Generalized metric spaces and mappings. Translated from the Chinese by the first author (Q310896): Difference between revisions

Already seventeen years ago, in my review of \textit{R. E. Hodel}'s article on the history of generalized metrizable spaces [in: Handbook of the history of general topology. Volume 2. Dordrecht: Kluwer Academic Publishers. 541--576 (1998; Zbl 0940.54003)], I remarked that ``the article covers most of the important developments in the theory of generalized metrizable spaces to a certain extent, but for a complete exposition a book would be necessary.'' At the time, I did not yet know that the first edition of Shou Lin and Ziqiu Yun's book on this very topic had already appeared in 1995, albeit in Chinese. With the publication of the expanded and translated third edition in 2016, Lin and Yun's book now holds the distinction of being the (too) long awaited first book on generalized metric spaces available in English, too. Since there is no generally accepted definition of what actually constitutes a generalized metric space, there is a large variety of classes of topological spaces and their interrelationships that could be studied in a book on this topic. Following the approach proposed by Alexandroff and Arhangel'skiĭ the authors concentrate on the following questions: (1) To what extent are generalized metric spaces images or pre-images of metric spaces? (2) Which classes of generalized metric spaces behave nicely with respect to continuous mappings? The book consists of three chapters, two appendices, and an extensive list of references with 496 entries; altogether on 328 pages. Chapter 1, entitled ``The Origin of Generalized Metric Spaces'' (51 pages), identifies the following (sometimes rather technical) concepts that have led to the definition of various classes of generalized metric spaces (in most cases spaces are assumed to be at least Hausdorff). (1) Modifications of distance functions, which lead to semi-metrizable and symmetrizable spaces. (2) Conditions on bases that are motivated by the Bing-Nagata-Smirnov metrization theorem, which result in the definition of Ceder's \(M_{i}\)-spaces. (3) Stratifications, which are used in the definition of stratifiable spaces, semi-stratifiable spaces, and \(k\)-semi-stratifiable spaces. (4) (Pseudo-)networks and (mod \(k\))-networks, which lead to \(\sigma^{\#}\)-spaces, \(\sigma\)-spaces, cosmic spaces, and strong \({\Sigma}\)-spaces. (5) \(k\)-networks and weak bases, which appear in the definition of \(\aleph_{0}\)-spaces, \(\aleph\)-spaces, and \(g\)-metrizable spaces. (6) Generalizations of countably compact spaces, which result in \(M\)-spaces, \(wM\)-spaces, \(w\Delta\)-spaces, \({\beta}\)-spaces, and \(q\)-spaces. Most of these generalized metric spaces are revisited in greater detail in Chapter 3. The central theme of Chapter 2 is aptly expressed in its title ``Mappings on Metric Spaces'': the authors provide quite a comprehensive collection of intrinsic characterizations of topological spaces that are images or pre-images of metric spaces under certain continuous mappings (94 pages). They start with the classical result of Hanai, Morita, and Stone that perfect mappings preserve metrizability, and prove that the perfect pre-images of metric spaces are precisely the paracompact \(M\)-spaces in the sense of Morita as well as the paracompact \(p\)-spaces in the sense of Arhangel'skiĭ. In the next section, sequential spaces are shown to be the quotient images of metric spaces. Similarly, Fréchet-Urysohn spaces are the pseudo-open images of metric spaces, and strongly Fréchet-Urysohn spaces are the bi-quotient images of metric spaces. Of course, the classical theorem of Ponomarev, that first countable spaces are the open images of metric spaces, is also presented. Following Foged, closed images of metric spaces are characterized as regular Fréchet-Urysohn spaces with a \(\sigma\)-hereditarily closure preserving \(k\)-network. The next sections of Chapter 2 contain characterizations of quotient \(s\)-images, pseudo-open \(s\)-images, open \(s\)-images, and closed \(s\)-images of metric spaces, where a continuous mapping \(f:X \rightarrow Y\) is said to be an \(s\)-mapping if \(f^{-1}(y)\) is separable for every \(y \in Y\). They also contain characterizations of quotient \(ss\)-images, pseudo-open \(ss\)-images, and closed \(ss\)-images of metric spaces, where \(ss\)-mappings are a certain modification of \(s\)-mappings. In the final sections of Chapter 2, some of the generalized metric spaces that were introduced in Chapter 1 are shown to be images of metric spaces under suitable mappings. In particular, developable spaces are the open \(\pi\)-images of metric spaces, and semi-metrizable spaces are the pseudo-open \(\pi\)-images of metric spaces. Moreover, metacompact developable spaces, i.e., spaces with a uniform base, are the open compact images of metric spaces, and regular \(\sigma\)-spaces are the images of metric spaces under \(\sigma\)-locally finite mappings. Chapter 3 is simply entitled ``Generalized Metric Spaces'' (112 pages). It consists of ten sections, nine of which are entirely devoted to exactly one class of generalized metric spaces (or one defining concept) and closely related variants. The focus is on characterizing these spaces, their interrelationships, their role in metrization theorems, and on the question whether they are preserved by closed mappings. The chapter begins with some results on spaces with point-countable covers. For example, it is shown that a topological space is metrizable if and only if it is a paracompact \(M\)-space with a point-countable pseudo-\(k\)-network. Moreover, the property of having a point-countable pseudo-\(k\)-network is shown to be preserved by perfect mappings. In the next section, \(\Sigma\)-spaces and their variants strong \(\Sigma\)-spaces, (strong) \(\Sigma^{*}\)-spaces, and (strong) \(\Sigma^{\#}\)-spaces are studied in detail. For instance, it is shown that every strong \(\Sigma^{*}\)-space with countable pseudo-character is a strong \(\Sigma\)-space, and that every space with a \(\sigma\)-cushioned (mod \(k\))-network, in particular every strong \(\Sigma^{\#}\)-space, is a \(D\)-space. Next, \(\sigma\)-spaces and semi-stratifiable spaces are investigated. Both classes, which the authors claim to be ``regarded as the most successful generalizations of metric spaces'', behave particularly well with respect to mappings. A section on \(k\)-semi-stratifiable spaces follows -- a subclass of the class of \(\sigma\)-spaces -- which have been studied mainly by Chinese topologists. Some of the results appear here for the first time in English. The next section is on Ceder's \(M_{i}\)-spaces. It contains the proof (following Junnila) that \(M_{2}\)-spaces and \(M_{3}\)-spaces (= stratifiable spaces) coincide. Mizokami's characterization of \(M_{1}\)-spaces as the paracompact \(\sigma\)-spaces in which every closed subset has a closure-preserving open neighborhood base is mentioned, but, because of its complexity, presented without proof. Whether \(M_{1}\)-spaces and \(M_{2}\)-spaces coincide is a problem that has now been open for 57 years. The main result of the next section on developable spaces and \(p\)-spaces is Jiang's theorem that every space with a strict \(p\)-sequence is a submetacompact space. Its consequence is a nice characterization of developable spaces as strict \(p\)-spaces with a \(G_{\delta}\)-diagonal. \(M\)-spaces and their generalizations \(M^{*}\)-spaces, \(M^{\#}\)-spaces, \(wM\)-spaces, and \(w\Delta\)-spaces are studied in the following section. There are further sections on \(\aleph\)-spaces and on \(g\)-metrizable spaces. The final section of Chapter 3 contains more than 60 open questions concerning generalized metric spaces. A nice touch is that nearly every section is accompanied by illustrative (counter)examples. The first of the two appendices discusses covering properties that are used in the preceding chapters and provides full proofs of characterizations of paracompact spaces, subparacompact spaces, metacompact spaces, submetacompact (= \(\theta\)-refinable) spaces, and meta-Lindelöf spaces (18 pages). The second appendix offers the authors' personal view of the history of generalized metric spaces (24 pages). Altogether this is an interesting book. Most of the proofs that are presented appear in a textbook (in English) for the first time. Anyone who intends to delve deeper into this area of set-theoretic topology should consider to start with this book. In writing the first book on generalized metric spaces, the authors have done a great job. Unfortunately, this cannot be said about the publisher. As mentioned above, the text has been translated from Chinese to English. And although the book is fairly readable, the text could have benefitted a lot from professional copy-editing. In my opinion it is the publishing house's responsibility to ensure that its books are grammatically and linguistically correct; at least this has traditionally been the case. Otherwise, what are we paying for when buying a book? Would we recommend this book to students as an example of good writing? Certainly not.