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.
Coxeter Descents, Poincaré Polynomials, and Growth — Examples
1 · Prerequisites
- Abelian Categories
- Absolute and Conditional Convergence; Rearrangement; Products
- Adjunctions Units and Counits
- Affine Algebraic Sets and Coordinate Rings
- Algebraic Closure, Embeddings, and Separability
- Algebraic Extensions, Extension Degree, and Finite Fields
- Associated Primes and Primary Decomposition
- Binary Operations, Monoids, Groups and Subgroups
- Bipartite Coxeter Elements and Ordered Root Complexes
- Canonical Roots, Signs, and Faithful Reflections
- Cardinal Arithmetic, Cofinality and the Alephs
- Cartan Subalgebras and Root Space Decompositions
- Categories, Functors and Natural Transformations
- Chain Complexes and Homology
- Chain Conditions, Semisimple Modules and the Wedderburn–Artin Theorem
- Chain Homotopy and the Homotopy Category
- Chains, Antichains, Sperner and Dilworth
- Compactness
- Compactness in Metric Spaces
- Completeness, Completion, and Uniform Continuity
- Complexification, Realification and Real Structures
- Congruences, the Integers Modulo n and the Chinese Remainder Theorem
- Conjugacy in Sₙ, Generation, and the Simplicity of Aₙ
- Connectedness
- Construction of the Natural Numbers
- Construction of the Real Numbers via Cauchy Sequences
- Construction of the Real Numbers via Dedekind Cuts
- Continuity, IVT, EVT, and Uniform Continuity
- Cosets, Index and Lagrange's Theorem
- Countability and Uncountability
- Coxeter Descents, Poincaré Polynomials, and Growth
- Coxeter Presentations, Exchange, and Reduced Word Theorems
- Cyclic Groups and Direct Products
- Delta Functors and Universality
- Depth and Cohen Macaulay Modules
- Derived Functors
- Determinants of Matrices over a Commutative Ring
- Diagonalisation and the Minimal Polynomial
- Divisibility, Euclidean Domains, Principal Ideal Domains and Unique Factorisation
- Divisibility, Greatest Common Divisors and Bézout's Identity
- Dual Spaces, Bilinear and Quadratic Forms, and Sylvester's Law of Inertia
- Eigenvalues, Eigenvectors and the Characteristic Polynomial
- Exactness and the Member Calculus
- Ext and Balanced Resolutions
- Filters and Ultrafilters
- Finite Counting, Factorials and Binomial Coefficients
- Finite Coxeter Diagrams and Complete Classification
- Finite Coxeter Invariants and Coinvariant Gradings
- Finite Fields and Cyclotomic Extensions
- Finite Reflection Arrangements and Spherical Coxeter Complexes
- Finite Weyl Invariants, Bruhat Order, and Kostant Harmonics
- Formal Power Series
- Foundations of the Real Numbers for Analysis
- Free Groups and Presentations
- Free Modules, Exact Sequences, Projective and Injective Modules
- Fundamental Trigonometric Identities
- Further Trigonometric Identities and Inverse Functions
- Graphs, Walks and Connectivity
- Group Actions, Orbits, Stabilisers and Cayley's Theorem
- Group Homomorphisms and the Isomorphism Theorems
- Hilbert Space Geometry and Riesz Representation
- Ideals, Quotient Rings and the Isomorphism Theorems for Rings
- Incidence Algebras and Möbius Inversion
- Inner Product Spaces, Gram-Schmidt, Projections and Adjoints
- Koszul Complexes and Regular Sequences
- Krull Dimension and Height Theorems
- Lie Algebra Representations, Enveloping Algebras, and PBW
- Limits and Colimits
- Limits of Real Functions
- limsup, liminf, and Subsequential Limits
- Linear Independence, Bases and Dimension
- Linear Recurrences and Rational Generating Functions
- Linear Transformations, Rank-Nullity and Quotient Spaces
- Localisation of Modules and Support
- Long Exact Sequences in Homology
- Maschke's Theorem, Complete Reducibility and the Structure of k[G]
- Matrices, the Matrix of a Linear Map, and Change of Basis
- Metric Spaces
- Modules, Submodules, Quotient Modules and the Isomorphism Theorems
- Monotone Functions, Discontinuities, and Continuity Sets
- Monotone Sequences, Bolzano-Weierstrass, and Cauchy Completeness
- Noetherian Rings and Hilbert Basis
- Normal Subgroups and Quotient Groups
- Normed and Banach Spaces
- Order, Zorn's Lemma, and the Axiom of Choice
- Ordinal Arithmetic and the First Uncountable Ordinal
- Ordinals, Cardinals, and Transfinite Recursion
- Parabolic Subgroups and Double Coset Geometry
- Partitions of Unity and Paracompactness
- Permutation Statistics, Inversions and Eulerian Numbers
- pi: the Equivalent Characterizations
- Polynomial Rings, the Division Algorithm and Roots
- Power Series and Real-Analytic Functions
- Preadditive and Additive Categories and Biproducts
- Prime Spectra and Radicals
- Primes, Euclid's Lemma and the Fundamental Theorem of Arithmetic
- Projective and Injective Resolutions
- Properties of the Integral and the Working FTC
- Real Forms and Reflection Geometry
- Rees Modules Artin Rees and Hilbert Samuel Theory
- Reflective Subcategories and the Adjoint Functor Theorems
- Relations, Functions, and Quotients
- Rings, Subrings, Integral Domains and Fields
- Rⁿ as a Normed Space; Vector-Valued Functions
- Root Systems, Dynkin Diagrams, and the Cartan-Killing Classification
- Roots, Rational Powers, and Classical Inequalities
- Separation Axioms: the Hierarchy
- Sequences and Limits
- Sequences and Series of Functions; Uniform Convergence
- Series: Convergence and the Nonnegative Tests
- Simple Field Extensions and the Construction of the Complex Numbers
- Simplicial Complexes and Simplicial Homology
- Sine, Cosine, and the Definition of Pi
- Solvable and Nilpotent Lie Algebras
- Splitting Fields
- Subobject Lattices Generators and the Grothendieck Axioms
- Subspaces, Products, and Quotients
- Suprema and Infima
- Symmetric Groups, Cycle Decomposition and the Sign Homomorphism
- Tensor Products of Modules
- The Cantor Set, Baire Category, and Measure Zero in ℝ
- The Complex Exponential and Euler's Formula
- The Derivative and the Mean Value Theorems
- The Determinant of a Linear Operator, Cofactors and Cramer's Rule
- The Diagram Lemmas in an Abelian Category
- The Exponential Function
- The Field of Fractions and Localisation
- The Fundamental Theorem of Finite Abelian Groups
- The Group Algebra and Representations of Finite Groups
- The Riemann Integral: Definition and Integrability
- The Spectral Theorem, Positive Operators and Singular Value Decomposition
- The Topology of Euclidean Space
- The Total Derivative in ℝᵐ → ℝⁿ
- The ZFC Axioms and the Basic Set Constructions
- Tits Cones, Chambers, and Parabolic Stabilizers
- Topological Spaces and Continuity
- Topology of ℝ
- Trees, Forests and Spanning Trees
- Triangularisation, Generalised Eigenspaces and Jordan Canonical Form
- Universal Properties, Representables and the Yoneda Lemma
- Vector Spaces, Linear Subspaces, Span and Direct Sums
- Weak Order, Inversions, and Lattice Operations
- Zariski Tangent Spaces, Regular Points, Smoothness, and Bertini
2 · Summary
This companion is a dependency leaf. Its examples use only the theory of coxeter-descents-poincare-polynomials-and-growth and that page's established prerequisite closure; no other theory page may depend on a supplier homed here.
Poincare products for Sn, Bn and Dn by explicit insertion re-derives the type-A products by inserting at each one-based position: this adds exactly inversions and yields . In the signed-permutation model, the type-B cosets over the subgroup fixing correspond to the choices , giving the factor and . For type D with , deleting the terminal node gives the parabolic; the even-signed Schreier graph has distances , twice, and , giving the factor and, by telescoping from , . The low-rank bases and checks are , , , , , , and . The product is , whose coefficient sum is . Evaluations at give , and for the symmetric, signed and even-signed permutation groups.
The A2 = S3 case: Steinberg inclusion-exclusion, degree product, and reciprocity uses the smallest-rank finite Coxeter system whose diagram is not a product of copies of . It relabels the library's on order-preservingly as permutations of , lists all six elements and lengths, and computes every right-descent interval directly and through inclusion-exclusion. The finite Steinberg identity is verified after clearing denominators; the degree product , group order , exponent sum , and reciprocity are checked explicitly. The finite enumeration uses no choice.
Infinite dihedral growth, the infinite Steinberg identity, and the failure of polynomial reciprocity uses the alternating normal forms to count one identity and two elements of every positive length, giving . Its spherical subsets are , and the infinite Steinberg identity becomes . In , , so a finite- reciprocity identity would force ; this fails for every . The alternating lengths are unbounded, so the group has no longest element. Its rank-one parabolic still has .
Each example states its hypotheses in full and verifies the displayed calculation; the infinite-dihedral example exhibits the exact point where the finite reciprocity argument stops. Exact item dependencies and source reading limits are recorded in research/coxeter-scaffold/inventory.json and research/plan-coxeter-groups-track.md.
3 · Logical flowchart
4 · Definitions, theorems and proofs
None yet.
5 · Examples, counterexamples and false statements
Poincare products for Sn, Bn and Dn by explicit insertion
Example
This example re-derives the products of Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (3) by explicit insertion in the same models, with . For the orbit labels below, write for the scaled dual coordinate functionals from that item; a vertex labelled denotes the corresponding orbit point .
(1) Type : inserting a letter into a permutation. Let . Use the order-preserving identification of the library's with permutations of ; under it the adjacent generator is and is the inversion number (The finite symmetric group , one-line notation, and cycle notation, Inversions, inversion number, the sign , and even and odd permutations, Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (1)(a), Proof 3.1). In this one-based notation, every and every position give a permutation by inserting the letter between positions and . The map is a bijection , and since the new inversions are exactly the pairs formed by with the entries to its right, Hence , so and .
(2) Type : inserting a sign. For , let be the signed permutation group acting on as in Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (1)(b) and (2)(b), with the subgroup fixing , as identified by Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (1)(b), (2)(b), and the parabolic-length fact in its Fact F16. Each has a unique decomposition with and the minimal element of its left coset ; the possible correspond bijectively to the orbit points labelled , and the distance (minimal coset length) of the representative sending to is , while for it is ; the list is obtained by moving along the chain , applying the sign change of the last coordinate, and returning along the negative chain. By The orbit of a dual fundamental functional: stabilizer, minimal coset length, Schreier distance, and the quotient formula (2),(4),
(3) Type : even signs. For , let be the even signed permutation group and let be the parabolic obtained by deleting the terminal node (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (1)(c),(2)(c)). The orbit of the corresponding dual fundamental functional has the Schreier graph obtained from the chains and by the cross edges and ; its distances are , then twice, then (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (2)(c)). Thus the quotient polynomial is . The orbit quotient formula gives ; since , this recurrence telescopes from in (4) to for . The cases are checked separately in (4).
(4) Bases and consistency checks. , , (a symmetric unimodal inversion-count sequence), , ; for one has , (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (3), Proof 5.1) with and , and gives a palindromic polynomial of degree whose coefficients sum to . The evaluations at recover , and .
Facts & Assumptions
Given: Integers , the symmetric group with its type presentation, the signed permutation groups and acting on with orthonormal basis , and the series .
Under the order-preserving identification of with permutations of , the type- generators map to adjacent transpositions and (The finite symmetric group , one-line notation, and cycle notation, Inversions, inversion number, the sign , and even and odd permutations, Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (1)(a), Proof 3.1).
The canonical model of is the full signed permutation group on , generated by adjacent coordinate exchanges and the sign change of coordinate ; the canonical model of is the even signed permutation group, with the exchange of coordinates , and () the exchange of coordinates (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (1)(b),(c)). In the model the subgroup generated by fixes and consists of all signed permutations of the remaining coordinates; in type , deleting gives the displayed parabolic.
In the type- and type- models, write for the dual coordinate functionals; the orbit points are labelled by . Their minimal-coset distances are on and on in type , and , twice, in type (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (2)(b),(c)).
For the dual-functional orbit of a deleted node, is the minimal length in the left coset (Left and right cosets and of a subgroup) and (The orbit of a dual fundamental functional: stabilizer, minimal coset length, Schreier distance, and the quotient formula (2),(4)).
As Coxeter systems, , and (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (3), Proof 5.1); the earlier product lemma gives , and the type- product (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (3)).
The length series is and has finite coefficients; for finite , evaluating at counts the elements of (Length generating series, descent-class series, spherical subsets, and the multivariate descent polynomial (1), The cardinality of a finite set).
For each standard parabolic , the restricted matrix presents the Coxeter system and its intrinsic length equals the ambient length restricted to (Support, intrinsic parabolic presentations, minimal coset representatives and length additivity, with the type-A identification (2)).
Verification
Type . Use the order-preserving relabeling of the library's as permutations of from [L1]. Deleting the letter from a permutation of inverts the insertion , so the map is a bijection . In the one-line form, inserting at position creates an inversion exactly with each of the entries to its right and with none of the entries to its left, and the relative order of the other entries is unchanged; hence by [L1]. Summing over the bijection gives . Since is the trivial permutation group, ; telescoping gives , which is after the index shift.
Type . By [L2] the parabolic with is the model, and [L7] identifies its intrinsic length with the ambient length; [L4] gives , the sum running over the orbit of [L3]; that orbit consists of one point at each distance , so the sum is . Hence , and telescoping from gives .
Type . For , by [L2] the parabolic obtained by deleting the terminal node is the model, and [L7] identifies its intrinsic length with ambient length; [L4] gives ; by [L3] the distances occurring are , the value twice and , so . Hence ; using the identity with in the induction step, and the base supplied by [L5], telescoping gives .
Small cases and evaluations. and follow from step 1.1; , a symmetric unimodal inversion-count sequence, and follow by expanding. By [L5], and , while the recursion of step 1.3 gives ; expanding, , which is palindromic of degree and has coefficient sum . Finally, evaluating the products of steps 1.1-1.3 at , where , gives , and . The insertion bijection of step 1.1 also gives by induction from . A signed permutation is specified by a permutation and independent signs, so ; for an even signed permutation the first signs determine the last, so . These are the group orders by [L2]. No invariant degrees or Choice are used.
The A2 = S3 case: Steinberg inclusion-exclusion, degree product, and reciprocity
Example
Let in its type Coxeter presentation with (The finite symmetric group , one-line notation, and cycle notation, Inversions, inversion number, the sign , and even and odd permutations). The library defines on ; the order-preserving relabelling identifies it with permutations of and preserves inversion number, so we write the adjacent generators as and . This is the smallest-rank finite Coxeter system whose diagram is not a product of copies of . The example checks the finite instances of parabolic factorization, descent inclusion-exclusion and the Steinberg identity of Finite descent parabolics, parabolic factorization, the Steinberg inclusion-exclusion identity, and rational growth, and of the degree product and reciprocity of The Poincare polynomial as a product of q-integers of the basic degrees, with longest-element reciprocity. The lengths, descents and reciprocity calculations use no choice. The basic-degree comparison assumes the Axiom of Choice (The Axiom of Choice) through the degree-table supplier cited in (3).
(1) Lengths and Poincare polynomial. The six elements are (length ), (length ), (length ) and (length ); hence , with (The longest element as the opposition of the chamber, and longest elements of finite parabolics (1)(ii)).
(2) Descent classes and Steinberg identity. The right descent sets are , , , , , , so and, with the interval convention of Length generating series, descent-class series, spherical subsets, and the multivariate descent polynomial (2), the nine interval series are The parabolic series are and . The Steinberg identity (Finite descent parabolics, parabolic factorization, the Steinberg inclusion-exclusion identity, and rational growth (4)) reads which is a polynomial identity after clearing denominators, and equivalently .
(3) Degree product. The basic degrees of are and (A regular Coxeter eigenvector determines the basic degrees: the exponent-residue identification and the complete degree tables for all finite Coxeter types (2), under the stated AC premise), so , (The cardinality of a finite set) and .
(4) Reciprocity. : the coefficient sequence is palindromic, in agreement with The Poincare polynomial as a product of q-integers of the basic degrees, with longest-element reciprocity (2).
Facts & Assumptions
Given: The symmetric group in the type presentation with simple generators , its length function , the Axiom of Choice for the basic-degree comparison, the descent sets and the series , of Length generating series, descent-class series, spherical subsets, and the multivariate descent polynomial.
Under the order-preserving relabelling , the library's is identified with permutations of and its adjacent generators are and (The finite symmetric group , one-line notation, and cycle notation). Direct composition gives the distinct one-line forms , , , , , for , respectively. These are all one-line arrangements of three entries. In the type- presentation, canceling makes every word alternating, and the braid relation reduces every alternating word of length at least four to a shorter word; thus the six listed words exhaust the presentation. Their inversion numbers are (Inversions, inversion number, the sign , and even and odd permutations); their lengths have these same values, since the length-two words are distinct from the identity and generators, while is distinct from all words of length at most two.
The element has length and is longest by [L1]; the finite longest-element result gives and for all (The longest element as the opposition of the chamber, and longest elements of finite parabolics (1)(ii),(iii)). Thus .
uses the inclusive interval convention (Length generating series, descent-class series, spherical subsets, and the multivariate descent polynomial (2)).
The factorization theorem gives unique length-additive parabolic factorizations; for , and (Finite descent parabolics, parabolic factorization, the Steinberg inclusion-exclusion identity, and rational growth (2)).
Under the Axiom of Choice (The Axiom of Choice), the independently determined basic-degree table gives for (A regular Coxeter eigenvector determines the basic degrees: the exponent-residue identification and the complete degree tables for all finite Coxeter types (2)).
The longest-element bijection and the length complementation in [L2] give for finite , by summing over ; this is the choice-free argument of The Poincare polynomial as a product of q-integers of the basic degrees, with longest-element reciprocity (2), Proof 1.2.
and , so the classical product gives (Classical Poincare products for A, B, D and I2(m) from the permutation and signed-permutation models (3)).
The cardinality of the six-element set is (The cardinality of a finite set).
Verification
Length and parabolic factorization. By [L1], the six elements exhaust and have lengths ; hence and by [L2]. For , [L5] gives and , so the length-additive factorization gives , agreeing with the direct enumeration.
Descent intervals. Reading right descents from the length table gives , , , , and . Thus the exact-descent series for are respectively . Summing these four classes over each inclusive interval gives , , , , and for , exactly the nine cases in the statement.
Inclusion-exclusion. Apply [L6] to the nine pairs . The needed quotients are , (the elements whose right descents omit ), and . The six symmetry classes of pairs yield, respectively, , for , , , , and . This verifies every interval value in the statement directly from the formula.
Steinberg identity. Substituting , and into the finite identity of [L7] gives . Clearing the denominator yields ; both sides equal . The reciprocity of [L9] gives , so this is also the rational-function identity displayed in the statement.
Degree product and reciprocity. By [L8], the basic degrees of are and , and by [L10] the classical product is ; this agrees with step 1.1. At , by [L11]. Also by step 1.1, and , so the coefficient sequence is palindromic as in [L9].
Infinite dihedral growth, the infinite Steinberg identity, and the failure of polynomial reciprocity
Example
Let be the infinite dihedral Coxeter group, so and (Coxeter matrices, the presented Coxeter group, reduced words, length, and standard parabolic subgroups, The rank-two block computation, exact dihedral orders, the signed reflection action, and ambient reducedness (3)(c),(4)). Then:
(1) Growth series. Every word reduces by canceling adjacent equal generators to an alternating word. For each , the two alternating words and have length ; for each , and have length . These words are reduced by The rank-two block computation, exact dihedral orders, the signed reflection action, and ambient reducedness (7), with the generators interchanged for the words beginning with ; their pairwise distinctness is proved in step 1.1 below using the infinite order of from (4). They exhaust the nonidentity elements, so has one element of length and exactly two of each length : The coefficientwise identity follows from the displayed coefficients by the Cauchy product (Formal power series over a commutative ring and the coefficient-extraction functional ), and has unit constant term (A formal power series is a unit exactly when its constant coefficient is a unit, Rational formal power series, proper presentations and reduced denominators).
(2) The infinite Steinberg identity. The spherical subsets are , with , , and ; the infinite case of Finite descent parabolics, parabolic factorization, the Steinberg inclusion-exclusion identity, and rational growth (4) reads i.e. , as the formal inverse of requires (A formal power series is a unit exactly when its constant coefficient is a unit).
(3) Failure of polynomial reciprocity. The reduced words in (1) have unbounded lengths, so has no longest element; the infinite-case result of Finite descent parabolics, parabolic factorization, the Steinberg inclusion-exclusion identity, and rational growth (1) also gives that no element has all descents. For comparison, the finite-case pattern is the reciprocity of The Poincare polynomial as a product of q-integers of the basic degrees, with longest-element reciprocity (2). Here the rational function satisfies Thus, for every , ; equality with would force in , which is impossible. The substitution is interpreted in the rational-function field, not as an element of . More generally, an infinite Coxeter group with finite has unbounded length (there are only finitely many words of bounded length), so its growth series is not a polynomial and the finite reciprocity theorem does not apply as a polynomial statement. This alone makes no assertion about rational-function reciprocity for other infinite Coxeter systems.
(4) Rank-one comparison. The parabolic has . The infinite growth series is not a finite product for any finite , since every such product is a polynomial while has infinitely many nonzero coefficients. With the conventions of the companion page, this same group is denoted .
Facts & Assumptions
Given: The presented group , the Coxeter matrix with on , its length function , and the series of Length generating series, descent-class series, spherical subsets, and the multivariate descent polynomial.
For the rank-two product has infinite order in and the elements are distinct involutions (The rank-two block computation, exact dihedral orders, the signed reflection action, and ambient reducedness (3)(c),(4)).
Every alternating word beginning with either or of length is reduced in the ambient group (apply the source also with interchanged) (The rank-two block computation, exact dihedral orders, the signed reflection action, and ambient reducedness (7)).
The defining relators allow adjacent equal letters to be canceled without changing the represented element; is the minimum length of a generator word representing (Coxeter matrices, the presented Coxeter group, reduced words, length, and standard parabolic subgroups).
For every the series is well-defined (Length generating series, descent-class series, spherical subsets, and the multivariate descent polynomial (1)); a series is rational if for polynomials with a unit (Rational formal power series, proper presentations and reduced denominators).
For infinite , no element has all descents (Finite descent parabolics, parabolic factorization, the Steinberg inclusion-exclusion identity, and rational growth (1)).
The finite-case pattern is with (The Poincare polynomial as a product of q-integers of the basic degrees, with longest-element reciprocity (2)); its longest-element reciprocity argument is choice-free and its basic-degree branch is not used here.
The Cauchy product is defined coefficientwise by finite sums, and a series with unit constant term has a unique formal inverse; here and have constant term (Formal power series over a commutative ring and the coefficient-extraction functional , A formal power series is a unit exactly when its constant coefficient is a unit).
Verification
Every word in can be shortened by canceling an adjacent pair or by [L3], so a shortest word is alternating. For each the two alternating words of length are , and for each the two of length are . They are reduced by [L2], so unequal lengths give unequal elements. Put , so and . At the same even length , equality of the two words would give and hence ; at the same odd length , equality would give and hence . Both contradict the infinite order in [L1]. Equivalently, the presentation admits the parity homomorphism sending both generators to , because each defining relator has even length, so even and odd words cannot coincide. Every element has one of these forms or is the identity. Thus and for every . By the Cauchy-product definition, : its coefficients are at degree , at degree , and at every degree . Since has unit constant term, in , and [L4] makes this a rational series.
A subset is spherical exactly when is finite. The three proper parabolics , and are finite; is infinite by [L1]. Thus the spherical subsets are exactly , with and , while all four subsets occur in the Steinberg sum of [L6]. Substituting from step 1.1 gives , the infinite case of [L6].
In , . If for some , then ; since , this forces , impossible in . Thus no such exists. The alternating words in step 1.1 have unbounded lengths, so has no longest element; the infinite-case statement of [L7] also says no element has all descents.
The series and are inverse to each other in : their product is , and both have constant term , so by [L9] as a formal inverse. This is the identity verified in step 2.1 and shows that is rational while has infinitely many positive coefficients.
The rank-one parabolic has . The reduced forms in step 1.1 with no right -descent are the identity and the alternating words ending in , exactly one of each length; hence . By [L5], , which is not a polynomial. No finite has , since the left side is a polynomial and the right has infinitely many nonzero coefficients. The same group is denoted .