2
$\begingroup$

Let $G$ be a simply connected (if necessary, compact Lie) group with finite center $Z$ and $p:G/\to G/Z$ be the canonical projection. Is there any way to know if every element in $\operatorname{Out}(G/Z)$ lifts to an element in $\operatorname{Out}(G)$?

I'm trying to understand who is $\operatorname{Out}(SU(3)/\mathbb{Z}_3)$.

Take $\varphi\in\operatorname{Out}(G/Z)$. From the fact that $G$ is simply connected we have that $p\circ \varphi$ lifts to a homomorphism $\tilde{\varphi}:G\to G$, but with kernel $Z$. Is there any way to turn this $\tilde{\varphi}$ into an automorphism?

$\endgroup$
6
  • $\begingroup$ This may be helpful: mathoverflow.net/questions/40666/… $\endgroup$
    – KhashF
    Commented Feb 24, 2021 at 14:38
  • 1
    $\begingroup$ Not true for the Heisenberg group. Is $Z$ finite? In that case $\tilde\phi$ is an automorphism. $\endgroup$ Commented Feb 24, 2021 at 14:39
  • $\begingroup$ Yes! Z is finite! I'm concerned about $Out(SU(3)/Z^3)$ $\endgroup$
    – Gomes93
    Commented Feb 24, 2021 at 14:47
  • $\begingroup$ @FriedrichKnop why is $\tilde{\varphi}$ an automorphism in this case? $\endgroup$
    – Gomes93
    Commented Feb 24, 2021 at 14:52
  • 7
    $\begingroup$ Yes, if $G$ is a simply connected Lie group with discrete center $Z$, then both $\mathrm{Aut}(G)$ and $\mathrm{Aut}(G/Z)$ can be identified to $\mathrm{Aut}(\mathfrak{g})$, and in both cases one gets Out by modding out by the canonical copy of $G/Z$ in $\mathrm{Aut}(\mathfrak{g})$. $\endgroup$
    – YCor
    Commented Feb 24, 2021 at 14:59

1 Answer 1

2
$\begingroup$

Maybe I will just answer the specific question the OP is interested in by giving a reference of sorts (in case it is helpful to the OP).

In Bourbaki Éléments de mathématique: groupes et algèbres de Lie. Chapitre 9 Exercise 3 one is asked to prove that $\mathrm{Out}(PSU(n))\cong \mathbb{Z}/2\mathbb{Z}$ for $n\geq 3$.

It seems to me that in this case one can show this explicitly. The non-trivial non-inner automorphism is represented by the Cartan involution $A\mapsto (A^{-1})^T$ on $PSL(n,\mathbb{C})$ which simplifies to $A\mapsto \overline{A}$ on $PSU(n)$.

$\endgroup$

You must log in to answer this question.

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