Book 8 · Chapter 8.7 · Monster Arc

The Pariah Groups
Exceptions to the Monster

Six simple groups that do not fit. Not subgroups of the Monster. Not quotients. Not relatives. Exceptions to the last exceptional thing. — after Conway and Norton, 1979

The classification of finite simple groups has two kinds of sporadics: the 20 groups in the Monster's "Happy Family" — subgroups or quotients of the Monster — and the 6 pariah groups, which are not related to the Monster in any way the classification can detect. The pariah groups are the genuine exceptions: the finite simple groups for which Monstrous Moonshine gives no prediction, for which no McKay-Thompson series exists (or exists with genus greater than zero), and for which no physical realisation has been found in the string-theoretic context of $V^\natural$.

· · ·

The Happy Family (20 Groups)

The Monster's Subordinates

The 20 Happy Family members: $\mathbb{M}$ (Monster) · $\mathbb{B}$ (Baby Monster) · $Fi_{24}'$ · $Co_1$ · $Co_2$ · $Co_3$ · $HS$ · $McL$ · $He$ · $Ru$ — wait, Ru is a pariah. The actual Happy Family: $\mathbb{M}, \mathbb{B}, Fi_{24}', Co_1, Co_2, Co_3, HS, McL, He, HN, Th, Fi_{23}, Fi_{22}, J_2, Suz, M_{11}, M_{12}, M_{22}, M_{23}, M_{24}$. All 20 are subgroups, quotients of subgroups, or sections of the Monster $\mathbb{M}$.

· · ·

The Six Pariahs

J₁ — Janko group
|J₁| = 175,560 = 2³ · 3 · 5 · 7 · 11 · 19
Discovered by Zvonimir Janko (1965) — the first new sporadic group in over 100 years. Acts on a 7-dimensional space over GF(11). No connection to the Monster. Its McKay-Thompson series is not a Hauptmodul for a genus-zero group — the moonshine correspondence fails.
J₃ — third Janko group
|J₃| = 50,232,960 = 2⁷ · 3⁵ · 5 · 17 · 19
Predicted by Janko (1969), constructed by Higman and McKay. Has a 9-dimensional faithful representation over GF(4). Thompson showed it is not a subgroup or section of any larger sporadic group including the Monster.
J₄ — fourth Janko group
|J₄| ≈ 8.6 × 10¹⁹
= 2²¹ · 3³ · 5 · 7 · 11³ · 23 · 29 · 31 · 37 · 43
Predicted by Janko (1975), constructed by Norton and others. The largest of the Janko groups. Has a 112-dimensional faithful module over GF(2). Despite its size, it is not a section of the Monster.
Ly — Lyons group
|Ly| ≈ 5.1 × 10¹⁸
Predicted by Richard Lyons (1972), constructed by Sims using a computer. Acts on a graph with 8,835,156 vertices. Named after Lyons; also called the Lyons-Sims group. Not in the Monster's family.
Ru — Rudvalis group
|Ru| = 145,926,144,000 = 2¹⁴ · 3³ · 5³ · 7 · 13 · 29
Predicted by Arunas Rudvalis (1973), constructed by Conway and Wales. Acts on a 28-dimensional space over GF(2). Despite being discovered by Conway (who also worked on the Monster), it is not a section of $\mathbb{M}$.
O'N — O'Nan group
|O'N| = 460,815,505,920 = 2⁹ · 3⁴ · 5 · 7³ · 11 · 19 · 31
Predicted by Michael O'Nan (1976), constructed by Sims. Has a 10-dimensional representation over GF(7). The O'Nan group is the pariah with the most active recent research: Moonshine for $O'N$ (2017) connects it to weight-$\frac{3}{2}$ modular forms, but this is not Monstrous Moonshine.
· · ·

The dm³ Reading of the Pariahs

In the dm³ framework, the Monster group $\mathbb{M}$ is the symmetry group of the fully completed G-chain: $G = U \circ F \circ K \circ C$ operating on a contact 3-manifold with all four operators acting at their full algebraic potential. The Happy Family members are the symmetry groups of sub-chains: systems where some but not all operators are active, or where the chain operates at reduced capacity.

The pariah groups, in this reading, are the symmetry groups of systems that run a version of the G-chain but with a different contact structure — not a restriction of the dm³ contact form, but an incompatible contact structure. These systems cannot be embedded in the Monster's contact geometry because their fold operator $F$ lives in a different topological class.

The O'Nan group's connection to weight-$\frac{3}{2}$ modular forms (rather than weight-0 Hauptmoduls as in Moonshine) suggests exactly this: its fold operator lives in a half-integer weight space, which in the contact-geometric language corresponds to a spin contact structure — a choice of square root of the canonical bundle on the contact 3-manifold — that is not compatible with the dm³ choice. The O'Nan group is the symmetry of a different fold.

· · ·

What the Pariahs Mean Physically

The physical interpretation of the pariah groups remains open. The Happy Family has physical realisations in string theory (the Monster), the Baby Monster ($\mathbb{B}$, connected to a specific orbifold of $V^\natural$), and the Mathieu groups ($M_{24}$, connected to K3 surfaces). The pariah groups have no known physical realisation as of 2026.

One possibility: the pariah groups govern systems that run a G-chain but fail to complete it — systems that get stuck at the fold $F$ or that enter the fold but cannot exit through the unfold $U$. These would be systems in "frozen" states: highly compressed, with a bifurcation surface ($K$) but no resolution. Black holes before evaporation, trapped enzymatic states, market microstructures that cannot clear — the physical fingerprints of pariah symmetry might be exactly the signatures of systems that compressed and folded but never unfolded.

The pariahs are not failures. They are the exceptions that prove the rule. If the Monster were the only symmetry, we would not know it was a symmetry. The pariahs are the contrast that makes the Monster visible.
CC BY 4.0 · G6 LLC · Pablo Nogueira Grossi · g6llc@proton.me · ORCID 0009-0000-6496-2186 · 2026