5
$\begingroup$

Let $p\colon E\to C$ be a continuous, surjective map between topological spaces with $C$ contractible. Suppose that $p^{-1}(c)$ is contractible for each $c\in C$. Is it true that $E$ is weakly contractible?

If not, what are some mild conditions that will guarantee that this is true? I am aware that this is true if $p$ is a (quasi-)fibration. Sorry if this is an easy question.

$\endgroup$
2
  • 10
    $\begingroup$ For a less pathological counterexample, consider the obvious bijection $[0,1]\cup (2,3]\to [0,2]$. $\endgroup$ Jun 20, 2014 at 12:40
  • 1
    $\begingroup$ Look up the Vietoris–Begle mapping theorem. $\endgroup$
    – Jeff Strom
    Jun 20, 2014 at 13:30

2 Answers 2

6
$\begingroup$

Here is the main theorem of "A Vietoris Mapping Theorem for Homotopy" by S. Smale:

THEOREM: Let $f:X\to Y$ be proper and onto, where $Y$ and $X$ are

  • $0$-connected
  • separable
  • locally compact
  • metric.

Assume $X$ is $LC^n$ (which I presume means locally $n$-connected) and that for each $y\in Y$, $f^{-1}(Y)$ is $LC^{n-1}$ and $(n-1)$-connected. Then

  1. $Y$ is $LC^n$ and
  2. $f$ is an $n$-equivalence.

Some comments:

  1. $f$ is proper means the preimages of compact sets are compact; this is automatic if $X$ is compact and $Y$ is Hausdorff.
  2. the local conditions on $X$ and $Y$ are satisfied if $X$ and $Y$ are CW complexes.

In your situation, if your $f$ satisfies the conditions of this theorem for all $n$, then we deduce that $f$ is an $n$-equivalence for all $n$ and hence that $E$ is weakly contractible because $C$ is contractible.

$\endgroup$
0
$\begingroup$

Check out:

MR0036509 (12,120e) Reviewed Borel, Armand; Serre, Jean-Pierre Impossibilité de fibrer un espace euclidien par des fibres compactes. (French) C. R. Acad. Sci. Paris 230, (1950). 2258–2260. 56.0X

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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