“…its largest semiartinian submodule) splits off. This is part of a general class of problems called splitting problems (relative to torsion theories), which have a long history of mathematical interest (see, for example, [G,K1,K2,Rot,T1,T2,T3]; also, [C,I1,IO,NT] for splitting problems for profinite and pseudocompact algebras). It was originally conjectured that if the Dickson torsion splits off in any left A-module for a ring A, then A is left semiartinian; this was proved to be false by a counterexample in [Co].…”