0
$\begingroup$

Let $X$ be a topological space such that $\pi_n(X)=H_n(X)=Z$. A continuous map $f: S^n \rightarrow X$ is an element of $\pi_N(X)=Z$ therefore $[f]_{\mathrm{homotopy}}$ is characterised by an integer $k$.

I am trying to show that $k$ is the degree of $f$. Is the following correct?

By Hurewicz theorem we have a group homomorphism $h:\pi_n(X)\rightarrow H_n(X)$. Let $[c]$ be a generator of $H_n(X)$, i.e. $[c]_{\mathrm{homology}}=1$. The Hurewicz map is \begin{equation} h([f]_{\mathrm{homotopy}})=f_*[c]_{\mathrm{homology}} .\end{equation} For $[f]\in\pi_n(X)$ on one hand we have, by definition of degree \begin{equation} h[f]=f_*[c]=\mathrm{deg}(f)\,[c]. \end{equation} On the other hand if $[f]=k$, since the Hurewicz map is a group homomorphism we also have \begin{equation} h[f]=h[k]=k\, h[1]=k\, [c]. \end{equation} It follows that the integer characterising the homotopy class of $f$ is the degree of $f$.

$\endgroup$
  • 1
    $\begingroup$ how do you know that $H_n(X)=\mathbb{Z}$? $\endgroup$ – Lee Mosher Aug 4 '14 at 18:49
  • $\begingroup$ @LeeMosher You are right, I started with $X=S^n$ and then incorrectly generalised, let me edit the question and put it as an hypothesis. $\endgroup$ – GFR Aug 4 '14 at 18:52
3
$\begingroup$

Here is a counterexample. Consider $X = \mathbb{R}P^3$, an orientable 3-manifold and so $H_3(\mathbb{R}P^3)=\mathbb{Z}$. The universal covering map $S^3 \to \mathbb{R}P^3$ induces an isomorphism on $\pi_3$ (by the long exact sequence of homotopy groups for a covering map). Since the identity map on $S^3$ represents a generator of $\pi_3(S_3)=\mathbb{Z}$ it follows that the universal covering map $S^3 \to \mathbb{R}P^3$ represents the generator of $\pi_3(\mathbb{R}P^3) = \mathbb{Z}$. But the universal covering map has degree 2.

In your proof, the expression $f_*[c]$ makes no sense, because the domain of $f_*$ is $\pi_n(S^n)$ or $H_n(S^n)$ (either one is OK because they are naturally isomorphic by the Hurewicz theorem) but $[c]$ is an element of $H_n(X)$.

$\endgroup$
  • $\begingroup$ Thinking about what you wrote it seems to me that my "proof" is only valid when there is a map $g: S ^n\rightarrow X^n$ which induces an isomorphism of the $n$th homotopy group AND the Hurewicz map from $\pi_n(X)$ to $H_n(X)$ is an isomorphism because then we have the commutative square \begin{equation}\begin{CD} \pi_n(S^n) @>g_*>> \pi_n(X)\\@VVhV @VVhV\\H_n(S^n) @>g_*>>H_n(X) \end{CD}\end{equation}. Is this correct? I am not strong in algebraic topology, so sorry if I write nonsense and thanks for your help! $\endgroup$ – GFR Aug 5 '14 at 8:50
  • $\begingroup$ mmm the commutative diagram does not want to compile. The top horizontal line is $\pi_n(S^n) \rightarrow \pi_n(X)$ via the map induced by $g$ on homotopy, the bottom horizontal line is $H_n(S^n)\rightarrow H_n(X)$ via the map induced by $g$ on homology. The vertical maps are the Hurewicz maps. $\endgroup$ – GFR Aug 5 '14 at 8:57
  • 1
    $\begingroup$ Under your assumption about the map $g$, then yes, this is true, and it follows from the theorem that the Hurewicz homomorphism is a natural transformation. $\endgroup$ – Lee Mosher Aug 5 '14 at 13:03

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.