3
$\begingroup$

I'm trying to show directly that $$ H_0(X, x_0) \cong \widetilde{H}_0(X) $$ where $x_0$ is a point in the topological space $X$ and $\widetilde{H}_0$ denotes the zeroth reduced (singular) homology group.

What I have so far gives me a wrong answer, I'd be grateful if anyone can point out the mistake and the right direction to go.

Attempted proof: As usual we have the chain complex $$ \cdots \longrightarrow \frac{C_1(X)}{C_1(x_0)} \stackrel{d_1}{\longrightarrow} \frac{C_0(X)}{C_0(x_0)} \longrightarrow 0$$ and by definition $$ H_0(X,x_0) = \frac{C_0(x)/C_0(x_0)}{\text{im}(d_1)}. $$

Now $$d_1\left(\frac{C_1(X)}{C_1(x_0)}\right) = \frac{d_1(C_1(X))}{C_0(x_0)}, \tag{$\ast$}$$ so $$ H_0(X,x_0) \cong \frac{C_0(X)}{d_1(C_1(X))} $$ by the third isomorphism theorem.

$C_0(X)$ is the free abelian group generated by the points of $X$, and $d_1$ sends the generators $$ \sigma : [0,1] \rightarrow X$$ of $C_1(X)$ to $(\sigma(1) - \sigma(0))$, so $$ d_1(C_1(X)) = \mathbb{Z}\{y-x\ |\ x,y \ \text{in the same path component of}\ X\},$$ and after modding out $C_1(X)$ by this we get that $H_0(X,x_0)$ is a direct sum of $\mathbb{Z}$'s, one for each path component of $X$. ($\ast\!\ast$)

But we know that $\widetilde{H}_0(X)$ is a direct sum of $n$ copies of $\mathbb{Z}$, where $n$ is one less than the number of path components of $X$. (For $X$ having a finite number of path components.)

There are two places in this argument I'm a bit suspicious of, namely the bits marked ($\ast$) and ($\ast\!\ast$), but I can't quite say what the problem might be. Help please?


Note that the same question is answered using chain homotopies here, but I'm looking for a more direct proof.

$\endgroup$
  • $\begingroup$ Did you figure it out? $\endgroup$ – Bruno Joyal Nov 17 '13 at 23:34
  • $\begingroup$ Cheers @BrunoJoyal, I did in the end :) Thanks. $\endgroup$ – Josh Chen Nov 18 '13 at 3:55
2
$\begingroup$

There is a problem with $(*)$: remark that $C_0(x_0)$ is not even a subgroup of $d_1(C_1(X))$, so it makes no sense to write the quotient. It's not a subgroup because $C_0(x_0)$ contains elements with nonzero "weight" (i.e. things which aren't in the kernel of the augmentation map). However, the image of $C_1$ lies in the kernel of the augmentation map. I'll let you think about what the image really is.

As for $(**)$, it's a little vague, so I hope that once you take care of $(*)$ you'll be able to clear it up as well.

Here's a general tip: instead of trying to prove that two gadgets are isomorphic by computing them side by side, focus on constructing a map from one gadget to the other, and then prove that your map is an isomorphism.

$\endgroup$

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.