Modules over discrete valuation domains (Q2426584)

The book is intended to be a systematic account on modules over discrete valuation domains including history. It develops the subject from scratch requiring minimal knowledge of algebra from the reader. The book contains an extensive bibliography for further reading. For the basic results, the base ring is not assumed to be commutative. For the advanced topics (starting form Section 19) only commutative base rings are considered. The following is a very brief summary of the main topics of the book. Chapter 1 introduces the most basic notions of module theory together with discrete valuation domains and the most easily available modules over them. Chapter 2 discusses important special classes of modules and their characterizations. Some examples: Injective modules are precisely the divisible ones (Theorem 6.5), and they are completely classified as direct sums of copies of two specific modules (Theorem 6.3). Basic submodules exist and they are unique up to isomorphism (Theorem 9.4). Pure-projective modules and pure-injective modules are partially characterized (Theorems 10.2 and 10.3), but the complete classification of pure-injective modules is omitted. Basic poperties of complete modules is developed in Section 11. The main subject of Chapter 3 is the endomorphism ring of divisible modules and complete torsion-free modules. It culminates in the identification of the lattice of left ideals of the endomorphism ring of a complete torsion-free module with the lattice of complete submodules with p-rank not exceeding the one of the big module (Theorem 15.4). A similar result is proved for divisible modules (Theorem 15.7) using the Harrison--Matlis equivalence between divisible modules and complete torsion-free modules (Theorem 13.1), which is also proven in this chapter. Chapter 4 deals with representations of rings as endomorphism rings. It shows only the tip of the iceberg, since a chapter cannot compete with whole books devoted to the subject like [\textit{R. Göbel} and \textit{J. Trlifaj}, Approximations and endomorphism algebras of modules (de Gruyter Expositions in Mathematics, Walter de Gruyter, Berlin) (2006; Zbl 1121.16002)]. First, some characterization theorems are deduced, a notable example is a ring theoretic characterization of rings appearing as endomorphism rings of a reduced torsion-free module over a complete discrete valuation domain (Theorem 18.4). Second, explicit realization theorems for a large class of rings are stated (Section 19) but the proof of the most advanced ones are omitted because they are too involved. As application, existence of many essentially indecomposable modules over complete discrete valuation domains are shown in Section 20. Section 22 presents some methods to obtain realization theorems for mixed modules from realization theorems for torsion-free modules. Chapter 5 studies indecomposable torsion-free modules of finite rank. The main tool introduced is the category of quasi-homomorphisms, leading to Arnold duality (Theorem 24.9). As a trivial example, Arnold duality identifies free and divisible modules. A less obvious example is studied in section 25: copurely indecomposable modules are exactly the duals of purely indecomposable modules. The limits of Arnold duality is also indicated: while purely indecomposable modules are easily listed, determining the isomorphism classes of copurely indecomposable modules is a difficult task. The final section of the chapter specializes to Nagata valuation domains, which are the discrete valuation domains where the quotient field of the completion of the domain is a finite extension of the quotien field of the domain itself. For Nagata valuation domains for which the degree of the above field extension is 2, indecomposable torsion-free modules of finite rank are completely classified (Theorem 26.1). Also, a complete classification is given for indecomposable torsion-free modules of rank at most 2 over any Nagata valuation domain (Theorem 26.2). Chapter 6 is devoted to the theory of Warfield modules. The central result is that Warfield modules are completely determined by cardinal invariants (Theorem 32.6). The path to this result is very well presented. A notable tool is defining a new category of modules with new morphisms (like the category of quasi-homomorphisms in Chapter 5) where Warfield modules completely decompose. The uniqueness of this decomposition produces the Warfield invariants for example. The only omission is the existence of Warfield modules with arbitrary invariants. Chapter 7 contains results of the type: every isomorphism of the endomorphism ring of two modules of certain kind is induced by an isomorphism between the modules. Notable examples when this is true are: (1) primary modules over a complete discrete valuation domain (Theorem 33.2), (2) reduced torsion-free modules over a complete discrete valuation domain (Theorem 33.5), (3) Warfield modules (Theorem 36.4). These are proved by the famous Kaplansky method. The authors also present variants when only topological isomorphisms between endomorphism rings are claimed to be induced by isomorphism of modules. A sample result: Every topological isomorphism between the endomorphism ring of a primary module or a reduced torsion-free module and and the endomorphism ring of an arbitrary module is induced by an isomorphism of the modules (Theorem 33.6(2)). Finally, Chapter 8 is an introduction to transitive and fully transitive modules. Loosely speaking, a module is transitive if any element is mapped to any element subject to some obvious restriction by an automorphism. A module is fully transitive if any element is mapped to any element subject to obvious restrictions by an endomorphism. Only basic examples are given for such modules, like divisble modules over any discrete valuation domain or torsion-free modules over a complete discrete valuation domain. The other results are about the relationship between the various notions of transitivity (not all of which mentioned here), mainly focusing on completely decomposable modules. Some sample results: (1) For a completely decomposable module, which is either a Warfield module or its torsion submodule does not have lements of infinite height, being transitive and being fully transitive are equivalent (Theorem 39.1). (2) A completely decomposable module is fully transitive if and only if its square is transitive (Theorem 39.5).