Nilpotent groups and their automorphisms (Q685985): Difference between revisions

Let \(G\) be a nilpotent group and \(\varphi\) an automorphism of \(G\) of prime order \(p\). A general problem of which a number of special cases are treated in the present book is to determine the structure of \(G\) when \(\varphi\) has certain preassigned properties. Typical results are G. Higman's theorem (1957) that if \(\varphi\) fixes only the identity then the nilpotency class of \(G\) is bounded by a function of \(p\) and J. L. Alperin's theorem (1962) that if \(G\) is a finite \(p\)-group and \(\varphi\) has \(p^ n\) fixed points then the derived length of \(G\) is bounded by a function of \(p\) and \(n\). A different kind of condition, and one extensively investigated by the author himself, is that \(\varphi\) be a splitting automorphism, meaning that \(aa^ \varphi\dots a^{\varphi^{p-1}}=1\) for all \(a\in G\). This superficially resembles the condition that \(a^ p=1\) for all \(a\in G\), and it is indeed related at a deeper level to A. I. Kostrikin's theorem (1959) that the nilpotency class of a finite \(d\)-generator group of prime exponent \(p\) is bounded by a function \(k(p,d)\) of \(p\) and \(d\). Substantial progress has been made in recent years in solving problems like the ones above and their generalizations, and in this the author has played a prominent part. The book gives a carefully detailed account of both new and old results in the area. The theory is developed ab initio and the book is largely self-contained (only in the final two chapters some outside results are required). Many of the proofs involve the associated Lie ring of a nilpotent group and for this reason the book is almost as much about Lie rings as about groups. Following the introductory chapters 1-3, chapters 4 (on Lie rings) and 5 (on groups) are concerned almost exclusively with automorphisms of prime order having a given number of fixed points. The standard proofs of the Higman and Alperin results and of well known extensions of these by V. A. Kreknin and Kostrikin are refined to give the following theorem of the author and Yu. A. Medvedev: if a nilpotent group \(G\) admits an automorphism \(\varphi\) of prime order \(p\) with \(q\) fixed points then \(G\) has a subgroup of \((p,q)\)-bounded index with \(p\)-bounded class. Perhaps the most original part of the book is chapter 6, which deals with groups acted on by a given group \(\Omega\): the action is by automorphisms but need not be faithful. One defines a variety of such \(\Omega\)-groups by identical relations that involve the elements of \(\Omega\) as well as those of \(G\). If \({\mathfrak M}\) is such a variety of \(\Omega\)-groups, the groups \(G\) in \({\mathfrak M}\) on which \(\Omega\) acts as the identity form a variety \(\overline{\mathfrak M}\) of groups in the ordinary sense. The author proves a very interesting local nilpotency theorem relating the groups in \({\mathfrak M}\) to those in \(\overline{\mathfrak M}\). When \(\Omega= \langle \varphi\rangle\) is cyclic of prime order \(p\) and \({\mathfrak M}\) is defined by the splitting identity cited earlier, \(\overline{\mathfrak M}\) becomes the (Burnside) variety of all groups of exponent \(p\). In this case the local nilpotency theorem yields the following result: a \(d\)-generator nilpotent \(p\)-group with a splitting automorphism of order \(p\) has class at most \(k(p,D)\), where \(k\) is the function in Kostrikin's theorem and \(D=d (p^ p-1)/ (p-1)\). A detailed study of the finite \(p\)-groups with a splitting automorphism of order \(p\) is undertaken in chapter 7. These groups are of particular interest because of their connections with the partitions of \(p\)-groups and with Hughes subgroups. The final chapter 8 ventures into new territory, presenting results about automorphisms of prime-power, rather than prime, order. The central result is as follows: a locally nilpotent \(p\)-group that admits an automorphism of order \(p^ k\) with exactly \(p^ n\) fixed points contains a subgroup of \((p,k,n)\)-bounded index that is solvable of \((p,k)\)-bounded derived length. The proof uses, amongst other things, the recently developed theory of powerful \(p\)-groups. The book gives an authoritative account of the recent advances in the subject-area. The discussions at the end of the later chapters point to some of the problems that remain open.