Weakly compact operators and strict topologies (Q392812): Difference between revisions

In this paper, \(X\) is a completely regular Hausdorff space, \(E\) a quasi-complete locally convex space with \(E'\) its topological dual, \(C_{b}(X)\) the space of all bounded, real-valued continuous functions on \(X\), \( \mathcal{B} \) the algebra generated by the zero-sets of \(X\) and \(\mathcal{B}a \) the Baire subsets of \(X\); also, \(M(X) = (C_{b}(X), \| \cdot \|)'\) and \( M_{t}(X), \; M_{\tau}(X)\), \(M_{\sigma}(X)\) are subsets of \(M(X)\) consisting of real-valued tight, \( \tau\)-smooth, and \(\sigma\)-smooth measures on \(X\). Using the fact that \(M(X)\) is a Dedekind complete vector lattice and \( M_{t}(X), \; M_{\tau}(X)\), \(M_{\sigma}(X)\) are projective bands in \(M(X)\), the purely finitely additive measures \(M_{pfa}(X)\) are defined to be the elements of the disjoint complement of \(M_{\sigma}(X)\) in \(M(X)\), the purely \(\sigma\)-additive measures \(M_{p\sigma a}(X)\) are defined to be the elements of the disjoint complement of \(M_{\tau}(X)\) in \(M_{\sigma}(X)\), and the purely \(\tau\)-additive measures \(M_{p\tau a}(X)\) are defined to be the elements of the disjoint complement of \(M_{t}(X)\) in \(M_{\tau}(X)\). Thus \(M(X)\) can be decomposed as: \[ M(X)= M_{pfa}(X) + M_{p\sigma a}(X) + M_{p\tau a}(X) + M_{t}(X). \] In this paper, the author extends this result to \(E\)-valued measures. As proved in [\textit{S. S. Khurana}, Georgian Math. J. 14, No. 4, 687--698 (2007; Zbl 1154.46025)], every weakly compact operator \(m: C_{b}(X) \to E \), just as in the real-valued case, gives a regular, finitely additive, \(E\)-valued measure on \( \mathcal{B} \) and the space of all such measures has natural subspaces of \(\sigma\)-smooth, \( \tau\)-smooth, and tight measures; these measures, coming from weakly compact operators, are also strongly bounded (they are also called exhaustive). The measure \(m\) is called purely finitely additive if \(e' \circ m \in M_{pfa}(X)\) for all \(e' \in E'\) and similar meanings for purely \(\sigma\)-additive measures and purely \(\tau\)-additive measures. The main result of the paper is: If \(m\) is the representing measure of a weakly compact operator \(m: C_{b}(X) \to E \), then \(m\) can be uniquely decomposed as \(m=m_{1}+m_{2}+m_{3}+m_{4}\), where \(m_{1} \in M_{t}(X)\), \(m_{2}\) is a purely \(\tau\)-additive measure, \(m_{3}\) is a purely \(\sigma\)-additive measure, and \(m_{4}\) is a purely finitely additive measure.