A group $G$ is perfect if $G=[G,G]$. For perfect groups, we know that the first group homology $H_1(G, \mathbb{Z})=G/[G,G]=0$. A group $G$ is called acyclic if its group homology $H_i(G, \mathbb{Z})=0$ for $i \geq 1$ and $H_0(G, \mathbb{Z})=\mathbb{Z}$. It is well-known that the group $A_5$ is a perfect group which is not acyclic. However, I am interested in computing group homology with rational coefficients. I would like to know an example of a perfect group which has non-trivial rational group homology. It seems to me that such a group cannot be a finite perfect group, because one can use the universal coefficient theorem $$0 \rightarrow H_i(G,\mathbb{Z})\otimes \mathbb{Q}\rightarrow H_i(G, \mathbb{Q})→Tor(H_{i-1}(G,\mathbb{Z}),\mathbb{Q})→0$$ and deduce the rational group homology vanish for $i \geq 1$ inductively. So I am wondering if a perfect group is actually rationally acyclic or a counterexample to this exisit?


You can find an example where second homology group of a perfect group is not zero in this paper of Milnor. p. 79.


Milnor consider the cohomology of the discrete group $PSL(2,R)$ (in his terminology $PSL(2,R)$ made discrete) and it shows that there is a class in $H^2$ which is not trivial. The duality implies that you have an homology class $\geq 1$ which is not trivial.

  • $\begingroup$ Thank you for your fast reply. Sorry that I have one more question. In this example, it seems that $dim H_*(G, \mathbb{Q})=1$, since the square of the generator is mapped to zero in $H^4$. I am still wondering if there exists an example of perfect group such that $dim H_*(G, \mathbb{Q})>1$. In particular, can $H_*(G, \mathbb{Q})>1$ has arbitrary dimension? $\endgroup$ – awivil Feb 18 '16 at 4:53

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.