Remarks on proficient groups. (Q2431539)

The primary aim of this article is to provide families of examples of finite perfect groups which are proficient, i.e., which admit profinite presentations with `few' relations. The results are closely related to the authors' papers on presentations of finite simple groups; [see Groups Geom. Dyn. 1, No. 4, 469-523 (2007; Zbl 1135.20024); J.\ Am.\ Math.\ Soc.\ 21, No. 3, 711-774 (2008; Zbl 1223.20024) and J.\ Eur.\ Math.\ Soc. (JEMS) 13, No. 2, 391-458 (2011; Zbl 1256.20008)]. Let \(G\) be a finite group. It is known that, if \(G\) has a (profinite) presentation with \(d\) generators and \(r\) relations, then \(r-d\geq d(M(G))\), where \(d(M(G))\) denotes the minimal number of generators of the Schur multiplier \(M(G)\cong H_2(G,\mathbb Z)\) of \(G\). A presentation (resp.\ profinite presentation) of \(G\) is called efficient (resp.\ proficient) if \(r-d=d(M(G))\). The group \(G\) is said to be efficient (resp.\ proficient) if it admits an efficient (resp.\ proficient) presentation. Efficient presentations have been studied for some time. Ongoing research is aimed, for instance, at finding efficient presentations for finite simple groups. At present the list of finite perfect or \(2\)-generated groups that are known to be efficient is rather limited. The notion of proficiency was introduced by \textit{K. W. Gruenberg} and \textit{L. G. Kovács} [Bull.\ Aust.\ Math.\ Soc.\ 60, No. 2, 177-189 (1999; Zbl 0961.20029)]. Every efficient finite group is proficient; it is an open problem whether there exist proficient finite groups which are not efficient. While many finite groups are proficient, e.g., all finite nilpotent groups, not every finite soluble group is proficient. Proficiency admits a useful cohomological characterisation; this interpretation plays a crucial role in the paper under review. The authors describe their work as an offshoot of their earlier papers on presentations of finite simple groups [op.\ cit.]. They show that several types of finite perfect groups are proficient, in particular infinitely many alternating groups, symmetric groups and their double covers. Indeed, they prove that \(G\) is proficient if \(G\) is one of the following: (a) a direct product of two or more simple alternating groups; (b) a direct product of finite quasisimple groups such that \(d(M(G))\geq 16\); (c) a direct product of quasisimple groups each of which is a covering group of \(\mathrm{PSL}(2,q_i)\) for some prime power \(q_i>3\); (d) \(\mathrm{Sym}(p+2)\) where \(p\equiv_3 2\) is prime; (e) \(\mathrm{Alt}(p+2)\), \(\mathrm{Sym}(p+2)\), \(2\mathrm{Alt}(p+2)\) or \(2\mathrm{Sym}(p+2)\) where \(p\equiv_4 3\) is prime; (f) \(\mathrm{SL}(2,q)\) where \(q\geq 4\) is a prime power; (g) \(\mathrm{PSL}(3,q)\) where \(q\equiv_3 1\) is a prime power; (h) \(\mathrm{PSL}(4,q)\) where \(q\) is an odd prime power. These criteria are obtained as applications of the authors' earlier results [see op.\ cit.], by a variation of methods and in combination with cohomological arguments. Indeed, the paper contains several auxiliary cohomological results, e.g.\ regarding the cohomology of amalgamated products. It also features a general construction of non-proficient finite perfect groups and several interesting conjectures. Finally, the authors take the opportunity to correct two (minor) errors in their earlier paper [Groups Geom.\ Dyn., op.\ cit.].