scientific article; zbMATH DE number 1076122

From MaRDI portal
Publication:4359564





There is in the literature a rich collection of theorems of this flavor: With suitable constraints on the topological groups \(G\) and \(H\), every measurable homomorphism \(h:G\to H\) is continuous. An early contribution in this vein of S. Banach (1932) achieves the conclusion for \(G\) and \(H\) separable metric and \(G\) complete with \(h\) taken ``measurable (B); \textit{K. Kuratowski} [Topologie (Warszawa, 1933; Zbl 0008.13202); p. 255] extended Banach's result to the case where \(h\) is Borel measurable in the usual modern sense, or even Baire measurable in the sense that for every \(V\) open in \(H\) there is \(U\) open in \(G\) such that \(U\Delta h^{-1} (V)\) is of category \(I\) in \(G\). Without citing Kuratowski's work, and with careful attention to Banach's unusual definition of measurable function, the present author proposes to elucidate and explicate Banach's result, thus in effect achieving Kuratowski's result with an additional hypothesis. From this he proposes to deduce a simplified proof of this result of \textit{D. Pickrell} [Proc. Am. Math. Soc. 102, 416-420 (1988; Zbl 0664.22014)]: If a representation \(\pi\) of the unitary group \(U({\mathcal H})\) of separable Hilbert space \({\mathcal H}\) into \(U(V)\) with \(V\) separable Hilbert space is continuous when \(U({\mathcal H})\) has its norm topology and \(U(V)\) its strong operator topology, then \(\pi\) remains continuous when both \(U({\mathcal H})\) and \(U(V)\) carry the strong operator topology. \{Reviewer's Remarks. (A) The author raises the question whether the use by \textit{C. C. Moore} [Trans. Am. Math. Soc. 221, 1-33 (1976; Zbl 0366.22005)] of the result of Banach (and Kuratowski) is logically secure and firmly grounded in the literature. Subsequent correspondence between the author and Professor Moore resolves this question completely in Moore's favor. In a jointly written ``Supplement to the present article which will appear in the same journal (J. Lie Theory), the present author and Professor Pickrell acknowledge this fact explicitly, and they achieve in addition a simplified statement and proof of yet another theorem of Banach-Kuratowski type. (B) In further correspondence with the author and the reviewer, Professor Moore has noted that the proposed proof of Pickrell's theorem, based on a purported isomorphism between two Borel fields, cannot be valid because in fact the fields in question have different cardinalities. With only a minor emendation, however, Pickrell's original proof (loc. cit.) remains valid\}.




Cited in
(19)








This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4359564)