On the cohomology of Young modules for the symmetric group

TL;DR

This paper applies topological methods to compute the cohomology of symmetric group modules, especially Young modules, reducing complex algebraic problems to representation theory and revealing stability phenomena in cohomology.

Contribution

It extends classical topological techniques to describe symmetric group cohomology with Young modules, providing explicit calculations and stability results, including the first complete cohomology description for a class of modules.

Findings

Complete cohomology description for many Young modules in characteristic two.

Stability of cohomology groups as tensor powers increase.

Criteria for vanishing cohomology of symmetric group modules.

Abstract

The main result of this paper is an application of the topology of the space $Q(X)$ to obtain results for the cohomology of the symmetric group on $d$ letters, $\Sigma_d$, with `twisted' coefficients in various choices of Young modules and to show that these computations reduce to certain natural questions in representation theory. The authors extend classical methods for analyzing the homology of certain spaces $Q(X)$ with mod-$p$ coefficients to describe the homology $\HH_{\bullet}(\Sigma_d, V^{\otimes d})$ as a module for the general linear group $GL(V)$ over an algebraically closed field $k$ of characteristic $p$. As a direct application, these results provide a method of reducing the computation of $\text{Ext}^{\bullet}_{\Sigma_{d}}(Y^{\lambda},Y^{\mu})$ (where $Y^{\lambda}$, $Y^{\mu}$ are Young modules) to a representation theoretic problem involving the determination of tensor products and decomposition numbers. In particular, in characteristic two, for many $d$, a complete determination of $\Hs Y^\lambda)$ can be found. This is the first nontrivial class of symmetric group modules where a complete description of the cohomology in all degrees can be given. For arbitrary $d$ the authors determine $\HH^i(\Sigma_d,Y^\lambda)$ for $i=0,1,2$. An interesting phenomenon is uncovered--namely a stability result reminiscent of generic cohomology for algebraic groups. For each $i$ the cohomology $\HH^i(\Sigma_{p^ad}, Y^{p^a\lambda})$ stabilizes as $a$ increases. The methods in this paper are also powerful enough to determine, for any $p$ and $\lambda$, precisely when $\HH^{\bullet}(\sd,Y^\lambda)=0$. Such modules with vanishing cohomology are of great interest in representation theory because their support varieties constitute the representation theoretic nucleus.