There is no largest proper operator ideal

From MaRDI portal



Abstract: An operator ideal is proper if the only operators of the form IdX it contains have finite rank. We answer a question posed by Pietsch (1979) by proving that there is no largest proper operator ideal. Our proof is based on an extension of the construction by Aiena-Gonz'alez (2000), of an improjective but essential operator on Gowers-Maurey's shift space XS (1997), through a new analysis of the algebra of operators on powers of XS. We also prove that certain properties hold for general mathbbC-linear operators if and only if they hold for these operators seen as real: for example this holds for the ideals of strictly singular, strictly cosingular, or inessential operators, answering a question of Gonz'alez-Herrera (2007). This gives us a frame to extend the negative answer to the question of Pietsch to the real setting.


If \(X, Y \) are Banach spaces let us denote by \(L(X, Y)\) the space of bounded linear operators from \(X\) to \(Y\). An operator \(R \in L(X, Y) \) is inessential if the operator \(\mathrm{Id}_X-TR\) is Fredholm for every \(T \in L(Y, X)\). \textit{A. Pietsch} [Operator ideals. Licenced ed. Amsterdam: Elsevier (North-Holland) (1980; Zbl 0434.47030)] conjectured that the operator ideal of inessential operators is the largest proper operator ideal. In the paper under review, the author shows that there exists a proper operator ideal which contains strictly the operator ideal of inessential operators. Moreover, he proves that there is no largest proper operator ideal. To obtain these results he uses and improves previous results of \textit{P. Aiena} and \textit{M. González} [Math. Z. 233, No. 3, 471--479 (2000; Zbl 0960.47009)]. He also proves that certain properties hold for general \(\mathbb C\)-linear operators if and only if they hold for these operators seen as real: A \(\mathbb C\)-linear operator is strictly singular (resp. strictly cosingular, inessential) if and only if it is strictly singular (resp. strictly cosingular, inessential) as \(\mathbb R\)-linear.



Cites work









This page was built for publication: There is no largest proper operator ideal

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6093629)