A row removal theorem for the Ext¹ quiver of symmetric groups and Schur algebras (Q4825629)

Consider the symmetric group \(\Sigma_r\) over an algebraically closed field \(k\). In characteristic zero, the simple \(k\Sigma_r\)-modules are the Specht modules (indexed by partitions of \(r\)). This paper is interested in the modular case when the characteristic of \(k\) is a prime \(p\), and the Specht modules are not simple in general. Here the simple modules (again classified by partitions) arise as heads or socles of Specht modules. A fundamental question is to determine the composition multiplicity of a simple module in a Specht (or dual Specht) module. The basis for this work are two theorems of \textit{G. D. James} [J. Algebra 71, 115-122 (1981; Zbl 0465.20010)] on this problem in the case when the associated partitions have equal first row (or part). The theorems state the multiplicity is the same as for the corresponding modules after removing the first row from the associated partitions. James' approach was to first work with related Schur algebras and then use the Schur functor to obtain the results for symmetric groups. The first main result of this paper uses James' techniques to deduce an injection on \(\text{Ext}^1\)-groups of simple modules over Schur algebras after row removal. Note that the extensions over Schur algebras are equivalent to certain extensions over a general linear group. In recent work of \textit{B. Parshall} and \textit{L. Scott} [Extensions, Levi subgroups, and character formulas (submitted for publication)], this result is extended to more general algebraic groups. The second main result is an analogous injection on \(\text{Ext}^1\)-groups of simple modules over \(k\Sigma_r\) under the restriction that the first part of the partitions involved are less than \(p\). To obtain this result, the author reproves weaker versions of James' results (with this partition restriction) using only the symmetric group (i.e., without translating to Schur algebras). This is done through the use of an interesting fixed point functor from \(k\Sigma_r\)-modules to \(k\Sigma_{r-m}\)-modules. The author studies this functor more fully in another paper [J. Algebra 280, No. 1, 295-312 (2004; Zbl 1079.20015)].