No label defined (Q4258034)

This is a very interesting survey article which comments on several topics, especially on the pure semisimple conjecture, giving some previous results of the author. (It is advisable to take a look at his articles ``An Artin problem for division ring extensions and the pure semisimplicity conjecture. I, II'' [part I: Arch. Math. 66, No. 2, 114-122 (1996; Zbl 0873.16010)].) The main result is a reduction theorem for the validity of the conjecture (cf. Theorem 3.4), stating (among others) that a necessary and sufficient condition is the following. For any pair of division rings \(F\), \(G\) and for any simple \(F\)-\(G\)-bimodule \(M\) with \(M_G\) finite and \(_FM\) infinite it is possible to construct an indecomposable module of finite right length over the hereditary ring \(\left(\begin{smallmatrix} F &M\\ 0 &G\end{smallmatrix}\right)\). As a corollary, it is given the interesting fact (see Theorem 3.5) that the conjecture fails in case one can find an overfield \(G\) of a field \(F\) with infinite left dimension and right dimension equal to 2 and with all iterated right dualizations of \(G\) having dimension 2. In this case, the AR quiver of the above mentionned hereditary ring is really strange for it has only two components, one infinite and one with only two (projective) modules. Thus, the pure semisimplicity conjecture is linked to an old (believed to be) Artin problem. For the entire collection see [Zbl 0910.00039].