Finite simple groups and localization (Q1852720)

In this context a localization is a group homomorphism \(f\colon H\to G\) with the property that every homomorphism from \(H\) to \(G\) extends to a unique endomorphism of \(G\). This paper is devoted to studying the case when \(f\) is injective and \(H\) and \(G\) are simple. In this case we can assume \(f\) is an inclusion, and then \(f\) is a localization if and only if (i) \(f\) extends to an inclusion \(\Aut(H)\to\Aut(G)\), (ii) if \(H\cong K\subseteq G\) then \(H\) and \(K\) are conjugate in \(\Aut(G)\), and (iii) \(C_{\Aut(G)}(H)=1\). An equivalence relation is defined on the set of finite simple groups by taking the symmetric-transitive closure of the relation ``\(G\) is a localization of \(H\)'' -- we can then ask about the equivalence classes (here called rigid components). The main result of the paper is that all the alternating groups are in the same rigid component, which also contains all \(L_2(q)\) and \(U_3(q)\), all sporadic groups (the exclusion of the Monster can easily be remedied), and some others. In fact, a great deal is known about inclusions of finite simple groups in one another, and the methods of this paper can be extended to show that this rigid component contains all groups \(S_{2n}(q)\), \(O_{2n+1}(q)\) and \(U_n(q)\) in dimension at least 6, most even-dimensional orthogonal groups \(O^\pm_{2n}(q)\) (\(n>4\) and \(q>3\) suffices), as well as a number of families of exceptional groups of Lie type.