TheoremStatement: AI-adaptedProof: AI-generated

How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Universal Brink–Howlett covering and its deck group

PDF p. 1: Theorems 1.3 and 4.7; PDF p. 2: Theorems 1.3 and 4.7; PDF p. 16: Theorems 1.3 and 4.7; PDF p. 17: Theorems 1.3 and 4.7

Statement

Let $(W, S)$ be a Coxeter system of arbitrary rank in its standard real root representation, with diagram vertex set $Γ_{0}$ , and fix arbitrary $J \subseteq Γ_{0}$ . Put $W_{I} = ⟨ s_{i} : i \in I ⟩$ and $N (W, J) = {Norm}_{W} (W_{J}) / W_{J}$ . Let $C h a m (J)$ be the algebraic label set of pairs $(x, I)$ with $W_{J} x = x W_{I}$ and $x$ of minimum length in $x W_{I}$ . Write $B H = B H (J)$ for the Brink–Howlett groupoid and $\tilde{B H} = \tilde{B H} (J)$ for the groupoid with these labels as objects and one arrow $(x, I) \to (y, K)$ labelled $x^{- 1} y$ . The functor $F : \tilde{B H} \to B H$ defined by $(x, I) \mapsto I$ and $x^{- 1} y \mapsto x^{- 1} y$ is a universal groupoid covering, with $Deck (F) ≅ N (W, J)$ and $B H ≅ \tilde{B H} / N (W, J)$ .

For the geometric interpretation, use the finite-rank relative-cone convention: the set $Γ_{0}$ is finite, while neither $W$ nor $W_{J}$ is required to be finite and $J$ is arbitrary. Then the objects upstairs are exactly the relative chambers, and the adjacent-wall arrows map to the Brink–Howlett generators $v_{a, I} : I \to K$ . Here $I$ is an associate of $J$ , $a \in Γ_{0} ∖ I$ , $D = [I, a]$ is the component containing $a$ in the diagram induced on $I \cup {a}$ , and $W_{D}$ is finite. The family is given by $v_{a, I} = w_{D ∖ {a}} w_{D}$ and $K = (I \cup {a}) ∖ {ι_{D} (a)}$ , where $w_{L}$ is the longest element of a finite standard parabolic $W_{L}$ and $w_{D} α_{i} = - α_{ι_{D} (i)}$ defines $ι_{D}$ . The projected adjacent crossings generate $B H$ .

Statement references

Coxeter-system, roots, length and weak-order conventions

Tits cone intersections, relative roots and arrangements

Algebraic chamber-label set

Brink–Howlett groupoid

Universal wall groupoid on relative chambers

Groupoid coverings, universal and normal covers

Root-permuting normaliser complement

Coxeter normaliser semidirect product

Labelling theorem for relative chambers

Standard expressions and their lengths

Facts & Assumptions

[F1]

The standard real root representation, positive roots and Coxeter length use the arbitrary-rank conventions of Coxeter-system, roots, length and weak-order conventions. The arbitrary-rank label set has exactly the coset and length conditions in the statement. The base groupoid is small and connected, with arrow $w : I \to K$ exactly when $w A_{K} = A_{I}$ , where $A_{L} = {α_{l} : l \in L}$ . The upstairs groupoid is small and nonempty, has singleton hom-sets and left-to-right label multiplication, and contains $(1, J)$ . Algebraic chamber-label set Brink–Howlett groupoid Universal wall groupoid on relative chambers

[F2]

The minimum-coset, canonical-simple-system transport, standard-parabolic identification and canonical-system generation assertions recorded in External input apply at arbitrary rank.

[F3]

The root-permuting subgroup $N_{J} = {n : n A_{J} = A_{J}}$ is a subgroup of the normaliser. The quotient homomorphism restricts to an isomorphism $N_{J} \to N (W, J)$ . This result is proved algebraically at arbitrary rank and does not depend on the covering theorem. Root-permuting normaliser complement Coxeter normaliser semidirect product

[F4]

A covering is a functor bijective on arrows with every fixed source; a nonempty connected covering with trivial vertex groups is universal. A free action on objects of a connected small groupoid gives a quotient covering with deck group isomorphic to the acting group. Groupoid coverings, universal and normal covers Independent groupoid covering and universal-cover input Quotient groupoids give normal covers

[F5]

In the finite-rank geometric convention, the algebraic labels identify with the relative chambers and every upstairs arrow factors into a finite sequence of adjacent crossings. Tits cone intersections, relative roots and arrangements Labelling theorem for relative chambers The universal wall groupoid is generated by adjacent crossings

[F6]

The indexed Brink–Howlett family is $v_{a, I} = w_{D ∖ {a}} w_{D} : I \to (I \cup {a}) ∖ {ι_{D} (a)}$ for $D = [I, a]$ with $W_{D}$ finite. For adjacent labels the finite-rank wall formula supplies such an index, the same target $K$ , and $y = x v_{a, I}$ . Standard expressions and their lengths Adjacent chambers and inverse wall-crossing elements

[F7]

In finite rank, the geometric normaliser action takes a label $(x, I)$ to the minimum representative of the coset $g x W_{I}$ , and preserves the crossing index. Normaliser action on labels, orbits and adjacency

External input

The following premises are imported from Dyer–Gimenez, with exact locators and reading limits in the second source reference. Here a reflection subgroup is a subgroup generated by reflections in roots, and $Π_{U}$ denotes its canonical positive simple-root system in the ambient standard based root system; $Φ^{+}$ is the ambient positive root set.

Dyer–Gimenez permits arbitrary rank in §1.4. For any reflection subgroup $U$ and $z \in W$ , Proposition 2.10(b),(c) gives a unique minimum $d$ in $z U$ and says that $y \in z U$ equals $d$ exactly when $y Π_{U} \subseteq Φ^{+}$ . Proposition 2.10(d) gives $Π_{z U z^{- 1}} = d Π_{U}$ . Proposition 2.11(c) identifies $Π_{W_{L}} = A_{L}$ for every subset $L$ of the simple-root indices, where $A_{L} = {α_{l} : l \in L}$ . A reflection subgroup is generated by the root reflections indexed by its canonical simple system, as also stated in Theorem 2.9 with $U = W_{Δ} = ⟨ s_{β} : β \in Δ ⟩$ .

Proof

Given: The arbitrary-rank algebraic data in the first paragraph of the Statement, and the additional finite-rank convention only for its second paragraph. Goal: Prove the universal covering, deck and quotient assertions, and the stated geometric interpretation on its specified domain.

1.1F1F2

If $(x, I)$ is a label, then $x W_{I} x^{- 1} = W_{J}$ and $x$ is the minimum of $x W_{I}$ . Apply [F2] to $U = W_{I}$ and $z = x$ . Canonical-system transport and the standard-parabolic identification give $x A_{I} = A_{J}$ . Conversely, an element $x$ with $x A_{I} = A_{J}$ is the minimum of $x W_{I}$ by the positivity criterion. Its conjugate subgroup $x W_{I} x^{- 1}$ has canonical simple system $x A_{I} = A_{J}$ , again by transport. The generation assertion in [F2] identifies this subgroup with $W_{J}$ . Thus such an $x$ is exactly a label with second component $I$ .

1.2F5F6

Under the finite-rank convention, [F5] identifies all objects upstairs with the relative chambers. For an adjacent arrow from $(x, I)$ to $(y, K)$ , [F6] supplies the finite component, index, target and equality $y = x v_{a, I}$ . Its image has label $x^{- 1} y = v_{a, I}$ and exactly the target in the indexed family. Thus the adjacent arrows project to the stated Brink–Howlett generators.

2.1F1step 1.1

For two labels $(x, I), (y, K)$ , step 1.1 gives $x^{- 1} y A_{K} = A_{I}$ , so $x^{- 1} y$ is a base arrow. Conversely, if $w : I \to K$ is a base arrow and $(x, I)$ is a label, then $x w A_{K} = x A_{I} = A_{J}$ . Step 1.1 makes $(x w, K)$ a label. These statements hold at arbitrary rank.

2.2F1F3step 1.1

For $n \in N_{J}$ , define $n \cdot (x, I) = (n x, I)$ . Step 1.1 shows that this is a label, since $n x A_{I} = n A_{J} = A_{J}$ . It is an action by invertible permutations and extends uniquely to the singleton-hom-set groupoid. It preserves every projected arrow label because $(n x)^{- 1} (n y) = x^{- 1} y$ . It is free on objects because $n x = x$ implies $n = 1$ . Two labels with the same second component are in one orbit: $n = y x^{- 1}$ satisfies $n A_{J} = y A_{I} = A_{J}$ and $n x = y$ . Conversely the action preserves the second component. Thus its orbits are exactly the fibres of $F$ .

3.1F1F4step 2.1

The proposed functor is well-defined by step 2.1. Its identity and composition laws follow from $x^{- 1} x = 1$ and $(x^{- 1} y) (y^{- 1} z) = x^{- 1} z$ . Fixing a source $(x, I)$ and a base arrow $w : I \to K$ , step 2.1 gives its lift to $(x w, K)$ with label $w$ . Any lift must end at $(y, K)$ with $x^{- 1} y = w$ , hence $y = x w$ , and its hom-set is a singleton. This proves the source-star bijection. Base connectedness and $(1, J)$ give a lift to an object over every base object.

3.2F3F7step 1.1step 2.2

Transfer this action along the group isomorphism $N_{J} \to N (W, J)$ in [F3]. In finite rank it agrees with the geometric action in [F7]: the representative $n$ of a quotient class sends $(x, I)$ to $(n x, I)$ , and step 1.1 makes $n x$ the minimum of $n x W_{I}$ , exactly the formula of [F7]. Hence the quotient action used here has the claimed geometric meaning on that domain.

4.1F1F4step 3.1

The upstairs groupoid is nonempty and connected and its vertex groups are trivial, by the singleton hom-sets in [F1]. Consequently the covering is universal by [F4].

4.2F4step 3.1step 2.2step 3.2

Apply [F4] to the free $N (W, J)$ action and write $Q$ for its quotient covering. By step 2.2, $F$ factors through $Q$ . The induced functor is bijective on objects, since its orbits are exactly the fibres and step 3.1 gives every base object. Each arrow orbit has a unique representative with any fixed source in its source orbit by [F4]; from that source the source-star bijection in step 3.1 gives exactly the base arrows. The induced functor is therefore also bijective on arrows and is an isomorphism. This proves the quotient assertion.

5.1F4step 4.2

The quotient-cover theorem gives $Deck (Q) ≅ N (W, J)$ . Since the induced quotient-to-base functor in step 4.2 is an isomorphism, $Q ϕ = Q$ holds exactly when $F ϕ = F$ . Their deck groups coincide, proving the asserted deck-group identification.

6.1F5F6step 3.1step 1.2∎

Every base arrow lifts by step 3.1 and its lift factors into adjacent arrows by [F5]. Projecting that factorization proves that the projected adjacent crossings generate the base. This establishes the geometric paragraph of the statement without imposing any sphericity condition on $W$ or $W_{J}$ .

Source notes

The article's Theorems 1.3 and 4.7 and their covering, deck and quotient conclusions remain the attribution. The article omits an explicit global rank qualifier. Its §2.2 follows Iyama–Wemyss for the dual representation; Proposition 3.6 invokes finite dimension, and Lemma 4.2 says that $I, K$ are finite. The current chamber-labelling supplier records that its independently sourced Iyama–Wemyss foundation starts with an $n \times n$ Coxeter matrix and allows infinite Coxeter groups. That independent source was not directly read in this adjudication. These observations are consistent with the finite-rank geometric convention already explicit in the current relative-cone definition; omission of a rank qualifier alone does not establish an intended infinite-dimensional geometric realization.

The local proof establishes the covering, deck and quotient conclusions at arbitrary rank, using the exact independent Dyer–Gimenez inputs and the existing algebraic normaliser splitting. The current splitting proof and its five algebraic helpers were read: they construct the complement by minimum coset representatives, without this covering theorem. Thus using that supplier here creates neither a logical nor a declared dependency cycle, despite the original article deriving its splitting corollary from the covering theorem. The arbitrary-rank lifting correspondence is proved locally, rather than applying the finite-rank chamber-morphism proposition outside its scope.

In finite rank, the retained chamber, adjacent-generator and generation conclusions apply the existing geometric suppliers and the constructed action agrees with the geometric normaliser action. No topology or chamber theorem for infinitely many generators is asserted. The original unqualified formulation remains recorded in the attribution reference; no original-article defect is established. The older Brink–Howlett, Lusztig, Howlett and Vinberg proofs have not been read for this adjudication.

Depends on

Used by · 5

Dependency tree · 2 levels

Each result is drawn at its shortest distance from this item. Arrows run from prerequisites to the items that use them; this item has a heavier outline. External source nodes link directly to the cited text. The view shows up to two dependency steps; open a prerequisite to continue.

External dependencies

Matthew Dyer and Harrison Gimenez, Root systems, Tits cones and imaginary cones of Brink–Howlett groupoids · §§1.3–1.4, PDF pp.3–4; §1.7, pp.4–5; Theorem 2.9, PDF p.22; Proposition 2.10(b)–(d) and Proposition 2.11(c), PDF p.23

Dyer–Gimenez permits arbitrary rank in §1.4. For any reflection subgroup $U$ and $z\in W$, Proposition 2.10(b),(c) gives a unique minimum $d$ in $zU$ and says that $y\in zU$ equals $d$ exactly when $y\Pi_U\subseteq\Phi^+$. Proposition 2.10(d) gives $\Pi_{zUz^{-1}}=d\Pi_U$. Proposition 2.11(c) identifies $\Pi_{W_L}=A_L$ for every subset $L$ of the simple-root indices, where $A_L=\{\alpha_l:l\in L\}$. A reflection subgroup is generated by the root reflections indexed by its canonical simple system, as also stated in Theorem 2.9 with $U=W_\Delta=\langle s_\beta:\beta\in\Delta\rangle$.

Hypotheses: Arbitrary-rank Coxeter system in its standard real based root system; arbitrary reflection subgroup U and arbitrary subsets of simple roots. No spherical, finite-rank or topological assumption.

Reading limits: Read the frozen exact v1 PDF text of §§1.3–1.4 and §1.7 (PDF pp.3–5), §2.8 and Theorem 2.9 (PDF p.22), and Proposition 2.10(b)–(d), Proposition 2.11(c) and their adjacent proof/reference paragraphs (PDF p.23). Visually inspected the supplied PDF page images 4 and 23 and checked the arXiv v1 bibliographic record. The standard root system is explicitly allowed in §1.3 and arbitrary rank in §1.4. The cited older proofs were not read or reconstructed. Only the displayed minimum-coset, canonical-system transport, standard-parabolic identification and generation assertions are imported; the covering argument is local.

Sources

Normalisers of parabolic subgroups of Artin–Tits groups and Tits cone intersections · PDF p. 2, Theorem 1.3 and following paragraph; pp. 15–17, Lemma 4.2, Definition 4.3, Propositions 4.4 and 4.6, Theorem 4.7 and Definition 4.9; p. 4, §2.2; p. 9, Proposition 3.6
Matthew Dyer and Harrison Gimenez, Root systems, Tits cones and imaginary cones of Brink–Howlett groupoids · §§1.3–1.4, PDF pp.3–4; §1.7, pp.4–5; Theorem 2.9, PDF p.22; Proposition 2.10(b)–(d) and Proposition 2.11(c), PDF p.23