On the complexities of some simple modules of symmetric groups (Q2334473)

Let \(G\) be a finite group and let \(\mathbb{F}\) be a field of prime characteristic \(p\). The \emph{complexity} of an \(\mathbb{F}G\)-module, denoted \(c_G(M)\), is the polynomial growth rate of a minimum projective resolution of \(M\). Thus \(c_G(M) = 0\) if and only if \(M\) has a periodic projective resolution. Let \(S_n\) denote the symmetric group of degree \(n\). The paper under review determines the complexity of the simple \(\mathbb{F}S_n\)-modules \(D^\lambda\) canonically labelled by the partitions \((n-1,1)\), \((n-2,2)\) and also \((n-k,k)\) where \(k \le 2p\) if \(p > 3\) and \(k \le 6\) if \(p=3\). The main results show that, with a small number of exceptions when \(p=2\), the complexity of any such module is equal to the \(p\)-weight of its block. It is known from the theory of rank varieties that the \(p\)-rank of the defect group of a block is an upper bound for the complexity of the modules it contains, so the author's result shows that, in almost all cases, these simple modules have the maximal possible complexity. The exceptional cases are that \(c_{S_n}(D^{(n-1,1)}) = w-1\) if \(p=2\) and \(n=2\) (mod \(4\)) and \(c_{S_n}(D^{(n-2,2)}) = w-1\) if \(p=2\) and \(n=5\) or \(n=6\), where in each case \(w\) is the weight of the relevant block. The proofs use rank varieties, stable generic Jordan types and results from the theory of Young modules.