There is at least 3 model structures on the category of topological spaces, the Quillen Model structure, the Storm model structure and the Mixed model structure. In the Mixed model structure $\mathsf{MixTop}$ ( mixed model structure), weak equivalence are weak homotopy equivalences, fibrations are Hurewicz fibrations and cofibrations are determined by the lifting property.

Suppose that we have a continuous map $f:X\rightarrow Y$ between cofibrant objects in $\mathsf{MixTop}$ and $f$ is a closed embedding i.e. $f(X)$ is closed subspace of $Y$ and $f:X\rightarrow f(X)$ is a homemorphism.

I was wondering if $f$ is a cofibration in $\mathsf{MixTop}$? Here is the nLab reference for the mixed cofibrations.


1 Answer 1


No. Let $X$ be an uncountable set and consider $I^X$ with the product topology. Then the inclusion $\{ 0 \} \to I^X$ is a closed embedding between (strongly) contractible spaces which are therefore mixed cofibrant. However, this map is not a Hurewicz cofibration since it would follow that $\{ 0 \}$ is the zero set of a continuous function $I^X \to I$ which would contradict the fact that $0$ has no countable neighbourhood basis in $I^X$. In particular, this map is not a mixed cofibration. (However, it is a Dold cofibration so pushouts along it are still homotopy pushouts with respect to both homotopy equivalences and weak homotopy equivalences.)

Added: Here is an example that is not even a Dold cofibration. Let $Y$ be the subspace of $\mathbb{R}^2$ that is the union of the line segment connecting $(0, 1)$ to $(0, 0)$ and the line segments connecting $(0, 1)$ to $(2^{-m}, 0)$ for all $m \in \mathbb{N}$. The inclusion $\{ 0 \} \to Y$ is a closed embedding, again between strongly contractible spaces. If it was a Dold cofibration, then it would admit a strong deformation retraction (this should be somewhere in the reference I mention in the comment). However, it is a standard exercise in topology that no such retraction exists.

  • $\begingroup$ Thank you very much, do you have a reference for Dold cofibrations? The map $f$ in the question is not a Dold cofibration ? $\endgroup$
    – Ilias A.
    Nov 9, 2015 at 12:54
  • $\begingroup$ The only reference I know is Section I.2 of Homotopietheorie by tom Dieck, Kamps and Puppe (in German, they call Dold cofibrations "Homotopiecofaserungen"). I suspect that closed inclusions between mixed cofibrant spaces are not in general Dold cofibrations, but I can't think of a counterexample at the moment. $\endgroup$ Nov 9, 2015 at 13:33

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.