Branching of modular representations of the alternating groups (Q1272459): Difference between revisions

Let \(p\) be a prime and let \(A_n\) denote the alternating group of degree \(n\). The purpose of this paper is to investigate the restriction to the subgroup \(A_{n-1}\) of a \(p\)-modular irreducible representation of \(A_n\) (that is, to investigate \(p\)-modular branching rules for \(A_n\)). In particular, the authors seek to describe those representations of \(A_n\) which remain irreducible on restriction to \(A_{n-1}\). The first problem to be solved is to describe in some way the \(p\)-modular irreducible representations of \(A_n\). These are obtained from the irreducible representations of the symmetric group \(S_n\), which, on restriction to \(A_n\), either split into two different irreducible representations of \(A_n\) or else remain irreducible. The solution of the Mullineux conjecture due to Ford and Kleshchev (1997) shows that, for odd \(p\), the former representations are described by the fixed points of the Mullineux involution acting on the set of \(p\)-regular partitions, whereas the others correspond to non-fixed points. For \(p=2\), a result of Benson (1988) describes the former representations in terms of special properties of certain 2-regular partitions of \(n\). We cannot explain the authors' results in any detail but we will briefly allude to the question of when an irreducible \(p\)-modular representation of \(A_n\) remains irreducible on restriction to \(A_{n-1}\). The corresponding question had been answered for \(S_n\) by Kleshchev in 1995, in terms of Jantzen-Seitz \(p\)-regular partitions of \(n\). When \(p=2\), these partitions are easily explained, since they are partitions into distinct parts all having the same parity (as conjectured by Benson in 1986). The authors' solution to the problem of when an irreducible \(p\)-modular representation of \(A_n\) remains irreducible on restriction to \(A_{n-1}\) is given in complete detail in their Theorem 5.10 when \(p\) is odd and it involves both the Mullineux involution and Jantzen-Seitz partitions. The question for \(p=2\) is not solved as completely as for odd primes, as the combinatorics is more complicated.