Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

In the paper A survey of D-spaces by Gary Gruenhage it is written that it is easily seen that any countably compact D-space is compact. However I'm not able to show it. Here is my attempt to prove this claim: Let $(X, \tau)$ be a countably compact D-space with topology $\tau$. Consider a function $N: X \rightarrow \tau$ with $x \in N(x)$ for every $x \in X$ (neighborhood assignment function). I think that to prove compactness of $X$ it is enough to show that open cover $\{N(x):x \in X\}$ of $X$ has finite subcover. Because $X$ is a D-space, there is a closed discrete subset $D$ of $X$ such that $\{N(x):x \in D\}$ covers $X$. It seem to me that to complete the proof (using countable compactness of $X$) I need to use the fact that $D$ is countable. However it is not true in general that close discrete set is countable. Thank you for your help! Michal Čihák

Note 1. The considered spaces are regular and $T_1$.

Note 2. A space $X$ is a D-space if whenever one is given a neighborhood $N(x)$ of $x$ for each $x \in X$, then there is a closed discrete subset $D$ of $X$ such that $\{N(x):x \in D\}$ covers X.

share|cite|improve this question
This looks like a duplicate of this question... is anyone for or against closing? – Zev Chonoles Mar 30 '12 at 19:00
@Zev: of course, they are related and the ideas are similar, but $\sigma$-compactness and countable compactness are quite distinct properties, so I'm against closing this question. – t.b. Mar 30 '12 at 19:17
@t.b.: Ah, I didn't read closely enough - nor did the person who flagged the question :) Thanks for pointing out the difference. – Zev Chonoles Mar 30 '12 at 19:19
@ZevChonoles Also, the implication goes the other direction in the linked question. – Alex Becker Mar 30 '12 at 20:47
up vote 1 down vote accepted

Suppose that your set $D$ is infinite. Pick a countably infinite subset of it, say $D_0$. For each $x\in D_0$ let $V(x)$ be a nbhd of $x$ such that $V(x)\cap D_0=\{x\}$; you can do this because $D_0$ is discrete. Moreover, $D_0$ is closed, so $X\setminus D_0$ is open, and $\{X\setminus D_0\}\cup\{V(x):x\in D_0\}$ is then a countable open cover of $X$ that clearly has no finite subcover: each $x\in D_0$ is covered only by the set $V(x)$. This contradicts the countable compactness of $X$, showing that in fact $D$ must be finite.

You are correct that it’s sufficient to look at open covers of the form $\{N(x):x\in X\}$ for nbhd assignments $N$. To see this, let $\mathscr{U}$ be any open cover of $X$. Then for each $x\in X$ let $N(x)$ be any element of $\mathscr{U}$ that contains $x$. Clearly $\{N(x):x\in X\}$ is a subcover of $\mathscr{U}$, so any finite subcover of $\{N(x):x\in X\}$ will automatically be a finite subcover of the original $\mathscr{U}$ as well.

share|cite|improve this answer
Thank you very much, your proof is very clear for me. – mcihak Mar 31 '12 at 6:58

It follows from the following characterization: a space $X$ countably compact if and only if every infinite subset of $X$ has an accumulation point. So in particular any discrete subspace is finite. By the way the characterization makes for a nice exercise.

share|cite|improve this answer

Note that any countable compact space $X$ is limit point compact. Now let $D \subset X$ be a closed discrete subset of $X$. Then $D$ does not have any limit points, since such a limit point would have to be in $X$, but $D$ is closed. Finally since $X$ is limit point compact it follows that $D$ is finite.

Let $\mathcal U$ be an open cover of $X$. Then for each $x \in X$ pick $U \in \mathcal U$ such that $x \in U$ and set $N(x)=U$. Now if $D$ is a closed discrete subspace of $X$ such that for $x \in D$ we have $N(X)$ covers $D$, then by the previous argument $D$ is finite, so this is a finite subcover of $\mathcal U$.

I am curious though if you can manage to do this without using the axiom of choice, since I seem to be invoking it in order to create the neighborhood function.

share|cite|improve this answer

We can do even better than making your set $D$ countable; $D$ is finite! Suppose your set $D$ is infinite, and choose some (necessarily closed) countably infinite discrete subset $D'$ of $D$. Define a neighborhood assignment function $N':D'\to \tau$ such that $x\neq y\implies N'(x)\cap N'(y)\neq \emptyset$, which is possible because $D'$ is discrete. Let $\mathcal O = \{N'(x)\cup (X\setminus D'): x\in D'\}$ and note that $\mathcal O$ is a countable open cover of $X$. Yet if we remove any open set $N'(x)\cup (X\setminus D')$ from $\mathcal O$, then $x\notin \bigcup \mathcal O$. Hence by countable compactness, $D'$ is finite, contradicting our choice of $D'$. Thus $D$ must be finite.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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