Large \(E\)-modules exist (Q1178926): Difference between revisions

A (right) module \(M\) over a ring \(R\) with 1 is an \(E\)-module if \(\hbox{Hom}_ \mathbb{Z}(R,M)=\hbox{Hom}_ R(R,M)\). The ring is an \(E\)-ring if \(R_ R\) is an \(E\)-module. The author answers affirmatively a question of C. Vinsonhaler: If \(R\) is an E-ring, are there arbitrarily large indecomposable \(E\)-modules over \(R\)? More precisely, the following theorem is proved. Theorem. Let \(R\) be a ring with 1 such that the additive group \(R^ +\) of \(R\) is slender and \(| R|<\kappa\) where \(\kappa\) is a regular uncountable cardinal less than the first measurable cardinal (if such exist). Let \(\lambda>\kappa\) be any cardinal with \(\lambda=\lambda^{2^ \kappa}\). Then there is an \(R\)-module \(G\) such that (1) \(| G|=\lambda\), (2) \(\hbox{End}_ \mathbb{Z} G=R\), (3) every \(R\)-submodule \(M\) of \(G\) with \(| M|<\kappa\) is imbedded in a cartesian power of \(R\), (4) \(G\) is slender, (5) if \(R\) is an \(E\)-ring then \(G_ R\) is an \(E\)-module. If \(R\) is an \(E\)-ring with no idempotents other than 0, 1, then \(G\) is directly indecomposable. The proof uses ``Shelah's Black Box''.