This question was originally posted to Math.StackExchange, but having got no response there, I'm reposting it here. I apologise if it is too elementary for this site. (Original post: https://math.stackexchange.com/questions/170310/calculating-hng-mathbbzg-as-co-homology-with-compact-support-of-a-prope )

I came across the following Exercise in Brown's book "Co-homology of Groups", and have been completely unable to solve it. If anyone could give me a hint that would point me in the right direction I would really appreciate it. I couldn't even think of a filtration of $X$ which might produce a useful spectral sequence.

Brown "Co-homology of Groups", Section VIII.7, p.209, Ex 4:

Prove the following generalisation of 7.5: Let $X$ be a contractible $G$-CW-complex with finite isotropy groups and only finitely many cells mod-$G$. Then $H^*(G, \mathbb{Z} G) \cong H_c^*(X, \mathbb{Z})$.

I presume that he means $H^*(G, \mathbb{Z} G) \cong H_c^*(X, \mathbb{Z})$ as $\mathbb{Z}G$-modules.

There is a hint:

First show, by a spectral sequence argument for instance, that $H^*(G, \mathbb{Z}G)$ can be computed from $\text{Hom}_G(C(X), \mathbb{Z}G)$.

Here $H_c^* $ is co-homology with compact support. Throughout Brown's book any action on a space is always rigid (set-wise and point-wise stabilisers of cells coincide). The "7.5" which he refers to is the statement that $H^*(G, \mathbb{Z}G) \cong H_c^*(X, \mathbb{Z})$ as $\mathbb{Z}G$ modules, where $X$ is a contractible, free, co-compact $G$-CW-complex.


1 Answer 1


Perhaps the "spectral sequence argument" Brown is referring to is the cohomology analogue of Proposition 7.3 in Chapter VII. It seems that for any $G$-CW-complex $X$ and $G$-module $M$, there should be a spectral sequence $$ H^p(G;H^q(X;M))\Rightarrow H^{p+q}_G(X;M). $$ It then should follow (using Proposition 5.3 in the same chapter) that if $f\colon X\to Y$ is a map of $G$-CW-complexes which induces an isomorphism on ordinary integral homology, then $f$ induces an isomorphism $H^\ast_G(X;M)\cong H^\ast_G(Y;M)$. In particular, if $X$ is the complex in the exercise and $Y$ is a point, we have $H^\ast_G(X;\mathbb{Z}G)\cong H^\ast(G;\mathbb{Z}G)$.

So far the finite isotropy assumption hasn't come in. But I believe this should be used in the second step, that is, showing that $\operatorname{Hom}_G(C(X);\mathbb{Z}G)$ has the same homology as $\operatorname{Hom}_c(C(X),\mathbb{Z})$ (generalising Lemma 7.4 of Chapter VIII).

I haven't checked the details yet, let me know if it works (or not)!

  • 2
    $\begingroup$ Thanks! I've written out your argument and it seems to work. $\endgroup$ Commented Jul 19, 2012 at 14:23

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.