1
$\begingroup$

Let $ G $ be a compact topological group which is quasisimple in the sense that $$ [G,G]=G $$ and $$ G/Z(G) $$ is simple as an abstract group. Must $ G $ be a Lie group?

This is a follow-up question to https://math.stackexchange.com/questions/4537401/compact-simple-group-which-is-not-a-lie-group

By Peter-Weyl theorem there must be a nontrivial continuous homomorphism from $ G/Z(G) $ to some $ U_n $. Since $ G/Z(G) $ is abstractly simple then this is an embedding and so we realize $ G/Z(G) $ as a compact and thus closed subspace of some $ U_n $. Thus $ G/Z(G) $ is a closed subgroup of $ U_n $ so it is a Lie group. So the question is equivalent to: let the compact topological group $ G $ be a perfect central extension of a compact simple Lie group. Must $ G $ be a Lie group?

$\endgroup$
4
  • 1
    $\begingroup$ Yes. This follows from the fact that $G/Z(G)$ has a finite fundamental group, so the size of the images of $Z(G)$ in unitary groups is bounded. $\endgroup$
    – YCor
    Commented Sep 29, 2022 at 14:01
  • $\begingroup$ @Ycor Could you expand on that? I understand that $ G/Z(G) $ has finite fundamental group since it is a simple Lie group. But I don't follow the argument about the bound on the size of the image of $ Z(G) $ for a nontrivial continuous homomorphism $ G \to U_n $ and how that all comes together to show that $ G $ is a Lie group. $\endgroup$ Commented Sep 29, 2022 at 14:13
  • $\begingroup$ If $G$ is infinite, it has a quotient $S=G/N_S$ which is a compact quasisimple Lie group, say with fundamental group of order $n_S$. Choose $S$ with $n_S$ minimal. If by contradiction $N_S\neq 1$, then using Peter-Weyl, we obtain a quotient that is a nontrivial finite connected cover of $S$, contradicting the minimality of $n_S$. $\endgroup$
    – YCor
    Commented Sep 29, 2022 at 14:24
  • $\begingroup$ @YCor would you be willing to write that up as an answer? $\endgroup$ Commented Sep 29, 2022 at 15:09

1 Answer 1

3
$\begingroup$

The group $G/Z(G)$ being simple, is a Lie group by Peter-Weyl. Hence it is either finite or connected. If it is finite, $G$ has center of finite index, hence has a finite derived subgroup. Since $G$ is perfect, this means that $G$ is finite.

Now suppose that $G/Z(G)$ is connected. Let $H$ be any Lie quotient of $G$ that is a cover of $G/Z(G)$. Then the image of $Z(G)$ in $H$ is central, and since the quotient has trivial center, we deduce that this image is precisely $Z(H)$. In particular, we see that $H=Z(H)H^0$. Since $G$ is perfect, so is $H$, and we deduce that $H=H^0$. Thus $H$ is a connected cover of $G/Z(G)$. Since $G/Z(G)$ has a finite fundamental group, say of order $d$, we deduce that $|Z(H)|\le d$.

This concludes: indeed if by contradiction $|Z(G)|>d$, then there exists a Lie quotient $H$ of $G$ in which the image of $Z(G)$ has cardinal $>d$ (by Peter-Weyl) and the above yields a contradiction.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .