two global flips, and no circularity of direction enters. $\blacksquare$ [T1; the hypothesis is the sequential structure itself.]

Definition 2.5 (orientation profile and projection). Fix the orientation profile $\omega$ by Lemma 2.4. The prograde successor map is defined only on shells with $C_j\ge3$; the orientation projection $\tau$ assigns $+1$ to prograde tangential moves, $-1$ to retrograde ones, and $0$ to radial moves and to tangential moves on degenerate shells.

Definition 2.6 (three quantities and initial values). Per transfer, the single-tick shell net flow is $\ell_j(\theta)=\sum\tau(m)$ over tangential moves in shell $j$; the cumulative account is $\Lambda_j(T)=\sum_{t\le T}\ell_j(\theta_t)$ with $\Lambda_j(0)=0$; the tracer unwound angular step account $w_{a,j}$ counts, per token, its signed tangential steps in shell $j$ in edge-step units, with $w_{a,j}(0)=0$; only on closed sojourn segments, where $w_{a,j}/C_j\in\mathbb Z$, is it called a winding number. Physical angular momentum $J_{\rm phys}$ is not a quantity of this paper; it belongs to the interface.

Definition 2.7 (vector account and explicit weights). The per-shell vector account $\boldsymbol\ell=(\ell_j)$ is the primary object; scalar totals exist only with explicit positive weights, $\ell_\alpha=\sum_j\alpha_j\ell_j$, $\alpha_j>0$. Three gauges: the edge-step gauge $\alpha=\mathbf 1$, combinatorially privileged since every move counts an integer; the turn gauge $\alpha_j=1/C_j$; the physical gauge $\alpha^{\rm phys}$, unestablished, deferred to the interface. A shared orientation grants only sign comparability across shells; magnitude comparison requires a declared weight.

Definition 2.8 (spin-order lattices). Four names: $\mathsf A_1$, prograde tangential moves only, radial and retrograde forbidden; $\mathsf A_2$, every atomic move has $\tau(m)\ge0$; $\mathsf A_3^\alpha$, every transfer has $\ell_\alpha(\theta)\ge0$; $\mathsf A_0^\alpha$, every transfer has $\ell_\alpha(\theta)=0$. $\mathsf A_0^\alpha$ is the zero fiber of $\mathsf A_3^\alpha$, listed as the neutral sector, not an independent fourth clamp.

Definition 2.9 (four observation layers). Four deterministic compression maps: the gross move record $\mathcal O_{\rm gross}(\theta)=M(\theta)$; the tangential net edge-current field $J^\omega_\theta$, defined on tangential edges of non-degenerate shells with the positive direction from the orientation profile, $J^\omega_\theta(e)=\#(e_+)-\#(e_-)$; the shell net flow $\ell_j(\theta)=\sum_{e\subset\text{shell }j}J^\omega_\theta(e)$; the weighted total $\ell_\alpha(\theta)$. The nesting and strictness of the zero fibers are the subject of Section 6; the kernel semantics is inherited from [P3] Section 6.

Definition 2.10 (generalized defect and the sequential trichotomy). On interface $j$, any inward parent $q$ of cell $i$ carries the offset $\delta_q(i)=(q-p_j(i))\bmod C_{j-1}$ and the generalized defect $D(i,q)=i-\tilde p_j(i)-\delta_q(i)$; at the primary parent $\delta=0$. Sequential trichotomy: if the primary map is non-constant the lift is unique and the increment vector well defined; if constant, a sequential lift exists but is not unique, the increment formula does not apply, and all cells share one parent; if the period sum is a proper multiple of $C_{j-1}$ the map is non-sequential (crossing).

3 Process Accounts, Non-Exactness, and the Directed Word Length

This section defines the account on legal transfers, proves that occupancy does not determine the instantaneous flow, proves that the cumulative account is not exact, and closes the distinction between process and state quantities with the directed word length. Hypotheses are restated at each item.

Proposition 3.1 (three accounts agree). Hypotheses: synchronous composite legality (Definition 2.2 b); the tracer rule that each move is charged to the token it carries; the orientation projection vanishes on both sides of degenerate shells (Definition 2.5). Then per tick, $$\sum_{e\subset\text{shell }j}J^\omega_{\theta_t}(e)=\ell_j(\theta_t)=\sum_a\bigl(w_{a,j}(t)-w_{a,j}(t-1)\bigr)\qquad(\forall j),$$ that is, the net edge-current account, the shell account, and the summed tracer unwound accounts take the same value each tick; the first equality is by Definition 2.9, the second is this proposition. Proof. Each tangential move carries exactly one token, and its $\tau$ enters simultaneously the net current of its edge, the net flow of shell $j$, and the unwound account of its token; the three sums are one finite sum in different orders, and exchanging the order gives the claim. Radial moves contribute zero to all three accounts, so tokens crossing shells create no boundary terms. $\blacksquare$ [T1]

Lemma 3.2 (bare occupancy does not determine flow). There exist a configuration $x$ and two legal transfers $\theta,\theta'$ from $x$ with $\boldsymbol\ell(\theta)\neq\boldsymbol\ell(\theta')$. Proof. On a non-degenerate saturated shell, the prograde and retrograde full-shell rotations are both legal occupancy-identity transfers (closed chains, Definition 2.2), with accounts $+C_j$ and $-C_j$. Hence the single-tick flow is not a function of the occupancy state. $\blacksquare$ [T1]

Theorem 3.3 (non-exactness). Hypotheses: the library is the occupancy-identity class of saturated transfers (Definition 2.2 c), containing at least one prograde rigid rotation of a non-degenerate saturated shell available to scheduling; the weight $\alpha$ is any positive weight. (i) Occupancy layer: no real-valued function $\Psi$ of occupancy states satisfies $\ell_\alpha(\theta)=\Psi(x_{\rm fin})-\Psi(x_{\rm init})$ for all library transfers, since the library contains occupancy-identity transfers with account $\alpha_jC_j\neq0$ (rotations), for which the right side is identically zero. (ii) Labelled layer: no such function of labelled states exists either, since $C_j$ prograde rotations form a closed history restoring all labels while accumulating $C_j\cdot\alpha_jC_j\neq0$, and any labelled-state function has zero net change along a closed history. (iii) Finite extensions: any finite state extension $E$ admitting an infinite lift of the rotation history has no exact primitive, since the lifted orbit inside finite $E$ is eventually periodic, one period accumulates a positive multiple of $\alpha_jC_j$, and any function on $E$ changes by zero over a period. $\blacksquare$ [T1]

Corollary 3.4 (sharpening). Finite states can carry orientation, phase classes modulo $C_j$, and discrete flow tiers; what the theorem excludes is only the global existence of a real-valued exact primitive; any extension that integrates the unbounded real-valued cumulative account exactly must be infinite. [T1]

Proposition 3.5 (two budgets). Radial side: the budget of inward atomic moves is finite and equals the potential drop $\Delta\Phi_r$, inherited from the ledger identity, Corollary 3.1a of [P3], together with its potential exclusion theorems; not reproved here. Tangential side, each branch conditional on the existence of a repeatable transfer: if the unclamped occupancy-identity library contains both rotations of some non-degenerate saturated shell, the cumulative account can run to $\pm\infty$; if the $\mathsf A_3^\alpha$ library contains a repeatable $\theta_+$ with $\ell_\alpha(\theta_+)>0$, it runs one-sidedly unbounded from zero; if the $\mathsf A_0^\alpha$ library contains a repeatable $v\in\mathrm{Spec}(\boldsymbol\ell)\setminus\{0\}$ with $\alpha\cdot v=0$, per-shell accounts oscillate unboundedly; absent such a vector no conclusion is drawn on that branch. Main sentence: radial descent has a finite countdown; the angular cumulative account has no countdown supplied by $\Phi_r$. [T1; the reading "the hole can rotate forever but cannot fall forever" is an assembled conditional reading, tier T2, valid where the branch hypotheses hold.]

Proposition 3.6 (directed word length; same group, different account). Hypotheses: saturated core; the $\mathsf A_1$ library, whose nonempty transfers are exactly disjoint unions of prograde rigid rotations of non-degenerate shells. Writing $\mathbf k=(k_j)\in\prod_j\mathbb Z_{C_j}$, $0\le k_jEqual reachable groups do not imply equal process cost: the distinction between process and state quantities closes here. [T1]

4 Rigid-Rotation Spectrum, Amplitude Bound, and Sign Symmetry

Theorem 4.1 (single-shell rigid quantization). Hypotheses: shell $j$ non-degenerate ($C_j\ge3$) and saturated; the transfer $\theta$ is a legal synchronous transfer whose moves lie entirely on the tangential edges of shell $j$ and which is occupancy-identity. Then every nonempty connected component of the move graph of $\theta$ is a directed cycle on the ring, and on a chordless ring the only directed cycles are the two-cycle on one edge (adjacent swap), the full prograde rotation, and the full retrograde rotation. Hence $$\ell_j(\theta)\in\{0,\pm C_j\},$$ with $\pm C_j$ attained iff $\theta$ contains the corresponding full-shell rigid rotation. Proof. Saturation and occupancy identity make the restriction a permutation of the ring cells; the partial-permutation clauses decompose it into directed cycles; on a chordless ring a cycle's support is one edge or the whole ring. $\blacksquare$ [T1]

Proposition 4.2 (single-tick spectrum of the radial lattice). Hypotheses: all shells saturated; the library is the disjoint union of intra-shell tangential occupancy-identity transfers (no radial moves). The vector spectrum is the product $\prod_{C_j\ge3}\{0,\pm C_j\}$ with degenerate components identically zero, and the weighted total spectrum is the signed subset-sum set $\{\sum_j\varepsilon_j\alpha_jC_j:\varepsilon_j\in\{-1,0,1\}\}$. Proof: shells do not interact; apply Theorem 4.1 shellwise. $\blacksquare$ Note: holes in the total spectrum depend on $\{C_j\}$ and $\alpha$; in the edge-step gauge the two-shell tower with capacities $(3,4)$ misses $\pm2,\pm5,\pm6$; the exact enumeration (C1-R) is in Appendix C. Spectra with holes are combinatorial spectra, not energy spectra. [T1]

Proposition 4.3 (universal amplitude bound and equality). Hypotheses: any legal synchronous transfer, arbitrary occupancy, open chains with vacancies included. In the edge-step gauge, $$|\ell_{\mathbf 1}(\theta)|\le\sum_{C_j\ge3}C_j=:M.$$ Equality holds iff the restriction of $\theta$ to every non-degenerate shell is a full-shell rigid rotation of one and the same direction; the degenerate sector may carry disjoint zero-account activity besides. Proof. A partial permutation lets each cell emit at most one move; only tangential moves on non-degenerate shells are charged, one unit each, so the bound is the count of non-degenerate cells. Equality forces every such cell to emit a same-sign tangential move; under the occupancy constraint this means, shell by shell, "every cell moves to its prograde (retrograde) neighbour", i.e. a full rigid rotation, and conversely. Degenerate moves are uncharged and do not obstruct equality. $\blacksquare$ [T1]

Definition 4.4 (sign involution). On occupancy-identity transfers (unions of closed chains, Definition 2.2 c), set $\rho(\theta)=\{b\to a:a\to b\in\theta\}$, reversing every move. Then $\rho(\theta)$ is again a legal occupancy-identity transfer and $\boldsymbol\ell(\rho\theta)=-\boldsymbol\ell(\theta)$. [D]

Lemma 4.5 (involution count and fixed points). On the occupancy-identity library $\rho$ is an involution, so the enumeration satisfies $$N_{\rm all}=N_0+2N_+,$$ with $N_0$ the number of zero-account transfers and $N_+$ the positive ones, positives and negatives paired. Fixed points: $\mathrm{Fix}(\rho)$ is exactly the unions of swaps; in the radial-move-free library $\mathrm{Fix}(\rho)=\{\boldsymbol\ell=\mathbf 0\}$ by Theorem 4.1; in general only $\mathrm{Fix}(\rho)\subseteq\{\boldsymbol\ell=\mathbf 0\}$, strict iff the library contains a compensating transfer with $\rho(\theta)\neq\theta$ and $\boldsymbol\ell(\theta)=\mathbf 0$. Proof: involution and sign reversal are immediate movewise; the count is orbit length one or two; fixed points are movewise self-reverse, i.e. swaps. $\blacksquare$ Note: for the onion of capacities $(1,3,4)$, the full library gives $1995=521+2\times737$ and the radial-free library $53=27+2\times13$; exact enumeration (C1-R) in Appendix C. [T1]

Lemma 4.6 (amplitude bound for a single shell with vacancies). Hypotheses: shell $j$ non-degenerate with $v\ge1$ vacancies; a legal synchronous transfer whose moves lie entirely on the tangential edges of shell $j$. Then $|\ell_j(\theta)|\le C_j-v$, with equality attained by the synchronous shift of all $C_j-v$ tokens. Proof. Each source emits at most one move and must be occupied, so the tangential step count is at most the token number $C_j-v$, each step $\pm1$, giving the bound; for equality let every token move to its prograde neighbour: each target receives at most one move and the final occupancy is the shifted configuration, still in $\{0,1\}$, hence legal, with account $C_j-v$. $\blacksquare$ [T1]

Three end-of-section clarifications. First, the quantization of Theorem 4.1 rests on saturation, shell-confined moves, and the chordless ring; remove any one and it can fail, and how it fails is the subject of the next section. Second, total spectra depend on the weight: the edge-step and turn gauges shape the same library differently, and magnitude comparisons must declare their gauge. Third, all spectra here are discrete combinatorial spectra; any bridge from combinatorial to physical spectra belongs to the interface and is not built here.

5 Interface Subgroups, Degenerate Shortcuts, and Rigidity Breaking

Closed-chain interface accounts are pinned to an integer subgroup, and its algebraic lattice spacing is legislated by capacities, increments, and parent offsets. This is the section's headline and the paper's second principal result. With the single-shell spectrum rigid (Theorem 4.1), once cross-shell moves enter the library the question is not how wide the library opens but which numbers the interface generates: on which lattice the new account values fall and with what spacing. The section first exhibits three families of primitive cycles as generators, then proves the local subgroup theorem and the global spacing, and closes with the spectral-lattice corollary and the ladder of spacings.

Lemma 5.1 (construction of primitive cycles). Setting: interface $j$ with the notation of Definitions 2.3 and 2.10; the library is the legal occupancy-identity class (Definition 2.2 c); all shells saturated. (i) Shared-parent three-cycle: if ring neighbours $i,i{+}1$ share an inward parent $q$ (through any parent), the closed chain $(j,i)\to(j,i{+}1)\to(j{-}1,q)\to(j,i)$ is legal with account $+1$; reversed, $-1$. (ii) Increment cycle: if the interface is sequential with non-constant primary map and both shells non-degenerate, then for each column $i$ the closed chain $\gamma_i$, descending from column $i{+}1$ to $p(i)+g_i$, walking $g_i$ retrograde steps below to $p(i)$, ascending to column $i$, and taking one prograde step above, has $g_i+3$ edges and account $1-g_i$; $g_i=0$ recovers the primary form of (i), and $g_i=2$ gives a five-edge cycle of account $-1$. (iii) Secondary-parent cycle: if cell $i$ has a secondary parent $q$ with offset $\delta=\delta_q(i)\ge1$, the chain descending from $i$ through $q$, walking $\delta$ retrograde steps below to $p(i)$, and ascending back to $i$ has account $-\delta$; $\delta=1$ gives a three-cycle. (iv) Degenerate clause: if the lower shell has $C_{j-1}\le2$ and the upper is non-degenerate, the orientation projection vanishes below and the constructed cycle has account $+1$ regardless of $g_i$. Proof: each construction is a simple cycle, hence under saturation a permutation of its cell set, satisfying the synchronous clauses; accounts are read off from the projection stepwise. $\blacksquare$ [T1]

Theorem 5.2 (interface subgroup theorem). Hypotheses: both shells non-degenerate; primary map sequential and non-constant; $\theta$ a legal occupancy-identity transfer supported on the shell pair $\{j{-}1,j\}$. Then its edge-step account lies in $$G_j\mathbb Z,\qquad G_j=\gcd\bigl(C_j,\;C_{j-1},\;\{|1-g_i|\}_i,\;\{\delta_q(i)\}_{i,q}\bigr),$$ and $G_j$ is exactly the greatest common divisor of the primitive simple-cycle accounts of the pair. Proof. Let $d$ divide all listed generators. The generalized defect satisfies the two congruences $$D(i{+}1,p(i{+}1))-D(i,p(i))=1-g_i\equiv0,\qquad D(i,q)-D(i,p(i))=-\delta_q(i)\equiv0\pmod d,$$ and its period drift $C_j-C_{j-1}\equiv0$, so all radial endpoints share a residue $c_j$ of $D$ modulo $d$. Define the $d$-valued potential $\psi(j{-}1,x)=x$ and $\psi(j,i)=i-c_j$, both modulo $d$, columns taken in the shared orientation of Definition 2.5. Then along each move: a tangential step changes $\psi$ by exactly its $\tau=\pm1$, the wrap being harmless since $d\mid C$; a radial step is uncharged and changes $\psi$ by $q-(i-c_j)\equiv c_j-D(i,q)\equiv0\pmod d$. The transfer decomposes into vertex-disjoint cycles (Definition 2.2), each of which returns $\psi$ to its start, so the total account is $\equiv0\pmod d$. Generation: $C_j$ and $C_{j-1}$ are realized by full rotations, $|1-g_i|$ by the increment cycles, $\delta_q(i)$ by the secondary cycles, all primitive simple cycles; hence the cycle-account gcd divides the generators and is no coarser than $d$; with the containment this gives equality. $\blacksquare$ [T1]

Corollary 5.3 (global spacing and degenerate shortcuts). Tail clause: if $\theta$ is supported on the non-degenerate tail and every interface satisfies the hypotheses of Theorem 5.2, set $d=\gcd_jG_j$, take the per-interface residues $c_j$, define recursively $s_j-s_{j-1}=c_j\pmod d$ with $s_{j_*}=0$ at the innermost tail shell, and put $\psi(j,i)=i-s_j$ modulo $d$; every radial edge changes $\psi$ by zero, every tangential edge by $\tau$, and every closed chain has account $\equiv0\pmod d$. Global clause: any occupancy-identity legal transfer on the whole graph has account in $G^{\deg}\mathbb Z$ with $$G^{\deg}=\gcd\bigl(d,\;\{1:\text{interfaces with degenerate lower and non-degenerate upper shell}\}\bigr),$$ degenerate interfaces contributing the generator $1$ through Lemma 5.1 (iv), so the containment is trivial there while the generation is real; without degenerate interfaces the tail clause applies. The degenerate core is a unit-generating shortcut: in the fully saturated occupancy-identity closed-chain library with at least one non-degenerate shell, any onion with a one- or two-cell inner shell has global algebraic lattice spacing one, and account quantization survives only on tails away from degenerate hubs. In-class witness: capacities $(2,12,4)$ give tail spacing two, measured tail cycle gcd two, yet a legal shortcut three-cycle of account $+1$ and a legal odd-account transfer $+5$; exact enumeration (C1-R) in Appendix D. $\blacksquare$ [T1]

Corollary 5.4 (spectral lattice). Hypotheses in full: fully saturated core; occupancy identity; synchronous partial permutation; the complete occupancy-identity library; every non-degenerate interface satisfying the hypotheses of Theorem 5.2. Then $$\mathrm{Spec}(\ell_{\mathbf 1})\subseteq G^{\deg}\mathbb Z\cap[-M,M].$$ Registered falsity outside the domain: a single same-shell hop into a vacancy has account $\pm1$ and escapes the lattice already on a tower of global spacing four; the lattice statement holds only for the closed-chain library and not for spectra with vacancies. Proof: Corollary 5.3 with Proposition 4.3. $\blacksquare$ [T1]

Ladder of spacings and the rigidity-breaking criterion. Theorem 5.2 splits the landscape in three. Spacing $G_j=C$ (equal capacities, single parent, increments all one): every interface closed-chain account is winding number times $C$, and the rigidity of Theorem 4.1 survives the cross-shell library. Spacing $1unit spacing implies neither a single-tick $\pm1$ nor a hole-free spectrum: there is a single-parent interface with $G=1$ whose primitive cycle accounts contain no $\pm1$ and whose full spectrum is riddled with holes (C1-R, Appendix D); the bridge from subgroup to packing is an open problem, deferred to Section 9.

Table 5.1 (four interface examples). Proofs and exact enumerations in Appendix D.

Interface structure Generator profile $G_j$ Spectral clamp
Equal-capacity single-parent uniform towerincrements all one, no offsets$C$spectrum $\subseteq C\mathbb Z$, rigidity survives
One-cell centredegenerate shortcut$1$all cells share the centre, unit tier
Congruent single-parent towerincrements in one residue classintermediatespectrum $\subseteq d\mathbb Z$, even-lattice witness on record
Default double parentsoffsets all one$1$breaking everywhere, confluent with the shared-parent language

6 Observational Compression and Zero Fibers

This short section establishes one thing: angular information degrades along a four-layer deterministic compression, the shape of the loss being exactly a nesting of zero fibers; net flow is the shallowest visible layer, not the only one. The four maps are those of Definition 2.9.

Proposition 6.1 (nesting and strictness of zero fibers). Hypotheses: legal synchronous transfers, arbitrary occupancy, any positive weight. Universal containments $$\ker\mathcal O_{\rm gross}\subseteq\ker J^\omega\subseteq\ker\boldsymbol\ell\subseteq\ker\ell_\alpha,$$ where $\ker$ denotes the class of transfers on which the layer vanishes and $\ker\mathcal O_{\rm gross}$ is the empty transfer. Each strictness is conditional on a witness. First, gross-visible with vanishing edge current: a same-edge adjacent swap (the two directions cancel on one edge) and a purely radial swap are both nonempty with $J^\omega\equiv0$. Second, current-visible with vanishing shell account: on one non-degenerate shell, two disjoint edges carrying one prograde and one retrograde legal hop in the same tick, e.g. capacity five, tokens at columns zero and three, vacancies at one, two, four, moves $0\to1$ and $3\to2$; the two edge currents are $+1$ and $-1$ while the shell account is zero. Third, shell-visible with vanishing total: neutral compensation, conditional on library and weight; in the turn gauge any counter-rotation of two non-degenerate shells, in the edge-step gauge only equal-capacity pairs; libraries without such a vector make this containment non-strict. Proof: containments hold since each layer is a function of the previous; the witnesses are computed layer by layer. $\blacksquare$ [T1]

Proposition 6.2 (three named blind zones). The invisibility of the angular account is confined to the following three observables, each named and finite. First, the occupancy projection is blind to the account: on the saturated core, the all-prograde and all-retrograde rotation schedules have identical occupancy trajectories tick by tick (the saturated constant) while their cumulative accounts diverge with opposite signs; enumerated witness (C1-R) in Appendix C. Second, the occupancy-quotient entropy is blind: the entropy count of [P3] is defined on the occupancy quotient and is invariant under every occupancy-identity transfer. These two follow because their domains are functions of occupancy or its quotient, while occupancy-identity transfers with nonzero account are in the library (Definition 2.2 c, Theorem 3.3). Third, the cross-shell release statistics of whole-shell synchronized ablation are blind to the account, by a separate argument: this statistic is not a pure occupancy function, reading each label's birth shell and release round; yet purely tangential scheduling never changes any label's shell membership, and whole-shell synchronized ablation releases strictly in shell order, so the cross-shell inversion statistics are invariant under the insertion of any tangential schedule; the rotation-blind inversion witness (C1-R) is in Appendix C. $\blacksquare$ [T1]

In sum: a non-vanishing coarse layer forces non-vanishing at every finer upstream layer; a non-vanishing fine layer may vanish under further compression. Visibility of the full move record does not guarantee visibility of the edge current; edge-current visibility does not guarantee shell-account visibility; shell-account visibility does not guarantee weighted-total visibility. That net flow is the shallowest visible layer, and not the only one, means precisely this: whatever leaves a trace at the net-flow layer leaves traces at all upstream layers, while upstream traces need not reach the net flow. This is also the positive reading of the third non-claim of Section 1: the paper does not call all observation blind to the angular sector; the blindness established is the three names of Proposition 6.2, and the upstream edge-current layer sees more.

7 The Angular Account Neither Alters the Core Shape Nor Supplies Radial Support

This section establishes the paper's negative results. Hypotheses are restated per item; both timing semantics appear, each declared.

Lemma 7.1 (mutual reachability under one-way sliding). Hypotheses: a non-degenerate shell $C\ge3$ with fixed occupancy $0

Theorem 7.2 (invariance of the core shape). Hypotheses: fixed token number $N$; sequential single-hop semantics; the free hop library clamped to prograde ($\mathsf A_2$ at the single-hop level, where $\mathsf A_3$ coincides with it since a single move's net is its step). Then the unique smooth sink strongly connected class established for fixed $N$ in [P3], Proposition 3.2, keeps all three of its properties. First, every non-smooth state still has a finite path strictly decreasing $\Phi_r$, since the legality of an inward drop depends on occupancy only, never on orientation. Second, the smooth sector remains closed. Third, the skin shell remains strongly connected under prograde hops, by Lemma 7.1; for shells with $C\le2$ the lemma does not apply and one cites instead the original vacancy-pushing argument in the proof of [P3] Proposition 3.2, valid for all $C\ge2$. Hence a tangential spin-order clamp does not change the shape of the core: existence, uniqueness, and closedness of the smooth sink are untouched. $\blacksquare$ [T1]

Proposition 7.3 (annular configurations supply no radial support). Call $x$ annular if $N>0$, shells $0$ through $h{-}1$ are empty ($h\ge1$), and all occupancy lies in shells $h$ through $H$. Hypotheses: synchronous semantics; the unclamped legal library; the annulus contains at least one legal tangential move (a rigid rotation of a saturated shell $C\ge3$, or a single hop next to a vacancy on an unsaturated shell). Three clauses.

First, channel and budget: the inward channel is always open in the sense that the lowest occupied shell has parents in empty shells; the number of inward atomic moves needed to reach the smooth sector is exactly $\Delta\Phi_r$ (the ledger identity, [P3] Corollary 3.1a) and is unchanged by any tangential interleaving, while the number of ticks can be stretched arbitrarily. What is invariant is the step count, not the tick count.

Second, persistence and account: every annulus with a legal tangential move admits an infinite purely tangential schedule that never schedules a radial step, so the cavity persists by scheduling refusal alone. The zero-account branch is an additional condition: if the library also contains a repeatable nontrivial zero-account transfer (a saturated-shell adjacent swap, a swap of two occupied neighbours, or any other nonempty element of $\ker\boldsymbol\ell$), the same cavity persists under tick-by-tick zero account. This is not universal; the precise condition is that either some other shell carries a repeatable nonzero counter-signed tangential step, or the shell itself already has a zero-account transfer; the registered counterexample is an annulus consisting of this one shell alone, one token and two vacancies at $C=3$, where every nonempty tangential step has $|\ell|=1$, no swap is available, and no tick-by-tick zero-account schedule exists. The account value is therefore no switch on the radial channel; rotation is purely concurrent.

Third, transience: every nonempty annulus, having an empty inner layer, is non-smooth, hence transient in the dynamics of Theorem 7.2 and outside the closed recurrent class.

Joint statement: the shape can be maintained indefinitely by scheduling, while nothing in the model lets the angular account enter radial legality or the radial budget; the lifetime of an annulus is the restraint of a scheduler or the fiat of a clamp, never a work of spin. The name "no centrifugal term" means exactly this sentence and makes no universal physical negation. $\blacksquare$ [T1]

8 Differential-Rotation Statistics Under Conditional Measures

This section corroborates the theorems by sampling statistics under declared measures. Every conclusion is relative to its measure, no measure is natural, and nothing here is claimed as a law; all values are C1-R (Section 1), with full protocols, seeds, exact moments, and complete tables in Appendix E and recomputation material in Appendix F.

Measures and protocols. The component universe comprises both rigid rotations of every shell, all swaps, and all simple cycles of length at most six taken in both orientations, quotiented by support and account vector, each element passing the legality clauses of Definition 2.2. The three families are the default-double-parent onion of capacities $(1,3,4,5,6)$, the congruent single-parent tower $(12,4)$, and the equal-capacity single-parent uniform tower $(4,4,4)$, with coordinates given by their global algebraic lattice spacing (Corollary 5.3). The free measure is uniform on the universe with independent ticks; the clamped runs use $\mathsf A_2$ per Definition 2.8, movewise; the compensation runs draw compensating pairs at declared mixing rates; the neutral runs use half zero-account single components and half disjoint zero-sum pairs; the skin run is a random-cell sequential sweep on a single ring, a token possibly moving twice per tick. Wherever exact moments exist, the exact value from direct summation over the universe is canonical and the empirical value is a check.

Table 8.1 (summary of differential-rotation statistics; pure combinatorial-measure sampling, not time evolution).

Object Setting Highlights
Normalized shell drifts, equal or not$\mathsf A_2$ uniform component measure, three familiescongruent tower equal ($0.164$ vs $0.166$); uniform tower mid-shell double ($0.200,0.372,0.197$); default family fast inside, slow outside ($0.071,0.152,0.116,0.036$)
Sign of neighbour-shell covariancefree uniform component measurecongruent tower strongly anticorrelated $-0.750$; uniform tower strongly correlated $+0.351$; default family mixed ($-0.004,+0.023,-0.067$)
Two forms of compensation$\mathsf A_3$ mixture, rates $0,\tfrac14,\tfrac12$intra-component base rate $0.440$; pairwise compensation rising to $0.968$; total account always ascending
Neutral internal flow strengthzero-account mixturemean $2.608$ (interval $2.582$ to $2.635$), total account identically zero
Per-tick increment variance ratesamerising outward, $0.661$ to $1.259$
Mobility-curve peaksingle ring $C=12$, random-cell sweeppeak at half filling ($v=6$, mobility $0.356$), both ends declining (Appendix E)

Four readings, each prefixed by its measure, all tier T2.

First. Under the length-six, support-and-account-quotiented, uniform component measure with $\alpha=\mathbf 1$, the sign of the neighbour-shell covariance is a scale weighing the truncated cycle census: compensating cycles dominating gives anticorrelation, same-sign long cycles dominating gives correlation, and the two in balance give mixed signs; the three families of Table 8.1 are its three readings. No claim of a natural coupling constant for general graph families is made.

Second. Under the same measure clamped by $\mathsf A_2$ of Definition 2.8, when the only positive-account components are rigid rotations, the normalized shell drifts are equal; once all-prograde mixed cycles enter the library, the profile is generically differential. The equality comes from the combinatorial symmetry of library and sampling measure, not from any dynamical equation introduced here.

Third. Under the declared $\mathsf A_3$ mixture at $\alpha=\mathbf 1$, compensation splits into two forms: intra-component (a nonnegative single component with negative shell entries) and pairwise (disjoint positive and negative components in one tick); local retrogression is financed by global gain, and the clamp forbids total retreat, never differential rotation. This is the statistical quantification of the one-sided branch of Proposition 3.5; changing the weight changes the sufficiency criterion.

Fourth. Under the half-zero-single, half-zero-pair measure, a static total does not imply static shells: the total account is identically zero every tick while the internal flow strength exceeds two per tick and per-shell accounts diffuse linearly. This quantifies the neutral branch of Proposition 3.5 and does not extrapolate to all $\mathsf A_0$ schedules.

9 Open Mathematical Problems and Physical Interface Debts

Problem 9.1 (packing). A natural conjecture says that after rigidity breaking the occupancy-identity spectrum must fill the interval $[-M,M]$. The conjecture is false: unit lattice spacing does not imply a hole-free spectrum, the boundary marker of Section 5 being on record, and the claim is withdrawn. The defensible form narrows to: for default-double-parent growing onions and for single-parent floor growing onions, the occupancy-identity spectrum is the full segment $[-M,M]$. Both families have global spacing one: the former by the secondary-parent offsets (Table 5.1); the latter likewise, since floor growth makes capacities strictly increase across interfaces, the increments of the non-decreasing lift sum to $C_{j-1}$ over $C_j$ terms, forcing a zero increment, hence a shared-parent pair and a $\pm1$ cycle (Lemma 5.1 (i)). Exhaustion of three families and constructive certificates for one large family support the conjecture (C1-R, Appendix D); a general packing proof is not established. Tier: Open, no load carried.

Problem 9.2 (complete characterization of primitive $\pm1$ cycles). The four constructions of Lemma 5.1 are sufficient, the subgroup of Theorem 5.2 is the obstruction; the intermediate band is real: there is a single-parent interface with spacing one whose primitive cycle accounts contain no $\pm1$ (Appendix D). The exact characterization of when a primitive $\pm1$ cycle exists is open.

Problem 9.3 (non-sequential account theory). Crossing primary maps admit no lift and the increment language fails; small examples show rich cycle accounts. An account theory over non-sequential parent structures is open.

The angular voice of the three eras (T2). The three eras of the earlier papers were told radially; their angular voice can be assembled sentence by sentence from the theorems above, each sentence citing its part. Infall era: the tangential bias of falling matter writes the account tick by tick while radial steps are uncharged, the two ledgers running concurrently (Definition 2.5, Proposition 3.1). The saturation tick: once a shell fills, single-hop semantics dies and closed chains become the only carrier (Definition 2.2); a single shell with $v\ge1$ vacancies has single-tick account bounded by $\pm(C-v)$ (Lemma 4.6); at the saturation tick the belly of the spectrum dies as its extremes are born, the interior values all extinguished and $\pm C$ first reachable (Theorem 4.1). Tracer shell coordinates freeze from then on (the label radial lock, [P3] Lemma 3.3), and the occupancy history decouples completely from the account history (Proposition 6.2, first zone). Frozen-cycling era: rigid rotation can be kept resident by scheduling, the single-shell spectrum is $\{0,\pm C_j\}$, and cross-shell spacing is legislated by the interfaces (Theorems 4.1 and 5.2). Evaporation era: the cross-shell release statistics of whole-shell synchronized ablation inherit the order of [P3] and are blind to the purely tangential account (Proposition 6.2, third zone); the single-ring mobility curve shows the fully occupied end and the depleted end as the two low-activity wings of one combinatorial mobility curve (Table 8.1, last row; C1-R and T2). Naming these two wings collapse and evaporation is a toy-level interface label only, tier T3, carrying no load.

Interface conjecture (T3, the whole paragraph load-free). The momentum layer of this model is truncated by convention and deferred to the interface, following the division of [P2]. From this one may conjecture that the four absences called no frame dragging, no ergoregion provenance, no inheritance law, and no centrifugal term share this single truncation as their root. The boundary of what is actually proved here, item by item: no centrifugal term has a T1 exact meaning, namely the joint statement of Proposition 7.3 that nothing in the model lets the angular account enter radial legality or the radial budget; the in-model form of no dragging follows from the same proposition while the physical name is not established; the provenance of ergoregions and the inheritance of angular momentum are not established at all and remain interface-level observations, since $J_{\rm phys}$ itself is unestablished and the $J$ of collapse has nowhere in this model to be delivered. That the four share one root is a conjecture, not a theorem of this paper.

10 Conclusion and Firewall

Three concluding sentences. First, angular net flow lives at the transfer level, and its cumulative account is in general not exact. Second, the rigid-rotation spectrum and its interface lattice spacing are legislated jointly by the local graph structure and the radial clauses. Third, angular activity neither alters the shape of the smooth core nor supplies radial support in this model.

Four interface debts, listed singly. Shell weights $\alpha^{\rm phys}$: the paper declares only the two combinatorial gauges, edge-step and turn; the physical gauge is unestablished. Time scale: no bridge from ticks to physical time is built, and every rate here is a per-tick rate. Energy interface: no link from accounts to energy is established, and spectra with holes are combinatorial, not energetic. Delivery of $J_{\rm phys}$: physical angular momentum is not a quantity of this paper; its state character, conservation, and inheritance lie outside the scope.

The scope sentence, restated: this paper establishes a combinatorial angular-transport account; the shell weights, time scale, and energy interface required to pass from this account to physical angular momentum are all left unestablished. The four non-claims of Section 1 hold throughout as at the start.

Appendix A Primitive-Cycle Constructions and Legality

Constructions are given in coordinate lists. Legality is checked clause by clause against Definition 2.2 (b): at most one move out per source, at most one in per target, final occupancy unchanged; under saturation each construction is a permutation of its cell set, so the clauses hold automatically.

(a) Shared-parent three-cycle (Lemma 5.1 (i)): moves $(j,i)\to(j,i{+}1)$, $(j,i{+}1)\to(j{-}1,q)$, $(j{-}1,q)\to(j,i)$; projections $+1,0,0$; account $+1$; reversal gives $-1$. (b) Increment cycle $\gamma_i$ (ii): $(j,i{+}1)\to(j{-}1,p(i){+}g_i)$; $g_i$ retrograde steps below to $(j{-}1,p(i))$; $(j{-}1,p(i))\to(j,i)$; $(j,i)\to(j,i{+}1)$; $g_i{+}3$ edges, projections $0,(-1)^{\times g_i},0,+1$, account $1-g_i$; $g_i=0$ reduces to (a) in primary form, $g_i=2$ is a five-edge cycle of account $-1$. (c) Secondary cycle (iii): $(j,i)\to(j{-}1,q)$; $\delta_q(i)$ retrograde steps below to $(j{-}1,p(i))$; $(j{-}1,p(i))\to(j,i)$; account $-\delta_q(i)$; $\delta=1$ a three-cycle. (d) Degenerate clause (iv): with $C_{j-1}\le2$ the projection vanishes below, so the account of (b) is $+1$ regardless of $g_i$. All cycles are simple with pairwise distinct support cells, hence primitive.

Appendix B Full Proof of the Global Recursive Potential

Full proof of the tail clause of Corollary 5.3. Suppose every tail interface satisfies the hypotheses of Theorem 5.2 and set $d=\gcd_jG_j$. First, existence of per-interface residues: by the two congruences in the proof of Theorem 5.2, all generalized defects at radial endpoints of interface $j$ are congruent modulo $d$; call the value $c_j$; it is independent of the lift representative since the period drift $C_j-C_{j-1}\equiv0\pmod d$. Second, recursion of shell constants: put $s_{j_*}=0$ at the innermost tail shell and $s_j\equiv s_{j-1}+c_j\pmod d$; well defined, one constant per interface. Third, the potential: $\psi(j,i)=i-s_j\bmod d$ with columns in the shared orientation of Definition 2.5; ring wrap-around is harmless since $d\mid C_j$. Fourth, movewise increments: a tangential step in shell $j$ changes $\psi$ by $\pm1$, exactly its projection and its account; a radial step $(j,i)\to(j{-}1,q)$ is uncharged and changes $\psi$ by $q-s_{j-1}-(i-s_j)\equiv q-i+c_j\equiv c_j-D(i,q)\equiv0\pmod d$, ascents symmetrically. Fifth, closure: an occupancy-identity legal transfer decomposes into vertex-disjoint closed chains (Definition 2.2), each returning every cell and hence $\psi$ to its start, so the total account is $\equiv0\pmod d$. $\blacksquare$ The global clause follows by adjoining the generator $1$ of the degenerate interfaces (Appendix A (d)), containment trivial and generation real.

Appendix C Exact Enumeration Checks for the Analytic Theorems (C1-R)

All numbers come from one standard run of the scripts listed in Appendix F and have passed independent recomputation (the meaning of C1-R, Section 1). Onion of capacities $(1,3,4)$: occupancy-identity single components $642$ (swaps $18$, directed cycles $624$), legal composites $1995$; involution counts $1995=521+2\times737$ and, radial-free, $53=27+2\times13$ (witnesses of Lemma 4.5); full spectrum the segment $[-7,7]$; minimal positive closed-path account $9$ for the directed word length (instance of Proposition 3.6). In the edge-step gauge the two-shell radial-lattice spectrum of capacities $(3,4)$ misses $\pm2,\pm5,\pm6$ (note to Proposition 4.2). Capacities $(1,3,4,5)$: full spectrum $[-12,12]$, $25$ points. Capacities $(1,3,4,5,6)$: amplitude $M=18$, constructive certificates for all values $0$ to $18$. Capacities $(1,2,8)$: degenerate-shell projection verified zero, full spectrum the segment of amplitude $8$. Interface-subgroup double computation: cycle-account gcds of the six interfaces of the standard growing families all equal the analytic formula of Theorem 5.2, all $1$; default double-parent equal towers $C=3$ to $6$ likewise. Blind-zone witnesses: the pair of schedules with identical occupancy trajectories and diverging accounts, and the rotation-blind ablation inversion order, are both on record (witnesses of Proposition 6.2).

Appendix D Boundary Examples and Counterexamples

(1) Equal-capacity single-parent uniform tower (Table 5.1, row 1): capacities $(4,4,4)$, identity primary map, increments all one, no offsets; Theorem 5.2 gives $G=\gcd(4,4,0)=4$; enumeration confirms all cycle accounts in $4\mathbb Z$ and the full spectrum likewise; rigidity survives the cross-shell library, the account being winding times capacity. (2) One-cell centre (row 2): $C_0=1$ makes all shell-one cells share the centre; shared-parent three-cycles give $\pm1$ and global spacing one. (3) Congruent single-parent tower (row 3): capacities $(12,4)$, primary $p(i)=3i$, increments all $3$, $G=\gcd(4,12,2)=2$; enumeration confirms the full spectrum in $2\mathbb Z$; the two-interface tower $(36,12,4)$ gives $2$ per interface with all bounded-length cycle accounts even. (4) Default double parents (row 4): offsets all one enter the generators, $G=1$; the secondary three-cycles of capacities $(4,4)$ with accounts $\mp1$ are the witness, confluent with the shared-parent language. (5) Intermediate band (Problem 9.2): capacities $(7,5)$, single parent, increments $(3,1,1,1,1)$, with neither a shared ring-neighbour parent nor an increment in $\{0,2\}$; $G=1$ while the primitive cycle accounts are $\{0,\pm2,\pm5,\pm7\}$ without $\pm1$; the full spectrum is $\{0,\pm2,\pm5,\pm7,\pm12\}$, amplitude $12$ with sixteen holes: the boundary marker that unit spacing implies no packing (Problem 9.1). (6) In-class witness of the degenerate shortcut (Corollary 5.3): capacities $(2,12,4)$, tail interface spacing $2$ with measured tail cycle gcd $2$, yet a legal shortcut three-cycle of account $+1$, a legal transfer of odd account $+5$, and full-graph cycle gcd $1$. (7) Sequential trichotomy (Definition 2.10): the constant-parent $(3,3)$ example has a non-unique lift, all cells share one parent, and $G=1$ follows directly; the crossing $(3,3)$ example with $p=(0,2,1)$ admits no lift, cycle accounts $\{\pm1,\pm2,\pm3,\pm4\}$, registered as material for Problem 9.3.

Appendix E Protocols and Full Tables of the Differential-Rotation Statistics

Protocols. Component universe: both rigid rotations of every non-degenerate shell, all swaps, and all simple cycles of length at most six in both orientations, quotiented by support and account vector, each element passing the legality clauses of Definition 2.2. Families and coordinates as in Section 8. Sampling: independent local random generators per block with registered seeds, free $501$, clamped $503$, compensation $507$, neutral $509$, skin $511+v$; independent ticks, $T=20000$ (four thousand per $v$ in the skin block). Environment Python 3.12.3, NumPy 2.4.4; deterministic blocks reproduce bitwise, stochastic blocks bitwise in the declared environment. Bootstrap: independent resampling, $B=400$, quantiles $2.5$ and $97.5$. Wherever exact moments exist, the exact value from direct summation over the universe is canonical and empirical values are checks.

Full tables. Universe: components $1398/32/114$ with rotations $8/4/6$; global spacings $1/2/4$, divisibility of all universe accounts asserted per family. Free block: per-shell drifts exactly zero, empirical agreeing; neighbour-shell covariances (exact) default family $(1,2)\,{-}0.004$, $(2,3)\,{+}0.023$, $(3,4)\,{-}0.067$, congruent tower $(0,1)\,{-}0.750$, uniform tower $(0,1)$ and $(1,2)$ both $+0.351$; rotation shares exact $0.006/0.125/0.053$. Clamped block: $\mathsf A_2$ components $537/6/35$, zero-account shares $0.201/0.670/0.231$; normalized drifts default family $(0.0709,0.1519,0.1155,0.0363)$, congruent tower $(0.1641,0.1655)$, uniform tower $(0.1997,0.3720,0.1972)$. Compensation block (rates $0/0.25/0.5$): compensation rates $0.000/0.247/0.500$, per-tick retrograde flow $0.440/0.698/0.968$, per-tick total gain $0.988/0.947/0.892$. Neutral block: internal flow mean $2.608$, median $2.0$, interval $(2.582,2.635)$; per-tick increment variance rates $(0.661,1.143,1.236,1.259)$. Skin block ($C=12$):

$v$ 1 2 3 4 5 6 7 8 9 10 11
mobility.096.179.246.297.331.356.350.331.295.226.133
increment variance1.262.393.344.154.484.694.624.283.762.751.60

Appendix F Recomputation Material and Hashes

The numerical content of this paper comes from one standard run of four scripts and one dependency from the earlier series; per-file SHA-256 values follow, and the package hash of the published bundle is registered in the publication-record metadata following the series double-anchor convention.

spin.py: 7a60653de726d98b6ddd00f0152f9922bcad051e3528316666f45f7d6b96278a. angular.py: b200a17d7497cbf352aad8d2082ab5673ecacf5251b6f313be3dfe86d8bd7faa. families.py: fb4bd9611d68177fdafb0621b6b8f3f5b7ac50ef31a984140f415a292efc2509. diffrot.py: b605ccbe7f89250a27d56c5f275550ca34b6274378d7efddbae709adbc3aaf64. Dependency arealaw.py belongs to the published package of [P3], SHA-256 39ca2ab687935caa69dfa1c831db59f05899fa064cd40b954968e7f128347c07. Script package SHA-256: 24312a41889bae7fbd455e8127c81ad69f42fec9eec354cb65d08eb35b95cae6. Standard run: in one directory execute python3 spin.py, python3 families.py, python3 diffrot.py (the latter two importing angular.py and arealaw.py); this reproduces all numbers of Appendices C and E; environment Python 3.12.3 with NumPy 2.4.4; deterministic blocks reproduce bitwise, stochastic blocks bitwise in the declared environment; all numbers passed independent recomputation, hence C1-R.

Appendix G Complete Status Table

D (definitions and model inputs): Definitions 2.1, 2.2, 2.3, 2.5, 2.6, 2.7, 2.8, 2.9, 2.10, 4.4; the component universe and sampling measures; the three gauges. T1 (conditional theorems, hypotheses restated in the text): Lemma 2.4; Proposition 3.1; Lemma 3.2; Theorem 3.3; Corollary 3.4; the main sentence of Proposition 3.5; Proposition 3.6; Theorem 4.1; Proposition 4.2; Proposition 4.3; Lemma 4.5; Lemma 4.6; Lemma 5.1; Theorem 5.2; Corollary 5.3; Corollary 5.4; Proposition 6.1; Proposition 6.2; Lemma 7.1; Theorem 7.2; Proposition 7.3. T2 (assembled conditional readings): the reading "rotate forever, fall never" of Proposition 3.5; the four readings of Section 8; the angular voice of the three eras and the two-wing combinatorial reading in Section 9. T3 and Open: Problems 9.1 (Open), 9.2, 9.3; the interface-conjecture paragraph; the physical naming of the two wings; shell weights, time scale, energy interface, and the delivery of $J_{\rm phys}$. C1-R: all numbers of Appendices C and E.

Acknowledgements

The author thanks several large-language-model systems for assistance in derivation checking, adversarial review, and independent numerical recomputation; all claims and all errors are the author's own.

Funding and competing interests: this research received no external funding; the author declares no competing interests.

References

[P1] Qin, Han. Area-Level Plateau and Self-Sufficient Window in a Frozen-Ledger Model: A SAE black-hole information-flow-topology toy model (Chinese and English editions). Zenodo. DOI: 10.5281/zenodo.20913514.

[P2] Qin, Han. Update Incompatibility in Discrete Black-Hole Interiors: Three-Domain Separation, Orbit Dichotomy, and the Bridge Gap in Radial-Potential Lifting (Chinese and English editions). Zenodo. DOI: 10.5281/zenodo.21499053.

[P3] Qin, Han. Exclusion of Uniform Strict Potentials and Stratified Monotonicity in Discrete Black-Hole Interiors: Refuting the Same-Shell Bridge, Nested Observation Kernels, and the Release-Order Fork (Chinese and English editions). Zenodo. DOI: 10.5281/zenodo.21521760.