MathOverflow is a question and answer site for professional mathematicians. 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

A colleague in algebra asked me this, and I couldn't answer it. On the Wikipedia page for "epimorphism" it is claimed that in the category of Hausdorff spaces and continuous maps, a function is epi if and only if it has dense range. The "if" case is easy, but I couldn't justify the "only if" case.

This boils down to: let Y be a Hausdorff space, and let X in Y be a closed subset not equal to Y, and not empty. Can you find a Hausdorff space Z and functions f,g:Y->Z such that f and g agree on X, but are not equal. I think, by using a quotient argument, you can assume that X is just a point.

share|cite|improve this question
I am not convinced you can reduce the general case to one where X is a point. You would need better separation than Hausdorff, or the quotient may not be Hausdorff. – Harald Hanche-Olsen Feb 1 '10 at 15:10
Sorry, yes, that's right: if you collapse X to a point x then it's not clear that you can separate x from other points. – Matthew Daws Feb 1 '10 at 15:15
If you can reduce to the case of a point, then you are done. Take Y=Z, f to be the identity and g to be the map sending all of Y to the point X. So all the challenge is in that reduction step. – David Speyer Feb 1 '10 at 16:27
@frb: Your second map is likely to be discontinuous. – Harald Hanche-Olsen Feb 1 '10 at 17:04
Answer withdrawn: sorry, missed the obvious fact that the map isn't open! – Charles Siegel Feb 1 '10 at 17:07
up vote 13 down vote accepted

Let $Y$ be a Hausdorff space, and let $X\subset Y$ be a closed subspace. Consider disjoint union of two copies of $Y$, and let $Z$ be the coequalizer of two embeddings of $X$ into it (that is, we glue two copies of $Y$ along $X$). Clearly, the two natural maps $\iota_{1,2}:Y\to Z$ coincide only on $X$. It is easy to see that $Z$ is Hausdorff.

Indeed, take $z_1,z_2\in Z$, $z_1\ne z_2$. The map $p:Z\to Y$ is continuous, so if $p(z_1)\ne p(z_2)$, we can take preimages of open neighborhoods of $p(z_1)$ and $p(z_2)$ to separate $z_1$ and $z_2$. It remains to deal with the case $z_{1,2}=\iota_{1,2}(y)$ for $y\in Y-X$. But the neighborhoods $\iota_{1,2}(Y-X)$ work.

share|cite|improve this answer

This is really a comment on t3suji's answer, but it's too long to be a comment as such.

t3suji's answer is the canonical one in the following precise sense. Let $e: X \to Y$ be a morphism in any category. It's an elementary exercise to show that the following conditions on $e$ are equivalent:

  1. $e$ is an epimorphism

  2. the square $$ \begin{array}{ccc} X &\stackrel{e}{\to} &Y \\ e\downarrow & &\downarrow 1_Y \\ Y &\stackrel{1_Y}{\to} &Y \end{array} $$ is a pushout

  3. for some morphism $f: Y \to Z$, the square $$ \begin{array}{ccc} X &\stackrel{e}{\to} &Y \\ e\downarrow & &\downarrow f \\ Y &\stackrel{f}{\to} &Z \end{array} $$ is a pushout.

I'll only use the equivalence 1 $\iff$ 3 here. The other implications are just scene-setting.

Suppose we want to show that a particular morphism $e$ is not epi. Assuming that there are enough pushouts around, we can argue as follows. Form the pushout square $$ \begin{array}{ccc} X &\stackrel{e}{\to} &Y \\ e\downarrow & &\downarrow f \\ Y &\stackrel{g}{\to} &Z. \end{array} $$ If $f \neq g$ then the implication 1 $\Rightarrow$ 3 tells us that $e$ is not epi. Moreover, this strategy is bound to work, in the sense that if $f = g$ then the implication 3 $\Rightarrow$ 1 tells us that $e$ is epi after all.

It only remains to see that this is indeed what t3suji did. In his/her situation, $e$ was the inclusion $X \to Y$. He/she then took the coequalizer of the two obvious maps $X \to Y + Y$ (where $+$ means coproduct, i.e. disjoint union). For elementary and totally general reasons, this is the same thing as taking the pushout just mentioned. The morphisms that t3suji called $\iota_1$ and $\iota_2$, I called $f$ and $g$. Finally, although t3suji's pushout is in the category of all topological spaces, he/she then verified that the space $Z$ is indeed Hausdorff, from which it follows that it's also a pushout in Hausdorff spaces.

So now you know, in principle, how to answer any question of the form "prove that such-and-such a morphism isn't epi".

share|cite|improve this answer
Of course, in general the Principle of Conservation of the Difficulty under Changes in the Statement will probably kick in in most cases! – Mariano Suárez-Alvarez Feb 1 '10 at 20:37
Mariano, you mean that rephrasing a question in a different language never makes it fundamentally easier? I don't think I agree, and that wasn't the point of my answer. I was just observing that t3suji's answer (man, typing that name is giving me trouble) embodies some kind of universal method for solving these problems. – Tom Leinster Feb 1 '10 at 21:31
Yep: I think this is a very nice observation by Tom. Actually, I sat in on an algebra grad course last year which dealt with some of this homological algebra. This popped into my mind immediately upon seeing t3suji's answer. So I think I disagree with Mariano: if someone had said "pushout" to me, and walked away, I think I would have come up with t3suji's idea pretty quickly... – Matthew Daws Feb 1 '10 at 21:47
@Tom Leinster: (man, typing that name is giving me trouble) Sorry for that --- I somehow got so used to it that I have entered it in Math Overflow without much thought. It feels sort of silly to change it now... @Matthew Daws: It's interesting that somehow pushouts are not the first thought after reading the original problem (at least not for me). So maybe Tom's beautiful abstract set-up should help me see it next time I run across it. – t3suji Feb 1 '10 at 22:00

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.