MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

In the elementary group theory we know that for the symmetric groups $S_n$, except $S_6$, we have $Aut(S_n) \cong S_n$. Then the following question is natural:

What is the necessary and sufficient condition for $G$ such that $Aut(G) \cong G$?

share|cite|improve this question
See also the question "Does Aut(Aut(...Aut(G) stabilize?", – Tom Church Aug 20 '10 at 23:09
up vote 13 down vote accepted

This answer is essentially a series of remarks, but ones which I hope will be helpful to you.

(1) There are two ways to interpret the condition that $G$ be isomorphic to its automorphism group: canonically and non-canonically.

a) Say that $G$ is complete if every automorphism of $G$ is inner (i.e., conjugation by some element of $G$) and $G$ has trivial center. In this case, there is a canonical isomorphism $G \stackrel{\sim}{\rightarrow} \operatorname{Aut}(G)$.

The linked wikipedia article gives some interesting information about complete groups. As above, by definition having trivial center is a necessary condition; all nonabelian simple groups satisfy this. On the other hand, an interesting sufficient condition is that for any nonabelian simple group $G$, its automorphism group $\operatorname{Aut}(G)$ is complete, i.e., we have canonically $\operatorname{Aut}(G) = \operatorname{Aut}(\operatorname{Aut}(G))$.

b) It is possible for a group to have nontrivial center and outer automorphisms and for these two defects to "cancel each other out" and make $G$ noncanonically isomorphic to $\operatorname{Aut}(G)$. This happens for instance with the dihedral group of order $8$.

2) It seems extremely unlikely to me that there is a reasonable necessary and sufficient condition for a general finite group to be isomorphic to its automorphism group in either of the two sense above.

But a lot of specific examples are certainly known: see for instance

in which the order of the outer automorphism group of each of the finite simple groups is given. So, for instance, exactly $14$ of the $26$ sporadic simple groups have trivial outer automorphism group, hence satisfy $G \cong \operatorname{Aut}(G)$.

I wouldn't be surprised if the outer automorphism groups of all finite groups of Lie type were known (they are not all known to me, but I'm no expert in these matters).

share|cite|improve this answer
The outer automorphism groups of all finite groups of Lie type are certainly known: see Gorenstein, Lyons & Solomon, volume 3. Note that if $S$ is a finite simple group and $Aut S$ is its full automorphism group, then $Aut S$ is complete (and it is the only almost simple group with socle $S$ which satisfies this). So, in particular $Aut (A_n)$ is complete in all cases, even $n=6$! – Nick Gill Oct 2 '12 at 13:29
By the way I disagree with your pessimism about characterising groups such that $G\cong Aut G$ but, at least right now, I can't back that up with a proof... – Nick Gill Oct 2 '12 at 13:29

There does not exist a reasonable necessary and sufficient condition for an infinite centerless group to be complete. More precisely, letting $V$ be the set-theoretic universe, there exists an infinite complete group $G \in V$ and a $c.c.c$ notion of forcing $\mathbb{P}$ such that $G$ has an outer automorphism in the generic extension $V^{\mathbb{P}}$. An example can be found in :

S. Thomas, The automorphism tower problem II, Israel J. Math. 103 (1998), 93-109.

In fact, it can be shown that the group $G$ in this paper satisfies the stronger property that $G \not \cong Aut(G)$ as abstract groups in $V^{\mathbb{P}}$. In other words, there does not even exist a ''non-canonical isomorphism''.

For more on the ``nonabsoluteness'' of the height of automorphism towers, see:

J. Hamkins and S. Thomas, Changing the heights of automorphism towers, Annals Pure Appl. Logic 102 (2000), 139-157.

share|cite|improve this answer
Simon, isn't the infinite symmetric group $Sym_\omega$ of V already an instance of this? After all, if you add a Cohen real $c$ by forcing (or any new real), then the $Sym_\omega$ of $V$ gains new automorphisms in $V[c]$, simply because there are new permutations of $\omega$ in $V[c]$. Do we know how the automorphism tower of $Sym_\omega^V$ is affected in $V[c]$? – Joel David Hamkins Aug 20 '10 at 13:39
But none of the new permutations normalize the old symmetric group which remains complete in any forcing extension. This is Theorem 2.2 in: S. Thomas, The automorphism tower problem II, Israel J. Math. 103 (1998), 93-109. – Simon Thomas Aug 20 '10 at 15:51
The group $G$ in my paper is the stabilizer $S_{(\mathcal{U})}$ of a nonprincipal ultrafilter $\mathcal{U}$ and the notion of forcing $\mathbb{P}$ is the corresponding Mathias forcing. – Simon Thomas Aug 20 '10 at 15:55
Yes, of course that's right! – Joel David Hamkins Aug 21 '10 at 0:18

The fact you are using above is really that all $S_n$, except for $n=2,6$, are complete groups. This means they are centerless, and all automorphisms are inner. Clearly, all complete groups $G$ are isomorphic to $Aut(G)$.

But the infinite dihedral group $D_\infty$ is centerless, yet not complete. We still have $D_\infty\cong Aut(D_\infty)$.

The dihedral group of order 8 $D_8$ is not centerless, and yet still $D_8\cong Aut(D_8)$.

I doubt very much there is a complete classification available.


share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.