Alphabeta Math
Pipeline-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.

✓ 14 results · all verified · 14 also independently AI-judged
Every result on this page is machine-checked by a proof checker and read in full and owner-audited; the judge is an additional, independent cross-model AI review of the proofs; all 14 also cleared it.

Specht Modules and the Irreducibles of the Symmetric Group

1 · Prerequisites

2 · Summary

This page constructs the Specht module Sλ inside the tabloid permutation module by applying the signed column antisymmetrizer to tabloids. It establishes covariance, the column-cancellation and dominance properties, and the James submodule theorem. Over C, the invariant tabloid pairing gives nondegeneracy and irreducibility; together with the class count for Sn, this identifies the Specht modules as a complete irredundant list of complex irreducibles.

The final results order tabloids and use adjacent-column Garnir relations to straighten polytabloids. Standard polytabloids then give an explicit basis of each Specht module, including the empty-shape convention.

3 · Logical flowchart

4 · Definitions, theorems and proofs

DefinitionDefinition: Literature-sourcedProof: Not applicablejudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Column antisymmetrizers, polytabloids, and Specht modules

Definition

Let n≥0, let λ⊢n, and let t be a λ-tableau (Partitions, English diagrams, and conjugation, Tableaux and standard tableaux). Use the left action of Sn on tableaux, tabloids, and Mλ (Young subgroups, tabloids, and permutation modules). For each γ∈Ct, let sgn⁡(γ) be its inversion sign in Sn. The order-preserving relabelling ιn:{1,…,n}→{0,…,n−1}, i↦i−1, carries each inversion pair (i,j) bijectively to (i−1,j−1); thus this sign is exactly the published inversion sign on the finite ordinal (Inversions, inversion number, the sign sgn⁡(σ)=(−1)inv⁡(σ), and even and odd permutations). For n=0, both groups are trivial and the sign is 1.

The column antisymmetrizer, polytabloid, and Specht space are κt:=∑γ∈Ctsgn⁡(γ)γ∈C[Sn],et:=κt⋅{t}∈Mλ,Sλ:=span⁡C{es:s is a λ-tableau}. The elements of the sum are in the finite subgroup Ct, and each acts by the declared left action on the tabloid basis, so these are well-defined finite expressions.

For every tableau, Ct∩Rt={1}: a permutation in both stabilizers preserves the row and column of each entry, and each row-column intersection contains at most one node. Thus γ{t} are distinct as γ ranges over Ct, and the coefficient of {t} in et is 1. In particular, et≠0. When n=0, the empty tableau has Ct={1}, so κt=1, et={∅}, and S∅=C.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Polytabloid covariance and the column sign rule

Statement

For every λ⊢n, n≥0, λ-tableau t, and σ∈Sn, one has κσ⋅t=σκtσ−1,eσ⋅t=σ⋅et. For every γ∈Ct, γ⋅et=sgn⁡(γ)et. Consequently Sλ is an Sn-submodule of Mλ and is generated by any one et.

Facts & Assumptions

Given: n≥0, λ⊢n, a λ-tableau t, σ∈Sn, and γ∈Ct.

[F1]

The column antisymmetrizer is the finite group-algebra sum κt=∑c∈Ctsgn⁡(c)c (Column antisymmetrizers, polytabloids, and Specht modules).

[F2]

The polytabloid is et=κt⋅{t} (Column antisymmetrizers, polytabloids, and Specht modules).

[F3]

Sλ is the complex span of all es for λ-tableaux s (Column antisymmetrizers, polytabloids, and Specht modules).

[F4]

Ct is the direct product of the symmetric groups on its pairwise disjoint column sets (Row and column stabilizers).

[F5]

Cσ⋅t=σCtσ−1 (Tableau stabilizers transform by conjugation).

[F7]

A λ-tableau is a bijection from [λ] to {1,…,n} (Tableaux and standard tableaux).

[F8]

The tabloid module has the tabloids as basis and carries the linear extension of their left Sn-action (Young subgroups, tabloids, and permutation modules).

Proof

technique · direct
1.1givenF1F4F6algebra

Write Bj for the set of labels in column j and Hj=S(Bj). By [F4], the Hj have pairwise disjoint supports and every c∈Ct has a unique factorization c=c1⋯cr with cj∈Hj. The product ∏j=1rκBj expands over these tuples, and by [F6] the coefficient at c1⋯cr is ∏jsgn⁡(cj)=sgn⁡(c). Hence κt=∏j=1rκBj, where κBj=∑h∈Hjsgn⁡(h)h. If n=0, both sides are the empty product 1.

1.2givenF6algebra

For every c∈Ct, [F6] gives sgn⁡(σcσ−1)=sgn⁡(σ)sgn⁡(c)sgn⁡(σ−1)=sgn⁡(c), since the homomorphism property makes sgn⁡(σ−1)=sgn⁡(σ)−1.

1.3givenF1F2F6algebra

For γ∈Ct, reindex [F1] by d=γc. Since sgn⁡(c)=sgn⁡(γ−1d)=sgn⁡(γ)sgn⁡(d), this gives γκt=sgn⁡(γ)κt. Applying both sides to {t} and using [F2] proves γ⋅et=sgn⁡(γ)et.

2.1givenF1F5step 1.2algebra

By [F5], conjugation by σ bijects Ct with Cσ⋅t. Reindexing [F1] by d=σcσ−1 and using step 1.2 gives κσ⋅t=σκtσ−1.

3.1givenF2F8step 2.1algebra

From [F2], step 2.1, and the left module action [F8], eσ⋅t=κσ⋅t⋅{σ⋅t}=σκtσ−1⋅(σ⋅{t})=σ⋅(κt⋅{t})=σ⋅et.

4.1givenF3F8step 3.1

Every spanning vector of Sλ has the form es by [F3], and step 3.1 sends it under σ to eσ⋅s, which is again a spanning vector. Thus Sλ is an Sn-submodule of Mλ.

5.1givenF3F7step 3.1algebra∎

Fix any λ-tableau t and let s be any other one. By [F7], the rule σ(t(i,j))=s(i,j) defines a unique permutation σ∈Sn; for n=0 it is the identity. Then s=σ⋅t and step 3.1 gives es=σ⋅et. Therefore the orbit of et spans all the generators of Sλ, so et generates Sλ as an Sn-module. This holds for every choice of t.

DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Invariant Hermitian product on a tabloid module

Definition

Let λ⊢n and write vectors in the finite tabloid basis of Mλ as x=∑TaTT and y=∑TbTT, where T ranges over the λ-tabloids (Young subgroups, tabloids, and permutation modules). Give this complex permutation space the Hermitian product ⟨x,y⟩:=∑TaT‾bT, conjugate-linear in the first argument and linear in the second. This is the Hermitian version of the tabloid-basis form: Chan defines the symmetric bilinear version on a permutation basis (Chapter 9, Definition 9.1 and Remark 9.2, printed pp. 31–32), while the complex Hermitian form used here is specified explicitly.

The tabloid basis is orthonormal. Also ⟨x,x⟩=∑T∣aT∣2, which is positive for every nonzero x, so the form is positive definite. Each σ∈Sn permutes the tabloid basis, hence ⟨σx,σy⟩=⟨x,y⟩. Thus the action is unitary and its adjoint is σ−1. Extend the group-algebra adjoint conjugate-linearly; since sgn⁡(γ)∈{1,−1} and γ↦γ−1 permutes Ct, κt∗=∑γ∈Ctsgn⁡(γ)γ−1=κt. Every column antisymmetrizer is therefore self-adjoint.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Column collision cancels antisymmetrization

Statement

Let t have shape λ and let {s} be a μ-tabloid of the same n. If two entries a,b lie in one row of s and in one column of t, then κt⋅{s}=0.

Facts & Assumptions

Given: An integer n≥0, partitions λ,μ⊢n, a λ-tableau t, a μ-tabloid {s}, and distinct labels a,b lying in one row of s and one column of t.

[F1]

The row stabilizer preserves each row set, and the column stabilizer preserves each column set (Row and column stabilizers).

[F2]

A tabloid records row sets, so the order of entries within each row is forgotten (Young subgroups, tabloids, and permutation modules).

[F3]

The column antisymmetrizer is κt=∑γ∈Ctsgn⁡(γ)γ (Column antisymmetrizers, polytabloids, and Specht modules).

[F6]

The library's 1,…,n sign convention is transported by the canonical relabelling from the finite-ordinal convention (Column antisymmetrizers, polytabloids, and Specht modules).

Proof

technique · direct
1.1givenF1F2construct

Put τ=(a b). Since a,b are in the same column of t, τ preserves every column set and lies in Ct by [F1]. Since they are in the same row of s, it preserves every row set of s; therefore τ⋅{s}={s} by [F1,F2].

1.2givenF3F4F5F6constructalgebra

The labels a,b are distinct, so n≥2. The canonical relabelling in [F6] preserves order and inversion number. If a<b, the one-line permutation for τ=(a b) has one inversion from (a,b) and two inversions for each intermediate label, so its inversion number is 1+2(b−a−1)=2(b−a)−1, which is odd. Thus sgn⁡(τ)=−1 by [F5]. Order permutations by one-line notation and take the least element in each right coset g⟨τ⟩ of Ct. These canonical representatives partition Ct into pairs g,gτ, whence κt=∑g(sgn⁡(g)g+sgn⁡(gτ)gτ)=∑gsgn⁡(g)g(1−τ) by [F3,F4]. The least representative is uniquely defined in each finite coset, so no choice principle is used.

2.1step 1.1step 1.2algebra∎

Applying the last expression to {s}, every term is zero because (1−τ)⋅{s}={s}−{s}=0 by step 1.1. Thus κt⋅{s}=0, as claimed.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Nonzero antisymmetrizer image detects dominance

Statement

For every n≥0, partitions λ,μ⊢n, and λ-tableau t, if κtMμ≠0, then λ dominates μ in the published order λ⊵μ.

Facts & Assumptions

Given: n≥0, λ,μ⊢n, a λ-tableau t, and the hypothesis κtMμ≠0.

[F1]

The μ-tabloids form a basis of Mμ, with the linear extension of the left action of Sn (Young subgroups, tabloids, and permutation modules).

[F2]

The column antisymmetrizer is the group-algebra element κt=∑γ∈Ctsgn⁡(γ)γ (Column antisymmetrizers, polytabloids, and Specht modules).

[F3]

If two entries in one row of s lie in one column of t, then κt⋅{s}=0 (Column collision cancels antisymmetrization).

[F4]

If every row of s meets each column of t in at most one entry, then for tableaux of shapes λ and μ one has λ⊵μ (Basic row-column incidence lemma).

[F5]

The notation λ⊵μ means that every prefix sum of λ is at least the corresponding prefix sum of μ, with both partitions padded by zeros (Dominance order on partitions).

Proof

technique · direct
1.1givenF1F2construct

By [F1,F2], κt acts linearly on Mμ. If it killed every μ-tabloid, it would kill their span Mμ, contrary to the hypothesis; therefore some μ-tabloid {s} satisfies κt⋅{s}≠0.

2.1step 1.1givenF3F4F5construct∎

By the contrapositive of [F3], no row of s contains two entries from one column of t. Thus the basic combinatorial lemma [F4] applies to these tableaux and gives λ⊵μ; by [F5] this is exactly the published dominance order in the statement.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

The antisymmetrizer image in its own tabloid module is one-dimensional

Statement

For every n≥0, partition λ⊢n, and λ-tableau t, the image of the column antisymmetrizer on the tabloid module is exactly the nonzero line

κtMλ=Cet,et≠0.

Facts & Assumptions

Given: n≥0, λ⊢n, and a λ-tableau t.

[F1]

The column antisymmetrizer is κt=∑γ∈Ctsgn⁡(γ)γ (Column antisymmetrizers, polytabloids, and Specht modules).

[F8]

The polytabloid is et=κt⋅{t} (Column antisymmetrizers, polytabloids, and Specht modules).

[F2]

The coefficient of {t} in et is 1 (Column antisymmetrizers, polytabloids, and Specht modules).

[F3]

The λ-tabloids form a basis of Mλ, and its Sn action extends linearly from the left action on tabloids (Young subgroups, tabloids, and permutation modules).

[F4]

The stabilizer of the tabloid {s} is Rs (Young subgroups, tabloids, and permutation modules).

[F5]

If a row of a tabloid contains two entries from one column of t, then κt sends that tabloid to zero (Column collision cancels antisymmetrization).

[F6]

If every row of a tableau s meets every column of t in at most one entry and s,t have shape λ, then there are ρ∈Rs and γ∈Ct with ρ⋅s=γ⋅t (Basic row-column incidence lemma).

[F7]

The sign function is a group homomorphism to {+1,−1} (The sign is a homomorphism Sn→{+1,−1}, surjective exactly when n≥2).

Proof

technique · direct
1.1givenF2

The coefficient of {t} in et is 1 by [F2], so et≠0.

1.2givenF4F5F6

Let {s} be any basis tabloid. If κt⋅{s}=0, its image already lies in Cet. Otherwise, [F5] implies that each row of s meets each column of t in at most one entry, and [F6] gives ρ∈Rs and γ∈Ct with ρ⋅s=γ⋅t; since ρ stabilizes {s} by [F4], this yields {s}=γ⋅{t}.

2.1step 1.2F1F3F7algebra

For any γ∈Ct, reindex the defining sum by d=cγ to obtain κtγ=∑c∈Ctsgn⁡(c)cγ=∑d∈Ctsgn⁡(dγ−1)d=sgn⁡(γ)κt, using [F1,F7] and sgn⁡(γ−1)=sgn⁡(γ); applying this to the tabloid equality of step 1.2 gives κt⋅{s}=sgn⁡(γ)et in its nonzero case. Thus every basis tabloid maps into Cet, and linearity with [F3] gives κtMλ⊆Cet.

3.1step 1.1step 2.1F8

Since et=κt⋅{t} by [F8], the vector et belongs to the image κtMλ; step 1.1 makes its span nonzero, so step 2.1 gives κtMλ=Cet and proves the statement.

∎

TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

James's submodule theorem over the complex numbers

Statement

For every n≥0, partition λ⊢n, and Sn-submodule U⊆Mλ, either Sλ⊆UorU⊆(Sλ)⊥, where orthogonality is for the invariant positive definite Hermitian tabloid product.

Facts & Assumptions

Given: n≥0, λ⊢n, and an Sn-submodule U⊆Mλ.

[F1]

The tabloids form a basis of Mλ, and the Sn-action extends linearly from the tabloid action (Young subgroups, tabloids, and permutation modules).

[F2]

An Sn-submodule is a linear subspace stable under each element of Sn (Subrepresentations, direct sums of representations, and irreducibility).

[F3]

The column antisymmetrizer, polytabloid, and Specht space are

κt=∑γ∈Ctsgn⁡(γ)γ,et=κt⋅{t},Sλ=span⁡C{es:s is a λ-tableau}

for each λ-tableau t (Column antisymmetrizers, polytabloids, and Specht modules).

[F4]

For every t, κtMλ=Cet and et≠0 (The antisymmetrizer image in its own tabloid module is one-dimensional).

[F5]

The Specht space is generated as an Sn-module by any one polytabloid et (Polytabloid covariance and the column sign rule).

[F6]

The tabloid product is conjugate-linear in its first argument and linear in its second (Invariant Hermitian product on a tabloid module).

[F7]

Each κt is self-adjoint for the tabloid product (Invariant Hermitian product on a tabloid module).

[F8]

This Hermitian product is positive definite: if x≠0, then ⟨x,x⟩>0 (Invariant Hermitian product on a tabloid module).

[F9]

The orthogonal complement is (Sλ)⊥={v:⟨v,s⟩=0 for every s∈Sλ} (Orthogonality and the orthogonal complement).

[F10]

Every shape has a canonical standard row-filled tableau t0; for n=0 it is the empty tableau (Young subgroups, tabloids, and permutation modules).

No form of the Axiom of Choice is used. The first branch uses only witnesses to one existential statement, and all group-algebra sums are finite.

Proof

technique · direct
1.1givenF1F2F3F4

If κtu≠0 for some u∈U and λ-tableau t, then [F4] gives κtu=cet with c≠0; by [F1]-[F3] the finite group-algebra sum κtu lies in U, so division gives et∈U.

1.2givenF3F7

Otherwise κtu=0 for every u∈U and every t; self-adjointness in [F7] and et=κt⋅{t} from [F3] give ⟨u,et⟩=⟨u,κt⋅{t}⟩=⟨κtu,{t}⟩=0.

2.1givenF2F5step 1.1

By [F5], the polytabloid from step 1.1 generates Sλ under Sn; stability of U from [F2] and step 1.1 therefore give Sλ⊆U.

2.2givenF3F6F9step 1.2

Since the et span Sλ by [F3] and the product is linear in its second argument by [F6], step 1.2 gives ⟨u,s⟩=0 for every u∈U and s∈Sλ; by [F9], U⊆(Sλ)⊥.

3.1givenF3F4F8F9F10step 2.1step 2.2∎

The cases “some κtu≠0” and “all κtu=0” are exhaustive; if both conclusions held, Sλ⊆U⊆(Sλ)⊥, so the nonzero canonical et0∈Sλ from [F3], [F4], [F10] would satisfy ⟨et0,et0⟩=0 by [F9], contradicting [F8]. Thus exactly one alternative holds.

LemmaStatement: AI-adaptedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Complex Specht modules have nondegenerate Hermitian self-pairing

Statement

For every λ⊢n, Sλ is nonzero and Sλ∩(Sλ)⊥={0} for the positive definite invariant Hermitian tabloid product.

Facts & Assumptions

Given: n≥0 and λ⊢n.

[F1]

The tabloid-basis Hermitian product is positive definite: ⟨x,x⟩=∑T∣aT∣2>0 whenever x=∑TaTT≠0 (Invariant Hermitian product on a tabloid module).

[F2]

Sλ is the complex span of the polytabloids es (Column antisymmetrizers, polytabloids, and Specht modules).

[F3]

For every tableau t, the coefficient of {t} in et is 1 (Column antisymmetrizers, polytabloids, and Specht modules).

[F4]

For every partition, the canonical row-filled λ-tableau exists (Young subgroups, tabloids, and permutation modules).

[F5]

The orthogonal complement is defined by vanishing of the inner product against every vector of the subspace (Orthogonality and the orthogonal complement).

Proof

technique · direct
1.1givenF2F3F4algebra

If λ=∅, take its empty tableau; otherwise take the canonical row-filled tableau t0 from [F4]. By [F2], et0∈Sλ, and by [F3] its coefficient at {t0} is 1. Thus et0≠0 and Sλ≠0.

2.1givenF1F2F5algebra∎

Let v∈Sλ∩(Sλ)⊥. By [F5], ⟨v,w⟩=0 for every w∈Sλ; taking w=v gives ⟨v,v⟩=0. Positive definiteness [F1] implies v=0. The zero vector belongs to both spaces by [F2] and [F5], so Sλ∩(Sλ)⊥={0}.

TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Complex Specht modules are irreducible

Statement

For every n≥0 and λ⊢n, the nonzero complex Sn-representation Sλ is irreducible.

Facts & Assumptions

Given: n≥0, λ⊢n, and a nonzero Sn-subrepresentation U⊆Sλ.

[F1]

Sλ is the complex span of its polytabloids, which lie in the tabloid module Mλ (Column antisymmetrizers, polytabloids, and Specht modules).

[F2]

Mλ is a finite-dimensional complex representation of Sn (Young subgroups, tabloids, and permutation modules).

[F3]

Sλ is an Sn-subrepresentation of Mλ (Polytabloid covariance and the column sign rule).

[F4]

For every Sn-submodule V⊆Mλ, either Sλ⊆V or V⊆(Sλ)⊥ (James's submodule theorem over the complex numbers).

[F5]

Sλ≠0 and Sλ∩(Sλ)⊥={0} (Complex Specht modules have nondegenerate Hermitian self-pairing).

[F6]

A subrepresentation is a linear subspace stable under every group element (Subrepresentations, direct sums of representations, and irreducibility).

[F7]

A representation is irreducible when it is nonzero and its only subrepresentations are 0 and the whole representation (Subrepresentations, direct sums of representations, and irreducibility).

No form of the Axiom of Choice is used.

Proof

technique · direct
1.1givenF1F2F3F4F6

By [F1]-[F3], Sλ is a finite-dimensional subrepresentation of Mλ. Since the given U is a subrepresentation of Sλ, [F6] makes U stable under every σ∈Sn also as a subspace of Mλ. Therefore [F4] applies to U.

2.1givenF4F5step 1.1

If the second branch of [F4] holds, then the given U⊆Sλ also satisfies U⊆(Sλ)⊥. By [F5], this forces U⊆{0}, contrary to the nonzero hypothesis. Thus the first branch gives Sλ⊆U.

3.1givenstep 2.1

The given inclusion U⊆Sλ and step 2.1 imply U=Sλ.

4.1givenF5F7step 3.1∎

By [F5], Sλ is nonzero, and step 3.1 shows that every nonzero subrepresentation equals Sλ. Thus [F7] gives that Sλ is irreducible.

TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Homomorphisms from Specht to Young permutation modules obey dominance

Statement

For λ,μ⊢n over C, a nonzero Sn-map ϕ:Sλ→Mμ implies λ⊵μ. At μ=λ, every Sn-map Sλ→Mλ is a scalar multiple of the inclusion; equivalently, Hom⁡C[Sn](Sλ,Mλ)≅C.

Facts & Assumptions

Given: n≥0, partitions λ,μ⊢n, and a complex C[Sn]-module homomorphism ϕ:Sλ→Mμ.

[F2]

Each Mν is a finite-dimensional complex representation of Sn (Young subgroups, tabloids, and permutation modules).

[F3]

The ν-tabloids form a basis of Mν, with the linear extension of the left Sn-action (Young subgroups, tabloids, and permutation modules).

[F4]

For a λ-tableau t, κt=∑γ∈Ctsgn⁡(γ)γ, et=κt⋅{t}, and Sλ is the complex span of the es (Column antisymmetrizers, polytabloids, and Specht modules).

[F5]

Every et is nonzero, since its coefficient at {t} is 1 (Column antisymmetrizers, polytabloids, and Specht modules).

[F6]

Sλ is an Sn-subrepresentation generated by any one et (Polytabloid covariance and the column sign rule).

[F7]

A subrepresentation is a linear subspace stable under each group element (Subrepresentations, direct sums of representations, and irreducibility).

[F8]

For a finite group over a field whose characteristic does not divide the group order, every subrepresentation has a complementary subrepresentation; this applies over C (Maschke's theorem for finite groups over fields whose characteristic does not divide ∣G∣).

[F9]

The space Hom⁡Sn(V,W) consists of complex-linear Sn-equivariant maps and is a subspace of the complex vector space of linear maps (Intertwiners, the spaces Hom⁡G(V,W) and End⁡G(V), equivalent representations, and faithful representations).

[F10]

A complex-linear map is Sn-equivariant exactly when it is a C[Sn]-module homomorphism (For a commutative ring R, R-linear G-actions are exactly the compatible left R[G]-module structures).

[F11]

If κtMμ≠0, then λ⊵μ (Nonzero antisymmetrizer image detects dominance).

[F12]

For every λ-tableau t, κtMλ=Cet (The antisymmetrizer image in its own tabloid module is one-dimensional).

[F13]

λ⊵μ means every prefix sum of λ is at least the corresponding prefix sum of μ (Dominance order on partitions).

[F14]

Every shape has a canonical standard row-filled tableau t0 (Young subgroups, tabloids, and permutation modules).

[F15]

No form of the Axiom of Choice is used. The proof uses one complement supplied by Maschke's theorem and existential witnesses from spanning families; it does not choose from an arbitrary indexed family.

Proof

technique · direct
1.1givenF1F2F6F7F8F15

By [F1] and [F15], Sn is a finite group; by [F2], Mλ is a finite-dimensional complex representation; by [F6]-[F7], Sλ is a subrepresentation. Since char⁡C=0 does not divide ∣Sn∣, [F8] gives a subrepresentation T with Mλ=Sλ⊕T.

2.1givenF6F7step 1.1

Define p:Mλ→Sλ by p(s+t)=s for s∈Sλ,t∈T. The direct sum makes p a well-defined linear projection with p∣Sλ=id; since both summands are stable, p(σ(s+t))=σs=σp(s+t) for every σ∈Sn, so p is equivariant.

3.1givenF9F10step 2.1

Set Φ:=ϕ∘p:Mλ→Mμ. By [F9]-[F10] and step 2.1, Φ is a C[Sn]-module homomorphism, and Φ∣Sλ=ϕ.

4.1givenF3F4F10step 3.1

If ϕ≠0, some λ-tableau t has ϕ(et)≠0 because the polytabloids span Sλ by [F4]. The tabloid {t} is a basis vector by [F3]; then [F4] and step 3.1 give ϕ(et)=Φ(et)=Φ(κt⋅{t})=κt⋅Φ({t})≠0, so κtMμ≠0.

5.1givenF11F13step 4.1

If ϕ≠0, step 4.1 and [F11] imply λ⊵μ, which by [F13] is the stated dominance order; if ϕ=0, the nonzero-map implication is vacuous.

5.2givenF5F6F9F12step 4.1

Suppose μ=λ and ϕ≠0, and use the tableau from step 4.1. By [F12], ϕ(et)=κtΦ({t})∈Cet; since et≠0 by [F5], write ϕ(et)=cet with c≠0. For every σ∈Sn, equivariance gives ϕ(σet)=cσet; by [F6], this extends linearly to all of Sλ. Hence ϕ=cιλ, where ιλ:Sλ↪Mλ is inclusion.

6.1givenF4F5F6F9F10F14step 5.2∎

If ϕ=0, it is 0ιλ; step 5.2 covers every nonzero map when μ=λ. Each scalar multiple of inclusion is equivariant by [F6]. The canonical t0 from [F14] has et0∈Sλ by [F4] and et0≠0 by [F5], so ιλ is nonzero. Therefore c↦cιλ is a linear bijection from C to Hom⁡C[Sn](Sλ,Mλ).

CorollaryStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Distinct complex Specht modules are inequivalent

Statement

If λ,μ⊢n and Sλ≅Sμ as complex Sn-representations, then λ=μ.

Facts & Assumptions

Given: n≥0, partitions λ,μ⊢n, and an isomorphism f:Sλ→Sμ of complex Sn-representations.

[F1]

Each shape ν⊢n has a canonical standard row-filled tableau tν (Young subgroups, tabloids, and permutation modules).

[F2]

The Specht space Sν is the complex span of its polytabloids, each lying in the tabloid module Mν (Column antisymmetrizers, polytabloids, and Specht modules).

[F3]

For every tableau t, the coefficient of {t} in et is 1; in particular et≠0 (Column antisymmetrizers, polytabloids, and Specht modules).

[F4]

For each ν⊢n, Sν is an Sn-submodule of Mν (Polytabloid covariance and the column sign rule).

[F6]

A complex-linear map of Sn-representations is equivariant exactly when it is a C[Sn]-module homomorphism (For a commutative ring R, R-linear G-actions are exactly the compatible left R[G]-module structures).

[F7]

A nonzero C[Sn]-module map Sα→Mβ implies α⊵β (Homomorphisms from Specht to Young permutation modules obey dominance).

[F8]

The dominance relation on partitions of n is antisymmetric: λ⊵μ and μ⊵λ imply λ=μ (Dominance order on partitions).

No form of the Axiom of Choice is used. The proof uses only the canonical tableaux in [F1] and the given isomorphism.

Proof

technique · direct
1.1givenF1F2F3F4F5F6

Let tλ be the canonical tableau from [F1]. By [F2]-[F3], etλ≠0 in Sλ. Let ιμ:Sμ↪Mμ be inclusion and set ψ:=ιμ∘f. Since f and inclusion are injective, ψ(etλ)≠0. By [F4]-[F5], both maps are Sn-equivariant, so [F6] makes ψ a C[Sn]-module homomorphism.

1.2givenF1F2F3F4F5F6algebra

The inverse f−1 is equivariant: for y=f(x), surjectivity and equivariance of f give f−1(σy)=f−1(σf(x))=f−1(f(σx))=σx=σf−1(y) for every σ∈Sn. Let tμ be the canonical tableau from [F1] and let ιλ:Sλ↪Mλ be inclusion. By [F2]-[F3], etμ≠0 in Sμ; since f−1 and inclusion are injective, ψ′:=ιλ∘f−1:Sμ→Mλ is nonzero. By [F4]-[F6], it is a C[Sn]-module homomorphism.

2.1step 1.1F7

Apply [F7] to the nonzero map ψ:Sλ→Mμ from step 1.1; it gives λ⊵μ.

2.2step 1.2F7

Apply [F7] with α=μ and β=λ to ψ′ from step 1.2. It gives μ⊵λ.

3.1step 2.1step 2.2F8∎

Steps 2.1 and 2.2 give both dominance relations, so [F8] yields λ=μ.

TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Specht modules classify the complex irreducibles of Sn

Statement

For every n≥0, the modules {Sλ:λ⊢n} form a complete irredundant list, up to isomorphism, of finite-dimensional irreducible complex Sn-representations.

Facts & Assumptions

Given: n≥0. Let Pn be the set of partitions of n, Cn the set of conjugacy classes of Sn, and In the set of isomorphism classes of finite-dimensional irreducible complex Sn-representations.

[F1]

A partition is a finite weakly decreasing list of positive integers with sum n (Partitions, English diagrams, and conjugation).

[F3]

Sn is finite and ∣Sn∣=n! (The Lehmer code gives ∣Sn∣=n! again).

[F5]

R is an ordered field and, for every integer m≥1, its canonical natural m⋅1R is positive (The reals form a totally ordered field, Canonical naturals are positive and strictly increasing).

[F7]

The natural-to-integer embedding preserves order, and zero divides no positive integer (The naturals embed in the integers, Invertibility of a positive natural scalar in a field).

[F8]

C is algebraically closed (The complex numbers are algebraically closed).

[F9]

For a finite group G and algebraically closed field k with char⁡k∤∣G∣, the finite set of irreducible representation classes has cardinality equal to the finite set of conjugacy classes (If k is algebraically closed and char⁡k∤∣G∣, the number of irreducible representations of G equals the number of conjugacy classes).

[F10]

Conjugacy classes of Sn are in bijection with the tuples of nonnegative integers (c1,…,cn) satisfying ∑k=1nkck=n; when n=0 the unique empty tuple indexes the identity class (The conjugacy classes of Sn are indexed by the tuples (c1,…,cn) with ∑kck=n).

[F11]

Each Sλ is a nonzero irreducible complex representation of Sn (Complex Specht modules are irreducible).

[F12]

If Sλ≅Sμ for λ,μ⊢n, then λ=μ (Distinct complex Specht modules are inequivalent).

[F13]

A representation is finite-dimensional as specified in A finite-dimensional representation ρ:G→GL⁡(V) over a field, and its degree, is irreducible when it is nonzero and has no proper nonzero subrepresentation (Subrepresentations, direct sums of representations, and irreducibility), and two representations are equivalent exactly when an invertible intertwiner exists (Intertwiners, the spaces Hom⁡G(V,W) and End⁡G(V), equivalent representations, and faithful representations).

[F14]

For finite sets, a bijection transports cardinality (The cardinality ∣A∣ of a finite set).

[F15]

If J⊆I and I is finite, then J is finite and ∣J∣=∣I∣ implies J=I (A subset of a finite set is finite, with ∣B∣≤∣A∣, and equality holds if and only if B=A).

[F16]

A map is injective when equal outputs force equal inputs, is surjective when its image is the codomain, and is bijective when it is both injective and surjective (Injection, surjection, bijection).

[F17]

For n=0 the only partition is the empty list (Partitions, English diagrams, and conjugation).

[F18]

Cardinality is defined for finite sets, and a finite set has cardinality only as a natural number (The cardinality ∣A∣ of a finite set).

[F19]

For n=0, the unique empty tuple indexes the identity class of S0 (The conjugacy classes of Sn are indexed by the tuples (c1,…,cn) with ∑kck=n).

[F20]

A finite set has cardinality zero exactly when it is empty (The cardinality ∣A∣ of a finite set).

No Axiom of Choice is used.

Proof

technique · counting
1.1F4F5F6

For each m≥1, [F5] gives m⋅1R>0. The embedding in [F4] sends this canonical real scalar to m⋅1C; if the latter were zero, injectivity would make the positive real scalar zero, a contradiction. Thus no positive integer multiple of 1C is zero, and [F6] yields char⁡C=0.

1.2F2F3F7F8F9F18F20

The identity permutation belongs to Sn, so [F3] makes Sn a nonempty finite group and [F20] gives ∣Sn∣>0. The natural-to-integer embedding in [F7] preserves positivity, so characteristic zero does not divide this order; [F8] supplies the algebraic-closure hypothesis. Applying [F9] shows that In and Cn are finite and ∣In∣=∣Cn∣.

1.3F1F10F17F19construct

For a tuple (c1,…,cn) from [F10], form the finite list containing ck copies of k for each k=n,n−1,…,1. Its entries are positive, weakly decreasing and sum to n, so [F1] makes it a partition. The inverse map sends a partition to the multiplicity ck of each part k. For n=0, [F17] and [F19] make both constructions the empty list/tuple. Thus the tuple set in [F10] is in bijection with Pn, and [F10] then gives Pn≈Cn.

1.4F11F12given

Define Φ:Pn→In by Φ(λ)=[Sλ]. Fact [F11] makes this a well-defined map into In, and [F12] makes it injective.

2.1F14step 1.2step 1.3

By [F14] and step 1.2, the bijection in step 1.3 transports finiteness and cardinality, so Pn is finite and ∣Pn∣=∣Cn∣=∣In∣.

3.1F15F16step 1.2step 2.1step 1.4

Let J:=Φ[Pn]⊆In. The map Φ is injective by step 1.4, and its corestriction to its image is surjective by the definition of J; [F16] therefore makes this corestriction a bijection. Hence ∣J∣=∣Pn∣=∣In∣ by step 2.1. Since In is finite by step 1.2, [F15] gives J=In. Thus Φ is surjective as well as injective.

4.1F11F12F13step 1.4step 3.1∎

The surjectivity in step 3.1 says every finite-dimensional irreducible complex Sn-representation is isomorphic to some Sλ; injectivity in step 1.4 says no two distinct partitions give isomorphic modules. These are exactly completeness and irredundancy, proving the statement.

DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Tabloid and column orders for Specht straightening

Definition

Fix a partition λ⊢n. If T is a λ-tabloid, let rT(a) be the row containing label a. For distinct tabloids T,U, let m be the largest label for which rT(m)≠rU(m). Define T>U⟺rT(m)>rU(m). This is the reverse lexicographic order on the row-index vectors (rT(n),…,rT(1)): any two distinct vectors have a largest differing coordinate, and the row numbers there are comparable. Thus it is a finite strict total order on the λ-tabloids.

A λ-tableau is column-standard when its entries strictly increase down each column. For a column-standard tableau t, let ct(a) be the column containing label a. Since entries within each column are sorted, the assignment a↦ct(a) determines t uniquely. For distinct column-standard tableaux t,u, let m be the largest label with ct(m)≠cu(m), and define

t≺u⟺ct(m)<cu(m).

This is a finite strict total order on the column-standard tableaux; in words, the largest label assigned to different columns is farther left in t than in u. For λ=∅, there is one tabloid and one tableau, and each is the sole element of its order.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Leading tabloid of a column-standard polytabloid

Statement

For a column-standard λ-tableau t, et has coefficient 1 at {t}, and every other tabloid in et is strictly below {t} in the fixed tabloid order. Consequently the standard polytabloids are linearly independent.

Facts & Assumptions

Given: A partition λ⊢n and a column-standard λ-tableau t.

[F1]

A tableau is column-standard when its entries strictly increase down each column (Tabloid and column orders for Specht straightening).

[F2]

The tabloid order compares the row of the largest label placed in different rows (Tabloid and column orders for Specht straightening).

[F3]

The polytabloid is the signed column sum et=∑γ∈Ctsgn⁡(γ)γ⋅{t} (Column antisymmetrizers, polytabloids, and Specht modules).

[F4]

A standard tableau has entries strictly increasing along rows and down columns (Tableaux and standard tableaux).

[F5]

Rt consists of the permutations preserving each row set of t (Row and column stabilizers).

[F6]

Ct consists of the permutations preserving each column set of t (Row and column stabilizers).

Proof

technique · direct
1.1givenF3F5algebra

If γ∈Ct and γ⋅{t}={t}, then [F5] gives γ∈Rt, so γ∈Ct∩Rt. A permutation in this intersection preserves both the row and column of every entry; each row-column intersection contains at most one node, so it fixes every label and is the identity. Thus the identity is the only term of et contributing to {t}, and its coefficient is sgn⁡(1)=1.

1.2givenF1F2F3F6algebra

Let γ∈Ct be nonidentity and let m be its largest moved label. Then γ−1(m)<m: the preimage differs from m, and if it were larger than m it would itself be a moved label larger than m. By [F6], γ−1(m) and m lie in the same column of t, so by [F1] the smaller label lies above m. Under the left action, γ⋅t places m in that higher node; every label larger than m is fixed by γ. Thus m is the largest label whose row changes, and [F2] gives {γ⋅t}<{t}. Every nonidentity term of et is therefore strictly below {t}.

2.1F2F4step 1.1step 1.2construct∎

Distinct standard tableaux have distinct tabloids: their entries are already increasing within each row by [F4], so each row set determines its row uniquely. In a nontrivial linear relation among standard polytabloids, choose the greatest leading tabloid among those with nonzero coefficient; the finite total order [F2] gives this element. By steps 1.1–1.2, its coefficient in the relation is exactly the nonzero coefficient of its own polytabloid, since every other participating leading tabloid is smaller and all its terms are smaller still. This contradicts the relation. Hence the standard polytabloids are linearly independent.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Adjacent-column Garnir relation over C

Statement

Let t be a λ-tableau. Let X lie among entries of column j and Y among entries of column j+1, with ∣X∣+∣Y∣>λj′. Put H=SX×SY, choose representatives T containing 1 for the left cosets gH in SX∪Y, and GX,Y:=∑g∈Tsgn⁡(g)g. Then GX,Yet=0 in Mλ over C.

Facts & Assumptions

Given: n≥0, λ⊢n, a λ-tableau t, adjacent columns j,j+1, subsets X,Y of their respective entries with ∣X∣+∣Y∣>λj′, and a left-coset transversal T for SX∪Y/H containing 1.

[F1]

The polytabloid is et=κt⋅{t}, where κt=∑γ∈Ctsgn⁡(γ)γ (Column antisymmetrizers, polytabloids, and Specht modules).

[F2]

For h∈Ct, h⋅et=sgn⁡(h)et (Polytabloid covariance and the column sign rule).

[F3]

If two entries lie in one row of a tabloid and in one column of a tableau, that tableau's column antisymmetrizer kills the tabloid (Column collision cancels antisymmetrization).

[F4]

Column j has height λj′, and these heights are weakly decreasing with j (Partitions, English diagrams, and conjugation).

[F5]

The column stabilizer preserves each column set (Row and column stabilizers).

[F6]

A tableau of shape ν⊢n is a bijection from [ν] to {1,…,n} (Tableaux and standard tableaux).

Proof

technique · direct
1.1givenF1F3F4F5F6algebra

Set Z=X∪Y and AZ=∑z∈SZsgn⁡(z)z, where SZ fixes labels outside Z. The columns are disjoint, so ∣Z∣=∣X∣+∣Y∣>λj′. If λj′=0, both X and Y are empty, contradicting this inequality; hence ∣Z∣≥2. For each γ∈Ct, every label in Z lies in one of the first λj′ rows of γ⋅{t}: its preimage under γ is in column j or j+1, and λj+1′≤λj′ by [F4]. Thus two labels of Z lie in one row of that tabloid. Form ν=(n−∣Z∣+1,1∣Z∣−1) and the ν-tableau u that places the labels of Z in increasing order down its first column and all remaining labels in increasing order along its first row; this is a tableau by [F6]. Its other columns are singletons, so Cu=SZ by [F5] and κu=AZ. Applying [F3] to u and each tabloid γ⋅{t} gives AZ(γ⋅{t})=0. Expanding et by [F1] now gives AZet=0.

2.1givenF1F2F5F7step 1.1algebra∎

Since H permutes labels within the two respective columns, H⊆Ct by [F5]. Define AH=∑h∈Hsgn⁡(h)h; [F2] gives AHet=∣H∣et. The left-coset decomposition SZ=⨆g∈TgH and multiplicativity [F7] give AZ=GX,YAH. Therefore step 1.1 yields 0=AZet=GX,YAHet=∣H∣GX,Yet. The positive integer ∣H∣ is nonzero in C, so division gives GX,Yet=0. The proof works for every supplied transversal T and makes no further choice.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Garnir straightening spans the complex Specht module

Statement

For every n≥0, partition λ⊢n, and λ-tableau t, the polytabloid et is a finite complex linear combination of standard λ-polytabloids. This is proved without using the later RSK identity.

Facts & Assumptions

Given: n≥0, λ⊢n, and a λ-tableau t.

[F1]

Column j has height λj′, and the column heights weakly decrease with j (Partitions, English diagrams, and conjugation).

[F2]

A tableau is standard exactly when its rows and columns strictly increase (Tableaux and standard tableaux).

[F3]

The left action is (σ⋅t)(i,j)=σ(t(i,j)) (Tableaux and standard tableaux).

[F4]

Ct consists of the permutations preserving each column set, so its elements act by permuting labels within columns (Row and column stabilizers).

[F5]
[F6]

Sλ is the complex span of all λ-polytabloids (Column antisymmetrizers, polytabloids, and Specht modules).

[F7]

Polytabloid covariance gives eσ⋅t=σ⋅et (Polytabloid covariance and the column sign rule).

[F8]

For γ∈Ct, γ⋅et=sgn⁡(γ)et (Polytabloid covariance and the column sign rule).

[F9]

A tableau is column-standard when its entries strictly increase down each column (Tabloid and column orders for Specht straightening).

[F10]

Column-standard tableaux have a finite strict total order, with s≺u exactly when the greatest label assigned to different columns is farther left in s than in u (Tabloid and column orders for Specht straightening).

[F11]

The adjacent-column Garnir relation accepts a supplied left-coset transversal containing the identity (Adjacent-column Garnir relation over C).

[F12]

If X,Y lie in adjacent columns and ∣X∣+∣Y∣>λj′, the corresponding Garnir sum annihilates et over C (Adjacent-column Garnir relation over C). For every such transversal T, (∑g∈Tsgn⁡(g)g)et=0

No Axiom of Choice (AC) is used. Column sorting, the inversion, and the Garnir representatives below are specified by unique rules on finite sets; there is no AC dependency to propagate.

Proof

technique · finite strong induction using Garnir straightening
1.1givenF3F4F5F7F8F9algebra

Given any λ-tableau t, sort the entries in each column increasingly to obtain the unique column-standard s with the same column sets. The rule π(t(i,j))=s(i,j) defines a unique π∈Ct with π⋅t=s, so by covariance and the column sign rule es=π⋅et=sgn⁡(π)et and hence et=sgn⁡(π)es. It remains to prove the claim for column-standard tableaux; when n=0, the unique empty tableau is already standard.

1.2givenF1F2F9algebra

Let s be column-standard but not standard. There is an adjacent row descent s(q,j)>s(q,j+1); take the lexicographically least such (q,j), put h=λj′, xr=s(r,j) for q≤r≤h, and ya=s(a,j+1) for 1≤a≤q. Because row q contains both boxes, q≤λj+1′≤h; column-standardness gives xq<⋯<xh, y1<⋯<yq, and xq>yq. Hence every x in X={xq,…,xh} exceeds every y in Y={y1,…,yq} and ∣X∣+∣Y∣=(h−q+1)+q=h+1>λj′.

2.1givenF7F11F12step 1.2algebra

Put Z=X∪Y and p=∣X∣. For each p-element subset A⊆Z, list X∖A={a1<⋯<ar} and A∖X={b1<⋯<br} and set gA=(a1 b1)⋯(ar br), with the empty product the identity. These are disjoint swaps and gA(X)=A. Since H=SX×SY preserves X, two elements of SZ lie in the same left coset gH exactly when their images of X agree; thus the gA form a canonical transversal and gX=1. Apply the Garnir relation and use covariance [F7] to rewrite its terms, isolating es=−∑A≠X, ∣A∣=psgn⁡(gA)egA⋅s.

3.1givenF3F4F7F8F9F10step 2.1algebra

For A≠X, the greatest element xA of X∖A is swapped with an element of Y and, under the left action, moves from column j of s to column j+1 of gA⋅s. Every other changed label is smaller: changed labels from Y are below every element of X, and xA is largest among the changed elements of X. Sort the columns of gA⋅s to obtain the unique column-standard uA with the same column sets; then xA remains in column j+1, so the finite column order gives s≺uA. The unique column permutation taking gA⋅s to uA, covariance, and the column sign rule give egA⋅s=±euA.

4.1givenF1F2F10step 1.2step 2.1step 3.1base

The base case is a greatest column-standard tableau. If it were nonstandard, step 1.2 would give nonempty X,Y, so 0<∣X∣<∣X∪Y∣ and step 2.1 has a nonidentity representative; step 3.1 would then construct a strictly later column-standard tableau. Thus the greatest tableau is standard, and its polytabloid already has the required form.

5.1givenF2F10step 2.1step 3.1step 4.1ih

For a column-standard s, assume as the induction hypothesis that every later u has eu equal to a finite complex linear combination of standard polytabloids. [ih] If s is standard the claim is immediate; otherwise step 2.1 expresses es as a finite sum of egA⋅s and step 3.1 rewrites each as ±euA with s≺uA. The induction hypothesis then proves the claim for s.

6.1

Finite reverse induction in the order of [F10] now proves the claim for every column-standard tableau; step 1.1 extends it to every tableau. Since Sλ is the span of all polytabloids by [F6], the standard polytabloids span Sλ. The empty shape is included, all sums are finite, and RSK is not used. [given, F6, F10, step 1.1, step 5.1, discharge-induction] □

TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30Open item page →

Standard polytabloids form a basis of a complex Specht module

Statement

For every n≥0 and partition λ⊢n, the family {et:t is a standard λ-tableau} is a C-basis of Sλ. In particular, dim⁡CSλ=fλ, including dim⁡CS∅=1.

Facts & Assumptions

Given: n≥0 and a partition λ⊢n.

[F1]

A λ-tableau is a bijection from its finite Young diagram to {1,…,n} (Tableaux and standard tableaux).

[F2]

A standard tableau has entries strictly increasing along rows and down columns (Tableaux and standard tableaux).

[F3]

The empty tableau is the unique standard tableau of shape ∅ (Tableaux and standard tableaux).

[F4]

fλ is the number of standard λ-tableaux (Tableaux and standard tableaux).

[F5]

A canonical standard row-filled λ-tableau t0 exists; for λ=∅ it is the empty tableau (Young subgroups, tabloids, and permutation modules).

[F6]

Two tabloids are equal exactly when their corresponding tableaux have the same row sets (Young subgroups, tabloids, and permutation modules).

[F7]

The tabloid order is a finite strict total order (Tabloid and column orders for Specht straightening).

[F8]

et∈Mλ for every λ-tableau t, and Sλ=span⁡C{et:t is a λ-tableau}; for the empty shape, et0={∅} and S∅=C (Column antisymmetrizers, polytabloids, and Specht modules).

[F9]

For a column-standard tableau, the coefficient of its own tabloid in its polytabloid is 1 and every other tabloid in it is strictly lower; the standard polytabloids are linearly independent (Leading tabloid of a column-standard polytabloid).

[F10]

Every polytabloid is a finite complex linear combination of standard polytabloids (Garnir straightening spans the complex Specht module).

[F11]

The span of a subset of a vector space is exactly the set of finite linear combinations of its elements (span⁡(S) is exactly the set of linear combinations of finite lists of elements of S, and span⁡(∅)={0V}).

[F13]
[F15]

The dimension of a vector space with a finite basis is the unique natural number equinumerous with that basis (Finite-dimensional vector space, and its dimension dim⁡FV; infinite-dimensional means having no finite basis).

No form of the Axiom of Choice (AC) is used. The tableau set is finite, the row-filled tableau is canonical, and Garnir straightening gives finite sums.

Proof

technique · direct
1.1givenF5F8

Let Eλ={et:t is a standard λ-tableau}. The canonical tableau t0 from [F5] makes the indexing set nonempty, and [F8] gives Eλ⊆Sλ.

1.2givenF2F6F7F9algebra

The family (et)t standard is linearly independent by [F9]. The increasing rows [F2] and the row-set characterization [F6] show distinct standard tableaux have distinct tabloids; by the total order [F7], one of {t},{u} is greater, with coefficient 1 in its own polytabloid by [F9] and coefficient 0 in the other, whose terms are below its smaller leading tabloid. Thus t↦et is injective.

2.1givenF8F10F11F12F13step 1.1

Put W=span⁡C(Eλ). Garnir straightening [F10] expresses every eu as a finite complex linear combination of members of Eλ, so [F11] gives eu∈W. Since W is a subspace [F12] containing all these generators, [F13] and [F8] give Sλ⊆W; conversely, step 1.1 gives Eλ⊆Sλ, so [F13] gives W⊆Sλ. Hence W=Sλ.

3.1givenF14step 1.2step 2.1

By [F14], steps 1.2 and 2.1 show that Eλ is a basis of Sλ.

4.1givenF1F3F4F8F10F15step 1.2step 3.1∎

By [F1], standard tableaux form a finite set, and by [F4] it has fλ members; step 1.2 makes t↦et injective, so ∣Eλ∣=fλ. Thus [F15] and step 3.1 give dim⁡CSλ=fλ. If λ=∅, [F3] gives the unique empty standard tableau and [F8] gives its nonzero polytabloid and S∅=C, hence the same singleton-basis argument gives dim⁡CS∅=1. All sums in [F10] are finite; no RSK identity or axiom of choice is used.

5 · Examples, counterexamples and false statements

None yet.

Sources