Powers of commutators and anticommutators. (Q1929889)

Let \(\mathbb N\) denote the set of natural numbers. For elements \(x,y\) of the ring \(R\), denote by \([x,y]\) and \(\langle x,y\rangle\), respectively, the commutator \(xy-yx\) and anti-commutator \(xy+yx\). For \(S\subseteq\mathbb N\setminus\{1\}\), call \(R\) a \(\text{CP}(S)\) ring if for each \(x,y\in R\) there exists \(n=n(x,y)\in S\) such that \([x,y]^n=[x,y]\); and for \(S\subseteq 2\mathbb N\), call \(R\) an \(\text{ACP}(S)\) ring if for each \(x,y\in R\) there exists \(n=n(x,y)\in S\) for which \(\langle x,y\rangle^n=\langle x,y\rangle\). A well-known theorem of \textit{I. N. Herstein} asserts that every \(\text{CP}(\mathbb N\setminus\{1\})\) ring is commutative [Can. J. Math. 9, 583-586 (1957; Zbl 0079.05403)]; and a result of the second author states that every \(\text{ACP}(2\mathbb N)\) ring is anti-commutative [Am. Math. Mon. 94, 162-165 (1987; Zbl 0619.16020)]. The original proofs of these results require the use of structure theory; the paper under review seeks proofs which do not, which we describe as elementary. The authors provide elementary proofs of \(\text{CP}(\{3\})\) rings and \(\text{CP}(\{4\})\) rings; and for \(S\subseteq 2\mathbb N\) they give an elementary proof that \(\text{CP}(S)\) rings are commutative if and only if all \(\text{ACP}(S)\) rings are anti-commutative.