A commutator description of the solvable radical of a finite group. (Q2475045)

This impressive technical paper (the bulk of the argument consists of careful analysis of the simple nonabelian groups) succeeds in delivering the most pleasant and elegant output. The authors characterize (Th. 1.3) for a finite group \(G\) the solvable radical of \(G\) as being the collection of all \(g\in G\) such that, for any 7 elements \(a_1,\dots,a_7\in G\) the subgroup generated by the 8 elements \(g,a_ig{a_i}^{-1}\), \(1\leq i\leq 7\), is solvable. As a consequence (Th. 1.4), a finite group \(G\) is solvable if and only if every 8 conjugate elements of \(G\) generate a solvable subgroup. The authors conjecture that in Th. 1.4 the number 8 can be reduced to 4 -- this would be the best possible improvement, as in the symmetric groups \(S_n\) (\(n\geq 5\)) any triple of transpositions generates a solvable subgroup. The technique the authors use is to define, for an integer \(k\geq 2\) the so-called \(k\)-radical elements. An element \(g\in G\) is a \(k\)-radical element if for any \(k\) elements of \(G\) the subgroup generated by the simple commutators of \(g\) with these elements is solvable. Then the CFSG is used for a case by case inspection in order to prove a powerful technical result, (Th. 1.11): A finite nonabelian simple group has no nontrivial 3-radical elements.