On arithmetical rings and the radical formula (Q614274)

Let \(R\) be a commutative ring with identity and \(M\) be a unitary \(R\)-module. A submodule \(N\subseteq M\) is called prime if \(rm\in N\) implies \(m\in N\) or \(rM\subseteq N\) whenever \(r\in R\) and \(m\in M\). The (prime) radical of a submodule \(N\) of the module \(M\), \(\mathrm{Rad}_M(N)\), is defined to be the intersection over all prime submodules \(K\) of \(M\) containing \(N\); or \(M\), in case there are no prime submodules containing \(N\). The envelope \(E_M(N)\) of \(N\) is the set \(\{x\mid x=rm,\;r^km\in N,\text{ for some }r\in R,\;m\in M,\text{ and integer }k \}\). \(M\) is said to satisfy the radical formula if \(\mathrm{Rad}_M(N)\) is equal to the submodule generated by \(E_M(N)\) for every submodule \(N\). The paper under review is about verifying which modules over which rings satisfy the radical formula. In spite of the generality of the problem, the paper is restricted to those rings or modules which are ``locally totally ordered'' or ``serial'', respectively. A commutative ring is locally totally ordered, also called ``arithmetical'', if, locally, its ideals are totally ordered by set inclusion; and serial modules are modules whose every two submodules are comparable. See also [\textit{C. U. Jensen}, Acta Math. Acad. Sci. Hung. 17, 115--123 (1966; Zbl 0141.03502)].