Some model theory over hereditary Noetherian domains (Q1279823)

This paper can be seen as a part of a general program (at the intersection of algebra and model theory) of applying techniques and concepts from mathematical logic (model theory) to understand both the internal structure of modules and the structure of the category of modules over some fixed ring \(R\). The point of intersection is in the study of (indecomposable) pure-injective modules, which have natural definitions and arise naturally as an object of study both in module theory and in model theory. The set of all isomorphism types of pure-injective modules over a fixed ring \(R\) may be endowed with a natural quasi-compact topology, forming the Ziegler spectrum of \(R\). The power of the techniques lies in the fact that methods of module theory and methods of model theory intermingle in a very natural way. Here those techniques are applied to modules over hereditary Noetherian domains, and in particular to differential polynomial rings and generalized Weyl algebras (in the sense of Bavula). The results here are all of an `analysis of complexity' nature, that is, either strong structure theorems are obtained, or an explicit description of why the situation is inaccessibly complicated is given. One goal is to describe the points of the Ziegler spectrum, that is, to classify the indecomposable pure-injective modules over \(R\) by some list of isomorphism invariants. A second aim is to understand the structure and complexity of the topology of the Ziegler spectrum. The Cantor-Bendixson rank (CB-rank) of this topology has natural module-theoretic and model-theoretic interpretations. The introduction to the paper gives a good description of the background leading up to the subject at hand; coupled with the extensive bibliography the first few pages of this article make an excellent (albeit brief) introduction to the subject. The article continues with a summary of necessary results from module theory and ring theory, and a development of the properties of the Ziegler spectrum as they occur in this setting. On the one hand, the authors show that if \(R\) is a principal ideal domain that is a \(V\)-ring (that is, all simple modules are injective) and not a division ring, then the CB-rank of the Ziegler spectrum is \(1\), and the points are just the simple modules and the direct summands of the pure-injective envelope of \(R\). The unique non-isolated point is the injective envelope of \(R\). In particular, if \(R\) is a differential polynomial ring over a universal field with derivation, then the Ziegler spectrum of \(R\) has precisely three points. On the other hand, the authors show that for a simple hereditary generalized Weyl algebra \(A\) over a field \(k\) of characteristic zero, the situation is entirely different. The Ziegler spectrum has no isolated points and so the CB-rank is undefined. The classification problem appears to be hopeless: if \(k\) is countable then the Ziegler spectrum has cardinality \(2^\omega\) and there is a superdecomposable pure-injective (that is, one with no indecomposable direct summands). They then go on to show that for a wide class of generalized Weyl algebras, the theory of modules interprets the theory of \(k\langle x,y\rangle\)-modules, and so hence interprets the word problem for groups. This applies in particular to such rings as \(A_1(k)\) and \(\text{Usl}_2(k)\) (the first Weyl algebra and the universal enveloping algebra, respectively).