An ordinal indexing on the space of strictly singular operators (Q532602): Difference between revisions

Let \(\mathcal{S}\mathcal{S}(X,Y)\) be the space of strictly singular operators from a Banach space \(X\) to a Banach space \(Y\). \textit{G.~Androulakis, P.~Dodos, G.~Sirotkin} and \textit{V. G.~Troitsky} [Isr. J. Math. 169, 221--250 (2009; Zbl 1168.47018)] defined for each \(1\leq \xi<\omega_1\) a subset \(\mathcal{S}\mathcal{S}_\xi(X,Y)\subset \mathcal{S}\mathcal{S}(X,Y)\) of \(\mathcal{S}_\xi\)-\textit{strictly singular} operators. In the paper under review, the author defines the ordinal index \[ i(\mathcal{S}\mathcal{S}(X,Y))= \sup_{T\in\mathcal{S}\mathcal{S}(X,Y)} \inf\{\xi:T\in\mathcal{S}\mathcal{S}_\xi(X,Y)\}. \] The author proves, in particular, that if \(X\) and \(Y\) are separable and the unit ball \(\mathcal{B}(\mathcal{S}\mathcal{S}(X,Y))\) is a Borel subset of \(\mathcal{B}(\mathcal{L}(X,Y))\) in the strong operator topology, then \(i(\mathcal{S}\mathcal{S}(X,Y))<\omega_1\). He asks if a converse statement is valid. The author applies his results to hereditarily indecomposable spaces.