Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

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

Are there any simple examples of subspaces of a normal space which are not normal?

I know closed subspace of a normal space is normal, but open subspace in most cases which I can think of are also normal.

share|cite|improve this question
There is a well known book Counterexamples in Topology that you can refer to for any standard counterexample. – Carl Mummert Oct 25 '12 at 2:57
Why are you considering only open and closed subspaces? – Chris Eagle Dec 21 '12 at 12:36
up vote 11 down vote accepted

One well-known example is $X=(\omega_1+1)\times(\omega+1)$, where $\omega_1$ is the first uncountable ordinal with the order topology, and $\omega$ is the first infinite ordinal. $X$ is the product of two compact Hausdorff spaces, so $X$ is compact and Hausdorff and therefore normal, but $X\setminus\{\langle\omega_1,\omega\rangle\}$ is not normal: the closed sets $\omega_1\times\{\omega\}$ and $\{\omega_1\}\times\omega$ cannot be separated by disjoint open sets. $X$ is often called the Tikhonov plank, though the name is also sometimes applied to $X\setminus\{\langle\omega_1,\omega\rangle\}$.

Added: We can also appeal to the theorem that a space $X$ is Tikhonov (i.e., $T_1$ and completely regular) iff it is homeomorphic to a subspace of some product of closed unit intervals. Every product of closed unit intervals is compact and Hausdorff, hence normal, but there are many examples of Tikhonov spaces that are not normal, and every non-normal Tikhonov space is an example of a non-normal subspace of a normal space. A couple of the better-known examples of non-normal Tikhonov spaces are the Moore plane and the Sorgenfrey plane. Another example is the space described in this answer, if $\mathscr{D}$ is taken to have cardinality $2^\omega=\mathfrak c$, which the addendum to the answer shows to be possible; non-normality of the space follows from Jones’s lemma.

share|cite|improve this answer
Sorry, am a newbie, and not familiar with "first uncountable ordinal", ω1, ω etc! – Jesse P Francis Oct 25 '12 at 1:20
@user73412: The space $\omega+1$ is homeomorphic to the subspace $\{0\}\cup\left\{\frac1n:n\in\Bbb Z^+\right\}$ of $\Bbb R$. Think of $\omega_1+1$ as a set $Y$ with a linear order $\le$ on it such that (a) every non-empty subset of $Y$ has a $\le$-least element; (b) $Y$ has a $\le$-greatest element called $\omega_1$; $Y$ is uncountable; and if $y\in Y\setminus\{\omega_1\}$, then $\{z\in Y:z\le y\}$ is countable. That won’t tell you how to work with it, but it may at least give you some idea of what it is. – Brian M. Scott Oct 25 '12 at 1:31

${\mathbb R}$ is homeomorphic to $(0,1)$ and Tikhonov theorem tells us that the product of countably infinitely many copies of an interval is compact Hausdorff so a subspace of a normal subspace is not always normal.

share|cite|improve this answer
I don't understand this example. What is the non-normal subspace? – JSchlather Oct 25 '12 at 2:00
Are you referring to my answer? – glebovg Oct 25 '12 at 2:20
Yes. I am referring to your answer. It seems to imply that $(0,1)$ is not normal. – JSchlather Oct 25 '12 at 2:35
But the product of uncountably many copies of ${\mathbb R}$ is not normal. – glebovg Oct 25 '12 at 15:07
${[0,1]^{\mathbb R}}$ is normal but ${(0,1)^{\mathbb R}}$ is not. – glebovg Oct 30 '12 at 14:33

Consider $ [0,1]^{\mathbb{R}} $ with the product topology. By Tychonoff theorem $ [0,1]^{\mathbb{R}} $ is compact. Since every compact space is normal, we have $[0,1]^{\mathbb{R}} $ is normal.

Note that $ (0,1)^{\mathbb{R}} $ is a subpace of $ [0,1]^{\mathbb{R}} $. Since the product of uncountably many copies of ${\mathbb{R}}$ is not normal and ${\mathbb{R}}$ is homeomorphic to (0, 1) we have $ (0,1)^{\mathbb{R}} $ is not normal. Therefore a subspace of a normal space is not necessarily normal.

share|cite|improve this answer

The topology $T = \{\emptyset, X , \{a\} , \{a ,b\}, \{a , c\}, \{a ,b ,d\}\}$ on $X = \{a, b , c , d \}$ is normal. Take $Y = \{a, b, c\}$. Then the topology on the set Y is $S = \{ \emptyset , Y , \{a\} , \{ a ,b \} , \{ a , c \}\}$ is not normal. Hence subspace of normal space need not be normal.

share|cite|improve this answer
Sorry, I think (X,T) itself is not normal. By definition, any normal space should be T1, and (X,T) is not T1. For example, pick a,d. There are no neighborhoods of a and d, disjoint from each other. – Jesse P Francis Dec 31 '12 at 10:30

The topology $T=\{\emptyset , X,\{a\},\{a,b\},\{a,c\}\}$ on $X=\{a,b,c,d\}$.

Take the subspace $Y=\{a,b,c\}$ with Topology $S=\{\emptyset,Y,\{a\},\{a,b\},\{a,c\}\}$. Then Y is not normal but $(X,T)$ is normal.

share|cite|improve this answer
$T$ is not a topology for $X$; $\{a,b\}$ is open, as is $\{a,c\}$, but $\{a,b\} \cup \{a,c\} = \{a,b,c\}$ is not. – JSQuareD Jan 6 '15 at 15:46

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.