Banach spaces of continuous functions as dual spaces (Q267835): Difference between revisions

The Banach spaces \(C(K)\) of continuous real- or complex-valued functions on a compact Hausdorff space \(K\) have been in the centre of interest ever since the inception of Banach space theory. Key results in the early period include the Banach-Mazur theorem (on universality of \(C[0,1]\)) and the Banach-Stone theorem (on isometrical isomorphisms between \(C(K)\)-spaces), to be followed, in the early 1950s, by the theorem due to Nachbin, Goodner, Kelley, and (in the complex case) Hasumi describing which target spaces can replace the scalar field in the Hahn-Banach extension theorem. This class of spaces, called \(1\)-injective spaces, turned out to consist of exactly those \(C(K)\)-spaces with \(K\) being Stonean (also known as extremally disconnected), which is the somewhat bizarre property that any two disjoint open subsets have disjoint closures. Stonean spaces also appear in the Dixmier-Grothendieck characterisation of those \(C(K)\)-spaces that are isometrically isomorphic to a dual Banach space; this happens if and only if \(K\) is a Stonean space and the normal measures (i.e., order continuous functionals) separate the elements of \(C(K)\). Such a compact space is called hyper-Stonean, and in this case the predual of \(C(K)\) is uniquely determined in a strong sense. In the 1970s, these results were systematically expounded in the by now classical books by \textit{H.E.~Lacey} [The isometric theory of classical Banach spaces. Berlin-Heidelberg-New York: Springer-Verlag (1974; Zbl 0285.46024)], \textit{H.H.~Schaefer} [Banach lattices and positive operators. Berlin-Heidelberg-New York: Springer-Verlag (1974; Zbl 0296.47023)] and \textit{Z.~Semadeni} [Banach spaces of continuous functions. Warszawa: PWN -- Polish Scientific Publishers (1971; Zbl 0225.46030)], but there are hardly any more recent references in the monograph literature dealing with those topics. Hence, the authors of the volume under review have set out to give a fresh outlook on the theory of \(C(K)\)-spaces with a special view on their status as dual or bidual Banach spaces, both isometrically and isomorphically. In doing so they pay reverence to the late William G. Bade whose unpublished notes [\textit{W.G.~Bade}, The space of all continuous functions on a compact Hausdorff space. Berkeley: University of California (1957)], [\textit{W.G.~Bade}, The Banach space \(C(S)\). Aarhus: Matematisk Institut, Aarhus Universitet (1971; Zbl 0224.46026)] influenced many researchers back in the 1960s and 1970s; their book also owes a lot to these notes. As is apparent from the brief synopsis above, the key notions involved are of topological and measure theoretic nature. Indeed, Stonean spaces naturally appear as Stone spaces of complete Boolean algebras. Therefore, to address the themes indicated above it is necessary to study topological notions, Boolean algebras, items in measure theory and of course Banach spaces, Banach lattices and \(C^*\)-algebras. A good portion of the text is devoted to these topics before \(C(K)\)-spaces are studied in detail in Chapter~6, including Banach spaces of Baire functions. A list of open questions concludes the book. The material is presented in a meticulous and extremely well polished way. Readers will especially appreciate the detailed references that accompany the text. As a matter of fact, very few results, e.g., concerning the RNP, are only quoted and not proved. Researchers of Banach spaces will profit a lot from this very welcome contribution to the literature.