Scribd Logo
Upload

Welcome to Scribd!

  • 13 views

    n n λ λ n n 1

    Uploaded by

    mqtrinh
    Speaker: Syu Kato Title: A homological study of Green polynomials Date: February 22, 2012 This talk discusses Green polynomials, which relate representations of the symmetric group to Springer fibers. Green polynomials count the multiplicity of representations appearing in the homology of Springer fibers. The talk reinterprets orthogonality relations for Green polynomials using graded Euler-Poincaré pairings between projective modules and Springer fiber homology modules. This allows the generalization of Green polynomials to any finite Coxeter group.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    n n λ λ n n 1

    Uploaded by

    mqtrinh
    13 views3 pages
    Speaker: Syu Kato Title: A homological study of Green polynomials Date: February 22, 2012 This talk discusses Green polynomials, which relate representations of the symmetric group to Springer fibers. Green polynomials count the multiplicity of representations appearing in the homology of Springer fibers. The talk reinterprets orthogonality relations for Green polynomials using graded Euler-Poincaré pairings between projective modules and Springer fiber homology modules. This allows the generalization of Green polynomials to any finite Coxeter group.

    Original Title

    20120222-kato

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Speaker: Syu Kato Title: A homological study of Green polynomials Date: February 22, 2012 This talk discusses Green polynomials, which relate representations of the symmetric group to Springer fibers. Green polynomials count the multiplicity of representations appearing in the homology of Springer fibers. The talk reinterprets orthogonality relations for Green polynomials using graded Euler-Poincaré pairings between projective modules and Springer fiber homology modules. This allows the generalization of Green polynomials to any finite Coxeter group.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    13 views3 pages

    n n λ λ n n 1

    Uploaded by

    mqtrinh
    Speaker: Syu Kato Title: A homological study of Green polynomials Date: February 22, 2012 This talk discusses Green polynomials, which relate representations of the symmetric group to Springer fibers. Green polynomials count the multiplicity of representations appearing in the homology of Springer fibers. The talk reinterprets orthogonality relations for Green polynomials using graded Euler-Poincaré pairings between projective modules and Springer fiber homology modules. This allows the generalization of Green polynomials to any finite Coxeter group.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 3

    Speaker: Syu Kato

    Title: A homological study of Green polynomials


    February 22, 2012

    1 Introduction.
    Let Irr(Sn ) be the set of isomorphism classes of complex irreducible representations of Sn , which
    we index by Lλ where λ is a partition of n. The partitions λ also index the orbits Oλ in the
    nilpotent cone N ⊂ gln via Jordan normal form. Hence we might expect a natural correspondence
    between Irr(Sn ) and the nilpotent orbits, and this is given by the Springer correspondence.
    Let T ⊂ B ⊂ G = GL(n, C) be a maximal torus and Borel subgroup, and let t ⊂ b ⊂ g be
    their Lie algebras. Set n = [b, b]. Define N e = {(gB, x) ∈ G/B × g | x ∈ gng −1 }. There is a
    e → N , which is the Springer resolution. We also have a natural projection
    natural projection µ : N
    e
    pr : N → G/B =: B which is a vector bundle.
    Fix xλ ∈ Oλ and consider µ−1 (xλ ) ⊂ N e and it maps injectively to its image in B under pr.
    We denote the image by Bλ , which is the Springer fiber of xλ . Note that Bλ is well-defined up to
    G-conjugation.
    Theorem 1.1 (Borel). C[t] ∼ = H• (B; C) → H• (B; C) is surjective (here C[t]] = Sym(t∗ ) where
    G
    deg(t∗ ) = 2) and the kernel is hC[t]S
    + i.
    n

    Note that modding out by the above deal is the same as killing all trivial Sn -representations
    besides the one in degree 0. Furthermore, H• (B; C) inherits an action of Sn .
    Theorem 1.2 (Springer, Hotta–Springer, De Concini–Procesi). There is a natural map H• (B; C) →
    H• (Bλ ; C) is
    1. surjective (De Concini–Procesi)
    2. There is a presentation of the kernel which gives H• (Bλ ; C) the structure of Sn -representation.
    3. H2 dim Bλ (Bλ ; C) ∼
    = Lλ as Sn -modules.
    4. The assignment Oλ 7→ Lλ is a bijection (Springer correspondence)
    H• (Bλ ; C) gives characters of “unipotent representations”, which is a subset of the irreducible
    representations of GL(n, Fq ). Since this is a finite group, it has orthogonality relations for its
    characters.
    Given partitions λ, µ, define the Kostka polynomial (or Green polynomials)
    X
    Kλ,µ (t) = [H2i (Bλ ; C) : Lµ ]ti ∈ Z[t].
    i≥0

    Define Ωλ,µ to be the Poincaré polynomal (in t) of HomSn (Lλ ⊗ Lµ ⊗ sgn, H• (B)). Also, define
    K = (Kλ,µ )λ,µ and Ω = (Ωλ,µ )λ,µ .
    Theorem 1.3 (Shoji). There exists a diagonal square matrix Λ such that K T ΛK = Ω (orthogo-
    nality).
    Remark 1.4. A generalization was given by Lusztig.
    (
    t2 dim Bλ if λ = µ
    Theorem 1.5 (Borho–Mac Pherson). Kλ,µ (t) = , where ≤ is dominance or-
    0 if λ ≤
    6 µ
    dering, i.e., λ ≤ µ if and only if Oλ ⊂ Oµ .
    In particular, since K is upper-triangular and Λ is diagonal, we can calculate them from Ω.
    Question: How can one integrate the perspectives of De Concini–Procesi and Shoki of the
    Springer correspondence?

    1
    2 Reinterpretation of orthogonality relations.
    H• (Bλ ; C) is an Sn -module. But since Bλ ⊂ B, there is a natural action of t∗ = H2 (B; C) also.
    These two actions combine to give an action of A = C[Sn ]⋉C[t]. We define deg(w) = 0 for w ∈ Sn
    and deg(x) = 2 for x ∈ t∗ . This equips H• (Bλ ; C) the structure of a graded A-module. The top of
    H• (Bλ ) is the trivial representation,
    L and the socle of H• (Bλ ) is Lλ in degree 2 dim Bλ .
    Define Mλ = H• (Bλ ) = i∈Z HBM
    BM
    i (Bλ ; C) which is a graded A-algebra. The top of Mλ is Lλ
    in degree 0. Let Pλ = C[t] ⊗ Lλ be the indecomposable projective cover.
    The proper analogue of De Concini–Procesi is to have a surjection (thus far, a speculation)
    Pλ → Mλ → 0 for all λ ⊢ n.
    Let h−, −igEP : K(A-gmod) × K(A-gmod) → Z[t, t−1 ] be the graded Euler–Poincaré character-
    istic, defined by X
    hM, N igEP = (−1)i gdim(ExtiA (M, N )).
    i∈Z

    Lemma 2.1. hPλ , Pµ igEP = ∆gdim(HomSn (Lλ ⊗ Lµ , H• (B)) where ∆ = gdim(C[t]Sn ) = ((1 −
    t4 )(1 − t6 ) · · · (1 − t2n ))−1 .
    Proof. hPλ , Pµ igEP = gdim(HomA (Pλ , Pµ )) = gdim(HomSn (Lλ , Pµ )).

    Note: The right hand side of the lemma is Ω up to ∆ and the switch t 7→ t−1 .
    We define an isomorphism A ∼ = Aop via w ∈ Sn 7→ w−1 and t∗ is mapped
    L to itself identically.
    ∗ ∗
    If M is a finite-dimensional graded A-module, then its graded dual M = i∈Z (Mi ) is naturally
    an Aop -module, and hence an A-module.
    Theorem 2.2. hMλ , Mµ∗ igEP = 0 unless λ = µ.
    Proof. Reinterpret Shoji’s theorem using lemma plus some linear algebra.

    Example 2.3. Set n = 2. Let (12 ) be the trivial module and (2) is the sign representation. Then
    M(1,1) = L1,1 [0] ⊕ L2 [2] and M(2) = L2 [0]. We have exact sequences

    0 → P1,1 (−4) → P1,1 → M1,1 → 0


    0 → P1,1 (−2) → P2 → M2 → 0.

    In particular, ExtiA (M1,1 , M2 ) = 0 for all i ≥ 0.

    3 Kostka systems.
    Let W be a finite Coxeter group with reflection representation t. Let AW = C[W ] ⋉ C[t] with
    deg(W ) = 0 and deg(t∗ ) = 2. Then Aw ։ C[W ] is a maximal semisimple quotient, which implies
    that Lχ (an irreducible W -module) are the irreducible graded A-modules up to shift. We define
    h−, −igEP as before.
    ` ` `
    Definition 3.1 (Phyla). An ordered subdivision Irr(W ) = O1 O2 · · · ON is called a phyla
    P of W . Given χ, χ′ ∈ Irr(W ), we define χ <P χ′ if and only if χ ∈ Oi and χ′ ∈ Oj with i < j,
    and χ ∼P χ′ if and only if χ, χ′ ∈ Oj .
    Definition 3.2 (P-trace). Fix (W, P). For all χ ∈ Irr(W ), we define the P-trace of Pχ as
    X
    Pχ,P = Pχ / image(f ).
    f ∈HomAW (Pχ′ ,Pχ )>0
    χ ′ ≤P χ

    2
    Definition 3.3 (Kostka system). Fix (W, P). Then {Kχ := Pχ,P }χ∈Irr(W ) is called a Kostka
    system if hKχ , Kχ∗′ igEP = 0 unless χ ∼P χ′ . (This is actually a condition on P.)

    Theorem 3.4. {Mλ }λ⊢n is a Kostka system where P is an arbitrary refinement of the dominance
    ordering on partitions.
    1. hMλ , Mµ∗ igEP = 0 unless λ = µ
    2. The defining equations of each Mλ is completely understood in a representation-theoretic way.
    In addition, we have Ext•A (Mλ , Mµ ) = 0 unless λ ≤ µ.

    Remark 3.5. 1. We can prove that all generalized Springer correspondences give Kostka systems.
    2. Via this approach, the “a-functions” are not essential in Springer theory.
    3. Ext•A (Mλ , Mµ ) = 0 combined with Borho–Mac Pherson implies that hMλ , Mµ∗ igEP = 0.

    You might also like