I read somewhere that if a group $G$ has only finitely many commutators then the commutator subgroup $D(G)$ is itself finite.

Do you know a proof of this result?

  • $\begingroup$ Let me figure this out: You want to show that $D(G)$ is finite if $G$ has finitely many commutators, right? But is $D(G)$ not defined as the subgroup of commutators? Am I misreading something? $\endgroup$
    – awllower
    Jan 24, 2013 at 14:22
  • 3
    $\begingroup$ @awllower: $D(G)$ is the subgroup generated by the commutators. The commutators themselves do not form a subgroup in general, see here for a discussion. $\endgroup$
    – Martin
    Jan 24, 2013 at 14:27
  • $\begingroup$ @Martin I see. Thanks for pointing out this. $\endgroup$
    – awllower
    Jan 24, 2013 at 14:43
  • $\begingroup$ @awllower $\endgroup$
    – Alexander Gruber
    Jan 24, 2013 at 17:33

1 Answer 1


I am sorry but I do not have a complete answer at the moment.

I will show that if $G$ is finitely generated and there are only finitely many commutators, then $D(G)$ is finite.

We will prove that the center of $G$, $Z(G)$ has finite index in $G$, then a theorem by Schur will allow us to conclude.

If $G$ is finitely generated, take $g_1,\ldots,g_n\in G$ to be a generating set. We have $$Z(G)=\bigcap_{i=1}^n Z(g_i),$$ where $Z(g)$ is the centralizer of $g$ in $G$.

It suffices to prove that every $Z(g_i)$ has finite index in $G$; this is the same as asking to verify that every element $g_i$ has finitely many conjugates.

Take $h\in G$. We have $hgh^{-1}=g[g^{-1},h]$. Since there are only finitely many commutators, every $g\in G$ has finitely many conjugates.

Hence $Z(g_i)$ is of finite index in $G$, for every $i=1\ldots n$.

Now, we have the following theorem

Theorem (Schur) If $[G:Z(G)]<\infty$, then $D(G)$ is finite.

I learned about this theorem in the italian book Gruppi by Antonio Machì. There is now a translated edition where it is Theorem 2.38 on page 84.

  • 9
    $\begingroup$ But if $G$ has only finitely many commutators $[g_i,h_i]$, then $D(G) = D(H)$ with $H = \langle g_i,h_i \rangle$ finitely generated, so the result is true for arbitrary $G$. $\endgroup$
    – Derek Holt
    Jan 24, 2013 at 16:30
  • $\begingroup$ You are right! Thanks $\endgroup$ Jan 24, 2013 at 17:27
  • 2
    $\begingroup$ I added a precise reference. Here are screen shots of the relevant passage: part 1 and part 2. $\endgroup$
    – Martin
    Apr 29, 2013 at 19:25
  • $\begingroup$ The same proof can be found in Problems in Group Theory by J. Dixon (problems 5.21 to 5.24). $\endgroup$
    – Seirios
    May 6, 2013 at 16:18

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.