Publications

Article: Dowker spaces revisited
[Show abstract] [Hide abstract]
ABSTRACT: In 1951, Dowker proved that a space X is countably paracompact and normal if and only if X ×I is normal. A normal space X is called a Dowker space if X × I is not normal. The main thrust of this article is to extend this work with regards αnormality and βnormality. Characterizations are given for when the product of a space X and (ω + 1) is αnormal or βnormal. A new definition, αcountably paracompact, illustrates what can be said if the product of X with a compact metric space is βnormal. Several examples demonstrate that the product of a Dowker space and a compact metric space may or may not be αnormal or βnormal. A collectionwise Hausdorff Moore space constructed by M. Wage is shown to be αnormal but not βnornal.08/2010; 
[Show abstract] [Hide abstract]
ABSTRACT: The general question, “When is the product of Fréchet spaces Fréchet?” really depends on the questions of when a product of α4 Fréchet spaces (also known as strongly Fréchet or countably bisequential spaces) is α4, and when it is Fréchet. Two subclasses of the class of strongly Fréchet spaces shed much light on these questions. These are the class of α3 Fréchet spaces and its subclass of ℵ0bisequential spaces. The latter is closed under countable products, the former not even under finite products. A number of fundamental results and open problems are recalled, some further highlighting the difference between being α3 and Fréchet and being ℵ0bisequential.Topology and its Applications 06/2010; 157(8):14851490. DOI:10.1016/j.topol.2009.07.014 · 0.59 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: The contrast between the effects of the Continuum Hypothesis (CH) and those of MA+¬ CH is well known. It is also known that the contrast is greatly magnified when CH is strengthened to the combinatorial axiom ♦ while MA+¬CH is strengthened to the Proper Forcing Axiom (PFA). U. Abraham and S. Todorčević [Fundam. Math. 152, No. 2, 165–181 (1997; Zbl 0879.03015)] obtained a number of PFAlike applications of a simple combinatorial axiom which they showed to be compatible with CH. Later, S. Todorčević published [Fundam. Math. 166, No. 3, 251–267 (2000; Zbl 0968.03049)] new applications of a very similar but more powerful axiom, compatible with CH, but requiring the use of large cardinals. These axioms are simpler than the Dcompleteness machinery used to establish their consistency with CH (see [S. Shelah, Proper and improper forcing. 2nd ed., Perspectives in Mathematical Logic. Berlin: Springer (1998; Zbl 0889.03041)]). In this paper, the authors give new applications of these two axioms and also introduce and apply new axioms related to them. The authors’ summary: “We investigate some combinatorial statements that are strong enough to imply that ♦ fails (hence the name antidiamonds); yet most of them are also compatible with CH. We prove that these axioms have many consequences in settheoretic topology, including the consistency, modulo large cardinals, of a Yes answer to a problem on linearly Lindelöf spaces posed by Arhangel’skiĭ and Buzyakova (1998).”Transactions of the American Mathematical Society 11/2009; 361(11). DOI:10.1090/S0002994709047059 · 1.10 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: We explore the relation between two general kinds of separation properties. The first kind, which includes the classical separation properties of regularity and normality, has to do with expanding two disjoint closed sets, or dense subsets of each, to disjoint open sets. The second kind has to do with expanding discrete collections of points, or fullcardinality subcollections thereof, to disjoint or discrete collections of open sets. The properties of being collectionwise Hausdorff (cwH), of being strongly cwH, and of being wD(ℵ1), fall into the second category. We study the effect on other separation properties if these properties are assumed to hold hereditarily. In the case of scattered spaces, we show that (a) the hereditarily cwH ones are αnormal and (b) a regular one is hereditarily strongly cwH iff it is hereditarily cwH and hereditarily βnormal. Examples are given in ZFC of (1) hereditarily strongly cwH spaces which fail to be regular, including one that also fails to be αnormal; (2) hereditarily strongly cwH regular spaces which fail to be normal and even, in one case, to be βnormal; (3) hereditarily cwH spaces which fail to be αnormal. We characterize those regular spaces X such that X×(ω+1) is hereditarily strongly cwH and, as a corollary, obtain a consistent example of a locally compact, first countable, hereditarily strongly cwH, nonnormal space. The ZFCindependence of several statements involving the hereditarily wD(ℵ1) property is established. In particular, several purely topological statements involving this property are shown to be equivalent to b=ω1.Topology and its Applications 12/2008; 156(2):151164. DOI:10.1016/j.topol.2008.05.021 · 0.59 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: If X is a separable 0dimensional metrizable space in which every compact subset is countable, then C(X) with the compactopen topology is stratifiable iff X is scattered. This answers a question of Gruenhage and lends credence to a conjecture of Gartside and Reznichenko.Topology and its Applications 04/2007; 154(7):14891492. DOI:10.1016/j.topol.2005.11.020 · 0.59 Impact Factor 
Chapter: Classes of compact sequential spaces
[Show abstract] [Hide abstract]
ABSTRACT: Without Abstract11/2006: pages 135159; 
Article: A nonmetrizable collectionwise Hausdorff tree with no uncountable chains and no Aronszajn subtrees
[Show abstract] [Hide abstract]
ABSTRACT: It is independent of the usual (ZFC) axioms of set theory whether every collec tionwise Hausdorff tree is either metrizable or has an uncountable chain. We show that even if we add "or has an Aronszajn subtree," the statement remains ZFCindependent. This is done by constructing a tree as in the title, using the settheoretic hypothesis ♦∗, which holds in Godel's Constructible Universe.01/2006; 47(3). 
Article: ω 1 compactness in type I manifolds
[Show abstract] [Hide abstract]
ABSTRACT: In this paper we establish that any wellpruned ω1tree, T, admits an ω1compact type I manifold if T does not contain an uncountable antichain. If T does contain an uncountable antichain, it has been shown that whether or not T admits an ω1compact manifold is undecidable in ZFC.Topology and its Applications 02/2005; 148(1):165171. DOI:10.1016/j.topol.2004.08.006 · 0.59 Impact Factor 
Article: Various topologies on trees
[Show abstract] [Hide abstract]
ABSTRACT: This is a survey article on trees, with a modest number of proofs to give a flavor of the way these topologies can be efficiently handled. Trees are defined in settheorist fashion as partially ordered sets in which the elements below each element are wellordered. A number of different topologies on trees are treated, some at considerable length. Two sections deal in some depth with the coarse and fine wedge topologies, and the interval topology, respectively. The coarse wedge topology gives a class of supercompact monotone normal topological spaces, and the fine wedge topology puts a monotone normal, hereditarily ultraparacompact topology on every tree. The interval topology gives a large variety of topological properties, some of which depend upon settheoretic axioms beyond ZFC. Many of the open problems in this area are given in the last section. 
[Show abstract] [Hide abstract]
ABSTRACT: We build a model of ZFC+CH in which every first countable, countably compact space is either compact or contains a homeomorphic copy of ω 1 with the order topology. The majority of the paper consists of developing forcing technology that allows us to conclude that our iteration adds no reals. Our results generalize the iteration theorems appearing in Chapters V and VIII of [19] as well as the iteration theorem appearing in [9]. We close the paper with a ZFC example (constructed using Shelah's club–guessing sequences) that shows similar results do not hold for closed pre–images of ω 2 .Transactions of the American Mathematical Society 01/2005; DOI:10.2307/3845191 · 1.10 Impact Factor 
Topology and its Applications 03/2004; 138(s 1–3):325–327. DOI:10.1016/j.topol.2003.11.004 · 0.59 Impact Factor

[Show abstract] [Hide abstract]
ABSTRACT: Strong αfavorability of the compactopen topology on the space of continuous functions, as well as of the generalized compactopen topology on continuous partial functions with closed domains is studied.01/2003; 1(1). 
Article: Applications of some strong settheoretic axioms to locally compact T 5 and hereditarily scwH spaces
[Show abstract] [Hide abstract]
ABSTRACT: The author continues a detailed analysis of the structure of locally compact spaces, in some sense, finding properties that ensure that an increasing ω 1 union of simple subspaces will again be simple. The settheoretic assumptions increase from Martin’s Axiom, to PFA, and advance in Section 3 to conditions that are not known to be consistent. Unfortunately the axiom scheme SSA + PFA seems not to be consistent as can be seen by a new result of P. Larson. While the results (methods) of Section 3 and 4 are worth reading, the careful reader will need to be mindful of this settheoretic caveat. The two main results give the general flavor of this article. Theorem 1, a consequence of MA(ω 1 ), states that a locally compact hereditarily strongly collectionwise Hausdorff and countably compact space can be written as an increasing chain of open sets, the larger containing the closure of the smaller, each is countably compact with hereditarily Lindelöf boundary. Section 2 basically contains a very nice proof of Theorem 2: PFA implies that if a space X is hereditarily strongly collectionwise Hausdorff and maps perfectly onto ω 1 , then one may remove some finite union of disjoint copies of ω 1 from X, leaving behind a paracompact subspace.Fundamenta Mathematicae 01/2003; 176(1):2545. DOI:10.4064/fm17613 · 0.45 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: A manifold is a connected Hausdorfi space in which every point has a neighborhood homeomorphic to Euclidean nspace (n is unique). A space is collec tionwise Hausdorfi (cwH) if every closed discrete subspace D can be expanded to a disjoint collection of open sets each of which meets D in one point. There are exactly two examples of 1dimensional nonmetrizable hereditarily normal, hereditarily cwH manifolds: the long line and the long ray. The main new result is that if it is con sistent that there is a supercompact cardinal, it is consistent that every hereditarily normal, hereditarily cwH manifold of dimension greater than 1 is metrizable. The modern settheoretic topology of manifolds can be said to begin with the 1975 construction by Mary Ellen Rudin of a nonmetrizable perfectly normal man ifold using the settheoretic axiom }. This solved a problem that had been posed by Wilder at the end of his 1949 textbook (11) and thereby made it possible to con sistently extend the wealth of algebraic topology techniques used by Wilder beyond the context of metrizable manifolds, at least consistently with the usual axioms of ZFC. Shortly thereafter, with the help of Phillip Zenor (7), Rudin was able to reduce the settheoretic axiom to the more familiar Continuum Hypothesis (CH). Then, in 1978, Rudin showed that the existence of perfectly normal nonmetriz able manifolds was independent of the usual axioms of set theory (6), by showing that they do not exist under MA(!1). A very natural question is whether \per fectly normal" can be generalized to \hereditarily normal" (= completely normal), especially if one is aware of the old custom of designating perfectly normal spaces as T6 spaces and completely normal ones as T5 spaces. However, the long ray and long line show that the straightforward generalization of Rudin's theorem cannot hold. They are linearly ordered (hence hereditarily normal) spaces that are locally like the real line but contain copies of !1 and hence are not metrizable. In higher dimensions, however, it is a completely difierent story, and the following is still unsolved.Topology and its Applications 08/2002; 123(1). DOI:10.1016/S01668641(01)00181X · 0.59 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: The main problem in this article is one of my alltime favorites. To drum up interest in it, I announced at the 1986 Prague Topological Symposium (Toposym) that I was offering a US$500 prize for a solution during the following ten years [9]. There was essentially no progress on the problem all during those ten years, and so at the 1996 Toposym I raised the award to US$1000 during the following ten years. Those ten years have almost passed with no progress on the problem at all to the best of my knowledge, and I am hereby removing all time limits on the $1000 award and am contemplating raising it. Here is the problem that is the focus of all this largesse: Problem 1. Does ZFC imply the existence of a separable, first countable, count1001 ? ably compact, noncompact Hausdorff (T 2) space? A mild putdown of general topology one hears from time to time is that there are too many adjectives in a typical problem or theorem. For me, however, one of the charms of general topology is that there are so many theorems and problems one can understand with no more than a typical undergraduate textbook in general topology as a resource. The adjectives used here definitely fall under that heading; the concepts are like second nature to many of us, and I have little mental pictures that I associate to each one to help keep arguments straight. I will soon cut down on the number of adjectives in the alternative wording below, but the ones in the original wording are all implicitly there. The usual topology on ω 1 satisfies everything except separability. The Novak–Teresaka space described in Vaughan's article [15] satisfies everything except first countability. If one refines the cofinite topology on ω 1 by making initial segments open, then the resulting space satisfies everything except T 2 and is T 1 . The remaining two properties are obviously necessary also to have an open problem. Also, the question mentions ZFC because there is a multitude of consistent examples of spaces as in Problem 1; see Sections 1 and 2. In fact, Problem 1 is one of a small but growing number of topological problems for which a negative answer is known to entail (2 ω =)c ≥ ℵ 3 , yet c = ℵ 3 has not been ruled out. See Section 4. 
[Show abstract] [Hide abstract]
ABSTRACT: Metrizability is an extremely strong property where trees are concerned, and it turns out that in many ways, monotone normality is the appropriate generalization when the trees have uncountable chains. We show that monotone normality is equivalent to the tree being the topological direct sum of ordinal spaces, each of which is a convex chain in the tree. Several metrization theorems are proven, some in ZFC, some just assuming ZF or “ZF + Countable AC”, and still others assuming ZFCindependent axioms, as well as theorems in a similar spirit with monotone normality of the tree as a conclusion. The property of being collectionwise Hausdorff plays a key role, and we obtain partial results on the still unsolved problems of whether it is consistent that every collectionwise Hausdorff tree or every normal tree is monotone normal.Topology and its Applications 11/1999; 98(1):269290. DOI:10.1016/S01668641(99)000449 · 0.59 Impact Factor 
Article: Gently Killing Sspaces
[Show abstract] [Hide abstract]
ABSTRACT: We produce a model of ZFC in which there are no locally compact first countable Sspaces, and in which 2^{aleph_0}<2^{aleph_1}. A consequence of this is that in this model there are no locally compact, separable, hereditarily normal spaces of size aleph_1, answering a question of the second author.Israel Journal of Mathematics 01/1999; 136(1). DOI:10.1007/BF02807198 · 0.66 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: Classical characterizations of four separable metrizable spaces are recalled, and generalized to classes of spaces which admit a uniformity with a totally ordered base. The AlexandroffUrysohn characterization of the irrationals finds its closest analogues for strongly inaccessible cardinals, while the other three spaces, including the Cantor set, find their most natural analogues for weakly compact cardinals. In addition, A.H. Stone's characterization of Baire's zerodimensional spaces is extended to give internal characterizations of all spaces γλ × D, where D is discrete and γλ has the initial agreement topology. The historical background for the AlexandroffUrysohn result is briefly surveyed.Topology and its Applications 01/1999; 91(1):123. DOI:10.1016/S01668641(97)002393 · 0.59 Impact Factor 
[Show abstract] [Hide abstract]
ABSTRACT: In 1975 E. K. van Douwen showed that if (Xn)n2! is a family of Hausdor spaces such that all nite subproducts Q n 
Article: Hereditary normality of γ N spaces
[Show abstract] [Hide abstract]
ABSTRACT: A γNspace is a locally compact Hausdorff space with a countable dense set of isolated points, and the rest of the space homeomorphic to ω1. We show that under the Open Coloring Axiom (OCA) no γNspace is hereditarily normal. This is the key to showing that some sweeping statements are consistent with (and independent of) the usual axioms of set theory, including: 1.(1) Every countably compact, hereditarily normal space is sequentially compact.2.(2) Every separable, hereditarily normal, countably compact space is compact and FréchetUrysohn.3.(3) The arbitrary product of countably compact, hereditarily normal spaces is countably compact. Not all of these conclusions follow just from MA + ¬ CH: a forcing construction is given of a model of MA + c = κ where κ is any cardinal ⩾ ℵ2 satisfying κ = 2Topology and its Applications 07/1995; 65(1):919. DOI:10.1016/01668641(94)00004M · 0.59 Impact Factor

Questions & Answers

Open Reviews

Research Feedback