Exact categories, big Cohen-Macaulay modules and finite representation type (Q2664604)

An important problem in the representation theory of algebras is to obtain a characterization of algebras of finite representation type. One of the first remarkable results in this direction shows that an artin algebra \( \Lambda \) is of finite representation type if and only if every \( \Lambda \)-module is a direct sum of finitely generated \( \Lambda \)-modules. This result is famous as Auslander-Ringel-Tachikawa theorem and is now about 50 years old [\textit{C. M. Ringel} and \textit{H. Tachikawa}, J. Reine Angew. Math. 272, 49--72 (1975; Zbl 0318.16006), Corollary 4.4]. A remarkable analogue of the Auslander-Ringel-Tachikawa theorem was first proved by [\textit{X.-W. Chen}, Adv. Math. 218, No. 6, 2043--2050 (2008; Zbl 1147.16015)], and later generalized by [\textit{A. Beligiannis}, Adv. Math. 226, No. 2, 1973--2019 (2011; Zbl 1239.16016)], for the subcategory of Gorenstein projective modules over an artin algebra \( \Lambda \). Beligiannis proved that any Gorenstein projective \( \Lambda \)-module is a direct sum of finitely generated ones if and only if the artin algebra \( \Lambda \) is virtually Gorenstein of finite Cohen-Macaulay type [Beligiannis, loc. cit.]. The latter result improved Chen's theorem [Chen, loc. cit.] for Gorenstein algebras. In this work, the authors, based on a suitable concept of (accessible) big Cohen-Macaulay modules, extends the Auslander-Ringel-Tachikawa theorem to arbitrary finite dimension. More precisely, they show that if \( \Lambda \) is a Cohen-Macaulay order over a complete \( d \)-dimensional regular local ring \( R \), then \( \Lambda \) is of finite type if and only if every (accessible) big Cohen-Macaulay \( \Lambda \)-module is a direct sum of finitely generated \( \Lambda \)-modules (c.f. Theorem 5.18). Recall that if \( R \) is a complete regular local ring of Krull dimension \( d \), then a Cohen-Macaulay order \( \Lambda \) over \( R \) is an \( R \)-algebra which is finitely generated and free over \( R \). Note that this result is a consequence of another important and remarkable result, called \textit{an Auslander-type ``splitting-big-objects'' Theorem}, that has a central role in this paper, Theorem 5.10. To state this theorem we need some notations and for this reason we refer the reader to page \( 3 \) of the work.