MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

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

We write $cl$ for the commutator length, i.e. the least number of commutators which multiply to a given element of a group.

Given an element $g$ in the commutator subgroup of the free group $G=F_2$ on two generators, is it true that $$cl_G(g) = \displaystyle \max_{\mbox{H < G finite index normal}} cl_{G/H} (g \mod H)$$ ?

share|cite|improve this question
On the RHS, you mean $\text{cl}_{G/H}$, right? – Andy Putman Jan 14 '13 at 18:49
Indeed, I changed that. – dlbb Jan 14 '13 at 19:37
up vote 11 down vote accepted

I think the answer is no and that actually the "finite quotient" commutator length (defined by your formula) is bounded on $[F_2,F_2]$.

Indeed by Nikolov-Segal, in the profinite completion $P$ of $F_2$, the derived subgroup $[P,P]$ is closed; since the set $C$ of commutators is compact, it follows by a Baire argument that $[P,P]$ is boundedly generated by $C$ (first get by Baire that some open neighborhood of the identity of $[P,P]$ has bounded commutator length in $P$ and then use compactness). Observe on the other hand that $[F_2,F_2]=[P,P]\cap F_2$, which is essentially trivial since the abelian group $F_2/[F_2,F_2]$ is residually finite. So in any finite quotient of $F_2$, the commutator length of $w\in [F_2,F_2]$ is bounded by the universal number that arises as upper bound of the commutator length of $[P,P]$ in $P$.

On the other hand, the commutator length of $F_2=\langle x,y\rangle$ is unbounded on $[F_2,F_2]$, as the commutator length of $[x,y]^n$ grows linearly (I think it's due to Bavard).

share|cite|improve this answer
Nice argument; however, looking at the Nikolov-Segal paper, it appears that the way that they show that $[P,P]$ is closed is to show that every element in it is a product of a bounded number of commutators: Of course, your argument shows why these things are equivalent. – Ian Agol Jan 15 '13 at 6:48
So, this must be a well known fact: 'There is a universal bound on the commutator length in all 2-generated finite groups'. What's the reference? Does it apply to n-generated finite groups for all n? – HJRW Jan 15 '13 at 8:17
@HW The reference is that same paper. See the last sentence of the abstract. – Sean Eberhard Jan 15 '13 at 9:43
Sean - ah yes, thanks. – HJRW Jan 15 '13 at 10:08
This is very surprising, thank you very much for this answer! – dlbb Jan 15 '13 at 12:21

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.