An application of the amalgam method: The \(2\)-local structure of \(N\)-groups of characteristic \(2\) type (Q1355559)

One of the fundamental theorems in the classification of the finite simple groups is Thompson's work on \(N\)-groups. This arose from the question to classify the minimal nonsolvable groups, i.e. the nonsolvable groups all of whose proper subgroups are solvable. It turned out that the more general problem to classify all simple groups all of whose local subgroups are solvable was the appropriate way. The main importance of that paper was that it proved that one is able to get general classification results just using the structure of local groups. This then became a pattern for large parts of the classification proof of the finite simple groups in general. Thompson's result was generalized by \textit{D. Gorenstein}, \textit{R. Lyons} [J. Algebra 38, 453-522 (1976; Zbl 0402.20012)], \textit{Z. Janko} [J. Algebra 21, 458-517 (1972; Zbl 0243.20013)] and \textit{F. Smith} [J. Algebra 34, 481-520 (1975; Zbl 0325.20009)] to the case that one just assumes that all 2-local subgroups are solvable. In the paper under review the author gives a new proof of the later result. For this he uses the amalgam method which has been developed over the last ten years. This method works very well as long as one focuses only at the local structure of the group and not at the global one. The theorem proved is the following. Let \(H\) be a finite group, \(S_0\) a Sylow 2-subgroup of \(H\) and \(B=C_{S_0}(\Omega_1(Z(J(S_0))))\). Assume furthermore that every 2-local subgroup of \(H\) containing \(B\) is solvable and of characteristic 2-type and that further there exist at least two maximal 2-local subgroups of \(H\) containing \(S_0\). Then \(H\) is of type \(L_3(2)\), \(Sp_4(2)\), \(G_2(2)'\), \(F_4(2)'\), \(M_{12}\), \(\Omega_6^+(2)\), \(\Omega_6^-(3)\) or \(\Omega_8^+(3)\), in particular \(|S_0|\leq 2^{15}\). Here of type \(X\) means that there is a pair of subgroups \(P_1,P_2\) in \(H\) such that \(O_2(\langle P_1,P_2\rangle)=1\), \(S_0\leq P_1\cap P_2\) and the structure of \(P_1,P_2\) is like that of a pair of maximal 2-local subgroups in \(X\). As a corollary one gets a classification of the simple groups of characteristic two type all of whose 2-local subgroups are solvable. For the proof the author starts with a pair of groups \(P_1,P_2\) with \(S_0\leq P_1\cap P_2\), \(O_2(P_i)\neq 1\), \(i=1,2\), and \(O_2(\langle P_1,P_2\rangle)=1\). Extending an argument of \textit{K. Gomi} [J. Algebra 108, 492-502 (1987; Zbl 0618.20009)] the authors gets even a pair with the additional properties (i) \(S_0\) is contained in a unique maximal subgroup of \(P_i\), \(i=1,2\). (ii) Either \(O^2(P_2)\vartriangleleft\vartriangleleft C_H(\Omega_1(Z(S_0)))\) or \(\Omega_1(Z(S_0))\) is neither normal in \(P_1\) nor \(P_2\). The first property reduces the structure of \(P_1\) and \(P_2\). If \(\Omega_1(Z(S_0))\) is neither normal in \(P_1\) nor \(P_2\), the amalgam method yields small modules for these groups, which then can easily be used to get the structure. If \(\Omega_1(Z(S_0))\vartriangleleft P_2\), then usually the amalgam method does not run smoothly. But here the subnormality which can be used to connect the \(P_i\) with the global structure of \(H\) overcomes a lot of problems. As the \(N\)-group paper became a model for the classification of groups of characteristic 2-type it can be hoped that the paper under review will become a model for a new proof in the revision of the classification based on the new amalgam method.