Suppose $Y$ is an $(n-1)$-connected space, $n>2$, so we have Hurewicz isomorphisms $\pi_n(Y)\cong H_n(Y)$ and $\pi_{n-1}(\Omega Y)\cong H_{n-1}(\Omega Y)$. Let a map $\alpha\colon X\to\Omega Y$ be given. Naturally it induces a map $\beta\colon X\times S^1\to Y$. I want to show the following diagram is commutative: $$\require{AMScd} \begin{CD} H_{n-1}(X) @>\times[S^1]>> H_n(X\times S^1)\\ @V\alpha_*VV @V\beta_*VV \\ H_{n-1}(\Omega Y) @<\cong<< H_n(Y). \end{CD} $$

Since $\Omega Y$ is path connected, we may homotope $\alpha$ to a based map. Then $\beta$ factors though the reduced suspension $\Sigma X$. If $X=S^{n-1}$ is a sphere, the commutativity would then follow from tracking down the definition of $\pi_n(Y)\xrightarrow{\cong}\pi_{n-1}(\Omega Y)$. However I don't know how this helps for the general case.

One can also phrase the question in cohomology in the obvious way. (In particular the cross product $\times[S^1]$ will be replaced by the slant product $/[S^1]$.)


1 Answer 1


The general case does follow from the case $Y=S^n$.

Without loss of generality $X=\Omega Y$ and $\alpha$ is the identity. Now the statement comes down to the assertion that the composition $$ \pi_{n-1}\Omega Y\to H_{n-1}\Omega Y\to H_n(\Omega Y\times S^1)\to H_n Y $$ corresponds to the Hurewicz map $\pi_nY\to H_nY$ via a natural isomorphism between $\pi_{n-1}\Omega Y$ and $\pi_nY$. This is true for all $Y$, and by naturality it is enough to prove it for $Y=S^n$.


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.