3
$\begingroup$

I am looking for a space as in the title, i.e.,

Is there a metacompact, normal, CCC space which is not Lindelof?

A space is ccc iff any family of pairwise disjoint open sets is at most countable.

A space $X$ is metacompact iff for any open cover $\mathcal U$ of $X$ there is a point finite refintement of $\mathcal U$.

Thanks.

$\endgroup$
5
$\begingroup$

Yes, such an example can be obtained using the Pixley-Roy hyperspace construction.

Given a topological space $X$, the Pixley-Roy hyperspace on $X$ ($PR(X)$) is defined as the space of all non-empty finite subsets of $X$ with the topology generated by sets of the form $[F,U]:=\{G \in PR(X): F \subset G \subset U \}$, where $F \in PR(X)$ and $U$ is an open subset of $X$ which contains $F$.

It is easy to see that $PR(X)$ is Lindelof if and only if $X$ is countable. Indeed, when $X$ is uncountable, the family $\{[\{x\}, X]: x \in X \}$ provides an open cover without countable subcovers. It's also easy to see that if $X$ has countable weight then $PR(X)$ is ccc.

Eric van Douwen proved that $PR(X)$ is always metacompact (provided $X$ is $T_1$) and that $PR(X)$ is a Moore space when $X$ is first-countable.

So for every uncountable set of reals $Z \subset \mathbb{R}$, the space $PR(Z)$ is not Lindelof, but it's Tychonoff, metacompact, ccc and a Moore space (you didn't ask for the latter, but it's a nice bonus to have! ;-)).

Only missing is normality. To get it we can use a standard trick, originally due to Bob Heath, which goes back to the Normal Moore Space Problem. Let $Y$ be a $Q$-set (that is an uncountable subset of the reals whose every subset is a relative $G_\delta$) with the additional property that $Y^n$ is a $Q$-set, for every $n \in \mathbb{N}$ (the existence of such a set of reals is consistent with and independent from ZFC). Then we can prove that $PR(Y)$ is normal.

For the proofs see the survey Lutzer, David J., Pixley-Roy topology, Topology proceedings, Vol. 3, No.1, Proc. Topol. Conf. Univ. Okla. 1978, 139-158 (1979). ZBL0435.54007.).

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.