On a problem of Faith

From MaRDI portal





For a ring \(R\) consider the property: (*) For an arbitrary pair \(x, y \in R\) there exist positive integers \(m(x, y), n(x, y)\) such that \(x^{m(x, y)}\) commutes with \(y^{n(x, y)}\). The authors show in a very ingenious way that if \(R\) has property (*) and no nil ideals then \(R\) is commutative. Even more, it is shown that if \(R\) is arbitrary with (*) then the set \(I\) of all nilpotent elements of \(R\) is an ideal of \(R\), with \(R/I\) commutative. This paper is the last in a long sequence of commutativity theorems by various authors, the previous best result is that of \textit{A. I. Likhtman} [Mat. Sb., Nov. Ser. 83(125), 513--523 (1970; Zbl 0212.37703)] who proved the same two theorems for the special case of (*) in which \(m\) and \(n\) are functions respectively of \(x\) and \(y\) alone. Note: The reviewer followed the proofs carefully and found only a couple of errors (trivial misprints) such as the exponents of \(b\) in the material after (3) on page 129.












This page was built for publication: On a problem of Faith

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4046216)