6
$\begingroup$

I'm interested in finite Iwahori-Hecke algebras.

If $\mathcal{H}$ is such a Hecke algebra, defined over $\mathbb{Z}[q^{\pm 1/2}]$, and $\Lambda$ an irreductible representation, there is the notion of a Schur element $S_\Lambda$. Roughly speaking, to $\Lambda$ you can associate a central element whose matrix in $\Lambda$ is given by $S_\Lambda$ times the identity.

I computed several of these Schur elements in type $A_n$ and I noticed a remarkable positivity property: all the coefficients in $q$ seem to be positive (see below)! I looked up in the literature, but I don't find explicitely this property.

Is there a concrete reference for this positivity? Does it hold in other types?

Here a small list of Schur elements:

  • For $\mathfrak{sl}_2$: $1+q$, $1+q^{-1}$
  • For $\mathfrak{sl}_3$: $q+1+q^{-1}$, $1+2q+2q^2+q^3$, $1+2q^{-1}+2q^{-2}+q^{-3}$
  • For $\mathfrak{sl}_4$: $[4]!$ (quantum factorial), $q^2+3q+4+3q^{-1}+q^{-2}$, $q^{-1}+2+2q+2q^2+q^3$, ...
Improve this question
$\endgroup$
8
  • 2
    $\begingroup$ How does one compute these? Is there a formula, or similar that you can reference? $\endgroup$
    – Per Alexandersson
    Mar 11, 2021 at 19:00
  • 2
    $\begingroup$ The representation $\Lambda$ is given by a left cell (at least in case $A_n$). Let $(C_w)$ denote the Kazhdan-Lusztig basis and let $(C^w)$ denote its dual basis, such that $tr (C_wC^v) = \delta_w^v$ where you use the standard trace. Now, for a given left cell $\Lambda$, the element $Z_\Lambda=\sum_{x\in \Lambda}C_xC^x$ is central in $\mathcal{H}$. Its action on $\Lambda$ gives the Schuer element $S_\Lambda$. There is also a formula, given in Neunhöffers article sciencedirect.com/science/article/pii/S0021869306001955, Definition 5.1. $\endgroup$
    – AThomas
    Mar 12, 2021 at 8:42
  • $\begingroup$ Of course you are no doubt aware of this, but there are many positivity phenomena in Kazhdan-Lusztig theory (e.g. arxiv.org/abs/1212.0791). $\endgroup$
    – Sam Hopkins
    Mar 12, 2021 at 13:22
  • 1
    $\begingroup$ In type A, the Schur element is a q-deformation of the product of the hook lengths. I would then naturally expect the Schur element to be the product of the corresponding quantum integers. $\endgroup$
    – Peter McNamara
    Mar 13, 2021 at 4:54
  • $\begingroup$ @PeterMcNamara: This is great! It seems indeed that the Schur elements are the produt of the quantum integers of the hook lengths. Is there a similar interpretation in other types? $\endgroup$
    – AThomas
    Mar 13, 2021 at 16:54

1 Answer 1

Reset to default
2
$\begingroup$

Together with Maria Chlouveraki, we determined the answer:

The Schur elements are positive in all classical types $A_n, B_n$ and $D_n$ (use the formula of Theorem 4.3. in https://hal.archives-ouvertes.fr/tel-01411063/document, which uses generalized hook lengths). For dihedral groups $I_2(m)$, Theorem 8.3.4. in the book of Geck-Pfeiffer on characters of finite Coxeter groups and Hecke algebras shows that the Schur elements have negative coefficients for $m>4$. For the exceptional cases, by brute force, there are Schur elements with negative coefficients only in type $H_4, F_4$ and $E_8$.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Did there end up being any connection to fake degrees? $\endgroup$
    – Sam Hopkins
    Mar 16, 2021 at 13:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.