On degenerations of tame and wild algebras (Q1345859)

Given \(k\), an algebraically closed field, and \(d\), a natural number, let us denote by \(\text{alg}(d)\) the affine variety of algebra structures on \(k^d\). The main result of this nice paper is the following Theorem: Every degeneration of a tame algebra in \(\text{alg}(d)\), under the action of the full linear group, is itself tame. The theorem is a partial generalization of \textit{P. Gabriel}'s famous result [see Lect. Notes Math. 488, 132-155 (1975; Zbl 0313.16034)]. The proof is based on a new characterization of tameness that uses the closed sets (depending on the two parameters \(z\) and \(t\)) consisting of modules of dimension \(z\) whose endomorphism rings have dimension not less than \(t\). A starting fact is that the dimension of these closed sets is upper semicontinuous as a function with domain \(\text{alg}(d)\). The reader interested in these techniques should take a look also at the author's Ph. D. Thesis [Tame distributive algebras and related topics (Univ. Bayreuth, 1993)] and some other significant papers like the following ones of \textit{J. A. de la Peña} [Commun. Algebra 19, 1795-1807 (1991; Zbl 0818.16013) and Fundam. Math. 137, 177-185 (1991; Zbl 0790.16014)].