# The order-46 observable ladder **Date:** 25 October 2026 **Status:** genus-one rail separation, modular transport and cyclic anomaly status closed; compatible normalization and genus-two sewing open **Evidence:** [observable-ladder ledger](../../../evidence/data/moonshine-order46-observable-ladder-v1.json) **Certificate:** [arithmetic and cross-ledger verifier](../../../verification/verify/verify_moonshine_order46_observable_ladder.mjs) ## Result The order-46 junction does more than prove that the 2A, 2B and 23 branches meet. It supplies a two-stage observable ladder. At order 23, the scalar McKay--Thompson series collapses both the generator orientation and the involution rail: \[ T_{23A}=T_{23B}. \] At order 46, ordinary Moonshine recovers the involution rail while still forgetting orientation within each rail. The published genus-zero groups are \[ 46AB:\quad 46+23, \qquad 46CD:\quad 46+. \] Here \(46+23\) denotes \(\Gamma_0(46)\) extended by the \(W_{23}\) Atkin--Lehner involution, while \(46+\) is the full Atkin--Lehner extension. Consequently \[ T_{46A}=T_{46B}, \qquad T_{46C}=T_{46D}, \qquad T_{46AB}\ne T_{46CD}. \] The ordinary order-46 trace therefore distinguishes the 2B/Leech rail from the 2A/Baby-Monster rail, but it does not orient either rail. ## 1. The root is encoded by the commuting pair Let \[ g=h^2,\qquad z=h^{23} \] for an order-46 element \(h\). The Chinese-remainder reconstruction is explicit: \[ \boxed{h=g^{12}z} \] because \(g^{12}z=h^{24}h^{23}=h^{47}=h\). Thus the order-46 element is not an auxiliary label added to the \(23\times2\) data. It is the unique combined insertion selected by that commuting cyclic pair. This explains why \(T_h\) can recover the involution rail even though \(T_g\) cannot. The order-23 square \(g\) forgets the parity component \(z\); the combined insertion \(g^{12}z\) restores it. ## 2. The desired twisted trace is a modular transport Generalized moonshine assigns a function \(Z(g_0,h_0;\tau)\) to commuting pairs, with an \(SL_2(\mathbb Z)\) action of the form \[ Z(g_0,h_0;\gamma\tau) \doteq Z(g_0^a h_0^c,g_0^b h_0^d;\tau), \qquad \gamma=\begin{pmatrix}a&b\\c&d\end{pmatrix}, \] where \(\doteq\) allows the projective multiplier fixed by the anomaly and normalization convention. Take \[ \gamma_{2,23} =\begin{pmatrix}1&11\\2&23\end{pmatrix}, \qquad \det\gamma_{2,23}=23-22=1. \] Starting from the ordinary twining pair \((1,h)\), it gives \[ (1,h)\gamma_{2,23}=(h^2,h^{23})=(g,z). \] Therefore \[ \boxed{ Z(g,z;\tau) \doteq T_h\!\left(\frac{\tau+11}{2\tau+23}\right) } \] up to the precise direction of the convention and its projective multiplier. The missing order-46 twisted--twining function is not 529 independent input series and not an arbitrary new function. Once a convention is fixed, it is a modular branch of a known order-46 Hauptmodul. ## 3. The anomaly bound closes, leaving normalization The order-23 restriction of the Monster anomaly is forced to vanish because its order divides both 24 and 23. The same argument does not trivialize the order-46 restriction: \[ \gcd(24,46)=2. \] At this coarse stage only a possible two-torsion obstruction remains. The published Moonshine notation closes it. An anomalous cyclic class is written \(n|h+\cdots\) with \(h>1\), whereas the two order-46 functions are written \(46+23\) and \(46+\). Both therefore have \(h=1\): they are normal, non-anomalous cyclic orbifolds. The restriction of the Monster anomaly to each order-46 cyclic subgroup is cohomologically trivial. This does not manufacture an absolute phase. A trivialization and the normalization of each generalized-moonshine function must still be chosen consistently, and inverse conjugacy classes may still be identified. But any remaining scalar is a convention/coboundary datum rather than a nontrivial \(\mathbb Z_2\) obstruction. In particular, the arithmetic possibility \(\gcd(24,46)=2\) is not realized by these two cyclic class pairs and cannot be identified with the genus-two Pfaffian sign. The immediate source-level calculation is therefore: 1. fix compatible anomaly trivializations and generalized-moonshine normalizations for 46A--46D; 2. transport \(T_{46AB}\) and \(T_{46CD}\) by \(\gamma_{2,23}\) (or its inverse under the chosen convention); 3. normalize the transported functions by a common principal-part convention and verify the power-five pairing described below; 4. compare the surviving sign/character with the 2A fermionic determinant line and the 2B order-eight lift kernel; 5. only then sew at genus two. ## 4. Symmetry and compression count The index of \(\Gamma_0(46)\) in \(PSL_2(\mathbb Z)\) is \[ 46\left(1+\frac12\right)\left(1+\frac1{23}\right)=72. \] The full Atkin--Lehner group at squarefree level 46 has four elements, while the \(+23\) extension has two. Thus the corresponding indices are 18 and 36. The 46CD function is invariant under the larger group and is consequently the more compressed scalar record. This is a group-theoretic measure of retained information; it is not a count of physical states. The observable hierarchy is now ```text order-23 scalar T_23 1 output: loses involution rail and generator orientation | | retain the order-2 component through h=g^12 z v order-46 scalar T_h 2 outputs: 46AB (2B rail) versus 46CD (2A rail) | | modularly transport within the same cyclic subgroup v generalized pair Z(h^2,h^23) 2 normalized scalar outputs: the inverse/power-five pairs remain collapsed | | scalar centralizer insertions are exhausted; | the linear/projective S4 action is also exhausted; | change compositional level to VOA products or genus-two sewing v candidate seam-sensitive vector ``` The pairing follows directly. Fifth power exchanges \(46A\leftrightarrow 46B\) and \(46C\leftrightarrow46D\). Under \(h\mapsto h^5\), \[ g=h^2\mapsto g^5,\qquad z=h^{23}\mapsto z^5=z. \] The ordinary functions satisfy \(T_h=T_{h^5}\) within each pair. Applying the same matrix \(\gamma_{2,23}\) to the same normalized Hauptmodul therefore gives the same normalized cyclic commuting-pair function for \((g,z)\) and \((g^5,z)\). A convention-dependent scalar cannot serve as an orientation observable. Thus cyclic generalized moonshine recovers the 2A/2B rail but does not recover the 23A/23B generator orientation. The subsequent complete \(C_{23}\times S_4\) census shows that every individual centralizer insertion also generates a cyclic subgroup and is equally scalar-degenerate. The next refinement must therefore change compositional level. The anomaly transgression to the centralizer is trivial, so even the full ordinary \(S_4\) character is determined by those five traces. Only cross-grade VOA operations or higher-genus sewing remain. ## 5. Consequence for the frontier programme The first frontier tranche can now be made smaller. There is no need to search for a free-standing order-46 generalized trace table before doing any work. Start from the two known Hauptmoduln, transport them by the explicit matrix and audit the compatible normalization. This is an FPT-002 inverse-detector problem: the order-46 insertion is the minimal extra channel already known to separate the two involution rails. It is also an FPT-019 complexity reduction: the unknown input is cross-grade VOA composition or a higher-genus refinement, not a 529-function oracle or another centralizer representation table. The physical boundary is unchanged. Modular transport is worldsheet/VOA transport, not real-time evolution, and its normalization phase is not automatically a laboratory parity signal. ## Sources and corpus links - Duncan, Griffin and Ono, [*Moonshine*](https://arxiv.org/abs/1411.6571), table of Monster classes and genus-zero groups - Carnahan, [*Generalized Moonshine IV: Monstrous Lie algebras*](https://arxiv.org/abs/1208.6254) - Tuite, [*Generalised Moonshine and Abelian Orbifold Constructions*](https://arxiv.org/abs/hep-th/9412036), for the \(n|h\) anomaly convention - [ATLAS Monster conjugacy classes](https://brauer.maths.qmul.ac.uk/Atlas/v3/spor/M) - [Joined order-46 bridge](joined-bridge-map.md) - [Frontier target crosswalk](frontier-target-crosswalk.md)