I am working on problems in the past Qual exams.

"Let $f:\tilde X\to X$ be a covering map between path-connected and locally path-connected spaces. Suppose $p$ is homotopic to a constant map. Prove that $\tilde X$ is contractible."

I remember doing this before, but I forgot the final touch: we have a homotopy $H_t : \tilde X \to X$ s.t. $H_0(\tilde x)=f(\tilde x)$ and $H_1(\tilde x)=x_0$ constant. We observe that $\operatorname{id}_{\tilde X}$ is a lift of $H_0=f$, hence there is a unique lift $\tilde H_t : \tilde X \to \tilde X$ of the homotopy $H_t$ s.t. $\tilde H_0= \operatorname{id}_{\tilde X}$ and $f\circ \tilde H_1=H_1=x_0$.\

I'm stuck here. It suffices to prove $\tilde H_1$ is constant. But I don't have much information about $\tilde H_1$ except that $\tilde H_t$ is a unique lift of the whole $H_t$. How do I conclude?

I don't know how to use the hypothesis "path-connected and locally path-connected" here since in Hatcher's book, these hypotheses usually needed in the Galois corespondence, which I don't think is relevant here. Even the lift above is due to Proposition 1.30 which does not need these requirements.


1 Answer 1


To be honest, I am not entirely sure why $\tilde{X}$ path-connected and $X$ locally path connected are assumed here. It seems like $\tilde{X}$ being connected is sufficient.

As you said, it is sufficient to show that $\tilde{H}_1$ is constant. Now note that since $f \circ \tilde{H}_1 = x_0$ is constant, we have that the image of $\tilde{H}_1$ is contained in the fiber of $x_0$ under the map $f$. But since $f$ is a covering map, this fiber is discrete and since $\tilde{X}$ is connected, the map needs to map into one connected component i.e. $\tilde{H}_1$ is constant.

  • $\begingroup$ $\tilde X$ connected is in fact sufficient. It is also necessary because contractible spaces are connected. $\endgroup$
    – Paul Frost
    May 2, 2020 at 9:18
  • $\begingroup$ so if $\tilde X$ is connected, the problem is solved. It does not mean this problem is wrong, right? $\endgroup$ May 2, 2020 at 14:59
  • 1
    $\begingroup$ Yes, the problem is correct. I'm just saying that it is not necessary to assume that $\tilde{X}$ is path connected, but connected suffices. $\endgroup$ May 2, 2020 at 15:02
  • $\begingroup$ oh right, i thought connected implies path-con, i clearly misremember this $\endgroup$ May 2, 2020 at 15:25

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.