-1
$\begingroup$

A topological space $X$ is called Artinian if the descending chain condition holds for open subsets of $X$. If the descending chain condition holds for open basis subsets of a Hausdorff space $X$ with the property that the intersection of two open basis is an open basis, can we prove that $X$ is Artinian?

Also, is there any proof for the fact that every Artinian Hausdorff space is finite?

$\endgroup$
12
  • 1
    $\begingroup$ Answer: NO. Any infinite discrete space provides a counterexample. (It's also a counterexample for Noetherian case). $\endgroup$
    – Wlod AA
    Apr 10, 2018 at 17:37
  • $\begingroup$ It'd be nice if you mentioned your editing (some Qs on MO fail mentioning their editing, thus making some answers look silly for no good reason). $\endgroup$
    – Wlod AA
    Apr 10, 2018 at 19:51
  • $\begingroup$ Your editing didn't help. The infinite discrete space is still a counterexample. $\endgroup$
    – Wlod AA
    Apr 10, 2018 at 19:54
  • 1
    $\begingroup$ @YCor As far as I can tell from the edit history, the question whether every artinian Hausdorff space is finite was not the original question but was introduced in version 3. The original question was whether the descending chain condition (DCC) for a basis implies DCC for the whole topology, and this is what WlodAA answered negatively, the counterexample being an infinite discrete space, which is (as you noted) not artinian but has a basis, the family of singletons (and $\varnothing$ to make it closed under intersection), with DCC. $\endgroup$ Apr 10, 2018 at 23:14
  • 1
    $\begingroup$ @WlodAA yes I know, but given that the other interpretation is also trivial... no need to lose more time discussing this. $\endgroup$
    – YCor
    Apr 11, 2018 at 9:47

1 Answer 1

1
$\begingroup$

If $X$ is Artinian, then for every $x \in X$ there is a minimal neighbourhood $U_x$ such that if $O$ is open with $x \in O$, then $U_x \subseteq O$.

This is a simple application of Zorn's lemma on the poset of open neighbourhoods of $x$, ordered by reverse inclusion. The Artianness of $X$ implies that all chains in this poset are finite and then the intersection is an upperbound. So a maximal element exists. If $O$ is then open and contains $x$, we have $O \cap U_x \ge U_x$ so that by minimality $O \cap U_x = U_x$ or $U_x \subseteq O$.

If $X$ is moreover Hausdorff or even $T_1$, for $x \in X$ we have that $U_x =\{x\}$ or otherwise $y \neq x$ with $y \in U_x$ would exist. But $x$ must have a neighbourhood missing $y$ and this contradicts $y \in U_x$. So $X$ is discrete. And a discrete space is Artinian iff it is finite.

$\endgroup$
6
  • $\begingroup$ Unless I overlooked something, the same proof works with $T_1$ in place of "Hausdorff". $\endgroup$ Apr 10, 2018 at 23:17
  • $\begingroup$ The QUESTION asks: can we prove that X is Artinian? But your answers ASSUMES that $X$ is Artinian. I don't see any logical relation between your answer and the Question. $\endgroup$
    – Wlod AA
    Apr 11, 2018 at 0:54
  • 2
    $\begingroup$ @WlodAA the last line asked “is there any proof that an Artinian Hausdorff space is finite?” $\endgroup$ Apr 11, 2018 at 3:52
  • $\begingroup$ @AndreasBlass I agree. The question asked for Hausdorff. But indeed such a space can at best be $T_0$ I think. $\endgroup$ Apr 11, 2018 at 3:54
  • $\begingroup$ @HennoBrandsma by "such a space" I guess you mean "a infinite artinian topological space". Sure there are $T_0$ examples and no $T_1$ example. $\endgroup$
    – YCor
    Apr 11, 2018 at 9:06

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