A universal P P -group of weight ℵ \aleph
Jan van Mill · 2025
Read A universal P P -group of weight ℵ \aleph by Jan van Mill (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-15855.txt.
Introduction
Jan van Mill, writing alone, proves a universality result for a group built from a classic remainder space, under the Continuum Hypothesis. The space is the Čech-Stone remainder of the natural numbers, a large compactification used in topology and set theory. The group consists of all its homeomorphisms, equipped with the G-delta topology, and is a P-group of a stated large weight.
Catalog record
| CREATOR | Jan van Mill |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-15855
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A universal P P -group of weight ℵ \aleph Jan van Mill Abstract . We show that under the Continuum Hypothesis, the topological group of all homeomorphisms of the Čech-Stone remainder of ω \omega with the G δ G_{\delta} -topology, is a universal object for all P P -groups of weight at most 𝔠 {\mathfrak{c}} . 1991 Mathematics Subject Classification 54D35, 54H11, 22A05 1. Introduction All topological spaces under discussion are Tychonoff Uspenskiy [ 23 ] proved in 1986 that there is a universal topological group G with a countable base. Universal in the sense that every topological group with a countable base is (topologically isomorphic to) a subgroup of G . It is the homeomorphism group \mathcal{H}(Q) of the Hilbert cube Q=\mathbb{I}^{\omega} , endowed with the compact-open topology. (Another such group is the group of isometries of the Urysohn universal space, with the topology of pointwise convergence, [ 24 ] .) The obvious question from [ 24 ] – whether there are universal topological groups of uncountable weight – is still open. Natural candidates are the homeomorphism groups of the Tychonoff cubes \mathbb{I}^{\tau} , where \tau is uncountable. Uspenskiy’s proof in [ 23 ] is based on Keller’s Theorem [ 11 ] from 1931 that all compact convex and infinite-dimensional subsets of Banach spaces, are homeomorphic to Q . There are good topological characterizations of the Tychonoff cubes \mathbb{I}^{\tau} for uncountable \tau by S̆c̆epin [ 19 ] (see also [ 2 , 7.2.9] ). However, a necessary condition in these characterizations is that the spaces under consideration are Absolute Retracts, which is problematic for arbitrary compact convex subsets of \mathbb{I}^{\tau} . (There are even compact convex subsets of \mathbb{I}^{\omega_{1}} that do not satisfy the countable chain condition.) This blocks the attempt to prove that the spaces \mathcal{H}(\mathbb{I}^{\tau}) are universal for all uncountable \tau by a generalization of the arguments in Uspenskiy [ 23 ] . Shkarin [ 21 ] proved that there are universal Abelian topological groups with a countable base. And if \tau is an infinite cardinal such that 2^{\lambda}\leq\tau for every cardinal \lambda<\tau , then there is a group that is universal in the class of all Abelian topological groups of weight at most \tau . As usual, we denote the Čech-Stone compactification of the discrete space \omega by \beta\omega , and \omega^{*}=\beta\omega\setminus\omega , its Čech-Stone remainder. A compact, zero-dimensional F -space of weight {\mathfrak{c}} in which each non-empty G_{\delta} -subset has infinite interior will be called a Parovičenko space . Here an F -space is a space in which cozero-subsets are C^{*} -embedded. For normal spaces, this simply boils down to the statement that disjoint open F_{\sigma} -subsets have disjoint closures. Under the Continuum Hypothesis ( \mathsf{CH} ), all Parovičenko spaces are homeomorphic to \omega^{*} , as was shown by Parovičenko [ 16 ] . Subsequently, van Douwen and van Mill [ 4 ] obtained the converse: Parovičenko’s characterization of \omega^{*} implies \mathsf{CH} . Parovičenko also showed that under \mathsf{CH} , every compact space of weight at most {\mathfrak{c}} is a continuous image of \omega^{*} , hence \omega^{*} is universal in the sense of ‘mapping onto’. Dow and Hart [ 5 ] proved that under \mathsf{CH} , the Čech-Stone remainder [0,\infty)^{*} of the half line [0,\infty) is universal among the continua of weight {\mathfrak{c}} , again in the sense of ‘mapping onto’. (It is an open problem whether in some model there is a topological characterization of [0,\infty)^{*} in the style of Parovičenko.) We are interested in \mathcal{H}(\omega^{*}) , the homeomorphism group of \omega^{*} with the compact-open topology. Like \omega^{*} itself, it is a mysterious object. Rudin [ 18 ] proved that under \mathsf{CH} , its cardinality is 2^{\mathfrak{c}} . And, under the same condition, it is simple by Fuchino [ 8 ] . It even remains simple if \aleph_{2} Cohen reals are added to a model of \mathsf{ZFC} + \mathsf{CH} , [ 9 ] . But Shelah [ 20 ] proved that it is consistent that all homeomorphisms of \omega^{*} are ‘trivial’ in the sense that they are induced by bijections between cofinite subsets of \omega ; hence, in this model, \mathcal{H}(\omega^{*}) has cardinality {\mathfrak{c}} in contrast to Rudin’s result, and is by van Douwen [ 3 ] not simple, in contrast to Fuchino’s result. The goal of this note is to prove that \mathcal{H}(\omega^{*}) exhibits similar behavior within the class of suitable topological groups as \omega^{*} does within the class of suitable topological spaces. Following Arhangel’skii and Tkachenko [ 1 ] , we will call a topological group whose underlying topological space is a P -space – that is a space in which every G_{\delta} -subset is open – a P -group . If G is a topological group, then G_{(\delta)} denotes its G_{\delta} -modification . The underlying set of G_{(\delta)} is G , and the G_{\delta} -subsets of G form a basis (for the open subsets) of G_{(\delta)} . It is easy to see that G_{(\delta)} is a topological group, [ 1 , 3.6.H] , and a P -space, hence is a P -group. Theorem 1.1 . Let G be a topological group of weight {\mathfrak{c}} . Then there is a Parovičenko space X such that G_{(\delta)} is (topologically isomorphic to) a subgroup of \mathcal{H}(X) . This is in line with van Douwen’s unpublished result that every P -space of weight at most {\mathfrak{c}} embeds in \omega^{*} ; see [ 6 ] for details and references. Theorem 1.2 ( \mathsf{CH} ) . \mathcal{H}(\omega^{*})_{(\delta)} is universal for the class of all P -groups of weight at most {\mathfrak{c}} . The following is in line with the results in [ 16 , 4 ] that we discussed above. Theorem 1.3 . \mathsf{CH} is equivalent to the statement that for all Parovičenko spaces X and Y , \mathcal{H}(X) and \mathcal{H}(Y) are (algebraically) isomorphic. We also discuss some topological properties of the group \mathcal{H}(\omega^{*}) . We show, among other things, that \mathcal{H}(\omega^{*}) is not a P -group but that all of its countable subsets are closed and discrete. The class of ‘small’ P -groups is rather limited, so our results on universality are not a strong addition to what is known. But they may indicate that the homeomorphism group of \omega^{*} is by itself an interesting object that is worth studying. 2. Preliminaries 2.1. Topology For any space X , we let \tau X denote its topology, and \mathcal{H}(X) its group of homeomorphisms. We let e_{\mathcal{H}(X)} denote the identity element of the group \mathcal{H}(X) , that is, the identity function {\mathop{\operator@font id}\nolimits}_{X} of X . If no confusion is likely, we suppress the index. We use standard terminology for cardinal functions. For example, if X is a space, then w(X) and \chi(X) denote its weight and character, respectively (see Juhász [ 10 ] ). If X is a space, then \beta X and X^{*}=\beta X\setminus X denote its Čech-Stone compactification and Čech-Stone remainder, respectively. If X is normal, then disjoint closed subsets of X have disjoint closures in \beta X , [ 7 , 3.6.4] . In fact, this property characterizes \beta X . It is well-known, and easy to prove, that the Čech-Stone compactification \beta X of X is zero-dimensional iff X is strongly zero-dimensional; that