The theorem of Baer and Kaplansky for mixed modules (Q1908473)

Many papers on endomorphism rings of modules deal with two kinds of problems which are connected with the work of Baer, Corner, Kaplansky and Wolfson. Either it is shown that the endomorphism ring determines the module or there are very many modules with the same endomorphism ring. The first kind of result holds only for restricted classes of modules like abelian \(p\)-groups (Theorem of Baer-Kaplansky). From this point of view the following main result of this paper is very surprising. Theorem. Let \(R\) be a complete discrete valuation ring with prime \(p\). If \(M\) is an \(R\)-module and \(M/p^\omega M\) is a (mixed) \(R\)-module of countable torsion-free rank and \(\text{End}(M)\cong\text{End}(N)\) (\(\text{End}( . )\) denotes the algebra of \(R\)-endomorphisms) for some \(R\)-module \(N\), then \(M\cong N\). Recall that \(p^\omega M=\bigcap_{n\in\omega}p^nM\) denotes the first Ulm-submodule of \(M\) and note that the isomorphism between \(M\) and \(N\) is not induced. Main tools are an existence theorem for small homomorphisms (in the sense of Pierce) and a detailed study of cotorsion hulls.