Let $(X,A)$ be a relative CW-complex. Consider the inclusion $$i:\ \ X\times \{0,1\} \cup A\times I \rightarrow X\times I.$$

I was wondering if this is a cofibration.

I guess it is, for there is a push out diagram $\require{AMScd}$ \begin{CD} A\times (0,1) @>>> X\times \{0,1\} \cup A\times I\\ @VVV @VVV\\ X\times(0,1) @>>> X\times I \end{CD} the left vertical arrow is obviously a cofibration. So is the right vertical arrow.

Is the above argument correct?

  • $\begingroup$ Why is this diagram a pushout diagram? $\endgroup$ – Paul Frost Feb 13 at 18:40
  • $\begingroup$ @PaulFrost Since $X\times I$ is the disjoint union of $X\times (0,1)$ and $X\times \{0,1\}\cup A\times I$ after identifying $A\times (0,1)$. Since all the maps are inclusions, the diagram commutes. $\endgroup$ – Aolong Li Feb 13 at 18:46

Your argument is not correct because the diagram in your question is not a pushout diagram. If it were one, then for any two maps $\phi : X \times \{ 0, 1 \} \cup A \times I \to Y$ and $\psi : X \times (0,1) \to Y$ which agree on $A \times (0,1)$ we would get a unique map $\chi : X \times I \to Y$ whose restriction to $X \times \{ 0, 1 \} \cup A \times I$ is $\phi$ and whose restriction to $X \times (0,1)$ is $\psi$.

Counterexample 1.

Any CW-complex $X$ can be regarded as relative CW-complex $(X,A) = (X,\emptyset)$. Take $Y = I$ and define $\phi : X \times \{ 0, 1\} \to I, \phi(x,t) = 0$ and $\psi : X \times (0,1) \to I, \psi(x,t) = t$. These maps agree on $\emptyset \times (0,1)$ but do not give you the desired $\chi : X \times I \to I$.

Counterexample 2.

Let $(X,A) = (I,\{ 0\})$. Define $\phi : I \times \{ 0, 1 \} \cup \{ 0 \} \times I \to I, \phi(x,y) = y$ and $\psi : I \times (0,1) \to I, \psi(x,y) = \min (x+y,1)$. These maps agree on $\{ 0 \} \times I$ but again you do not the desired $\chi : I \times I \to I$.

A correct proof can be based on the observation that $(X \times I, X \times \{ 0, 1 \} \cup A \times I)$ is also a relative CW-complex.

By the way, it is a general theorem that if an inclusion $A \hookrightarrow X$ is a cofibration, then so is $X \times \{ 0, 1 \} \cup A \times I \hookrightarrow X \times I$. The proof requires some deeper properties of cofibrations. See Theorem 6 of

Strøm, Arne. "Note on cofibrations II." Mathematica Scandinavica 22.1 (1969): 130-142.


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.