A Provisional Solution to the Normal Moore Space Problem

Peter Nyikos

Proceedings of the American Mathematical Society · 1980 · 20 citations · 7 references

Concepts

Abstract

The Product Measure Extension Axiom (PMEA), whose consistency would follow from the existence of a strongly compact cardinal, implies that every normalized collection of sets in a space of character less than the continuum is well separated.Consistency of PMEA would thus solve many well-known problems of general topology, including that of whether every first countable normal space is collectionwise normal, as well as the normal Moore space problem.1. Introduction.Ever since it was raised by F. Burton Jones in 1933 [14], the problem of whether every normal Moore space is metrizable (often referred to simply as "the normal Moore space problem") has been one of the most famous and intensively researched problems in general topology.A few of the milestones in the vast literature on this problem are: [14] Jones's own article, which includes the proof that the axiom 2H° < 2Hl [now known to be independent of the usual (Zermelo-Fraenkel with Choice) axioms of set theory] implies that every separable normal Moore space is metrizable; [3] R. H. Bing's 1951 paper, which includes a proof that every collectionwise normal Moore space is metrizable and the first example of a normal space which is not collectionwise normal; [1], [2] the papers of P. S. Alexandroff and A. V. Arhangel'skiï giving various characterizations of metacompact Moore spaces (see §2) and raising the question of whether the normal ones are metrizable; [12] R. W. Heath's proof that the existence of a separable nonmetrizable normal Moore space is equivalent to the existence of a ß-set of real numbers, from which also follows the existence of a metacompact nonmetrizable normal Moore space; [16], [22] the Silver-Rothberger proof that the existence of Q-sct is consistent with (and therefore independent of) the usual axioms of set theory; and [9], [10] W. Fleissner's proof that GödePs Axiom of Constructibility implies every normal Moore space-indeed, every normal space of character < c-is collectionwise Hausdorff.A more detailed survey of the history of the problem may be found in the articles of Mary Ellen Rudin [15], [16] and Franklin Tall [22].Work on the normal Moore space problem has led to other questions which will probably be of greater interest to a general audience than the original problem.One is whether every first countable normal space is collectionwise normal.Since Moore spaces are first countable, an affirmative solution to this problem would

References

7