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

how a free group of rank greater than or equal to 3 contains every free group of countable ranks as a sugroup?

share|cite|improve this question
math.stackexchange would have been more appropriate for this question. – Alain Valette Sep 8 '11 at 19:06
Given that so many people agree with Alain, why has no one else voted to close this question? – HJRW Sep 13 '11 at 14:03

Actually a free group of rank 2 contains a free group of countable rank as a subgroup, namely the commutator subgroup. The easiest way to see this is to view the free group as the fundamental group of a wedge of 2 circles. The covering space associated to the commutator subgroup is the Cayley graph of ZxZ, or if you like, the grid formed by the integer lattice in R^2. Clearly the fundamental group of this covering space is free of countable rank since there are countably infinitely many edges off a spanning tree. You can take the x-axis and all vertical lines as the spanning tree.

share|cite|improve this answer

In case an explicit description of such a subgroup would help, let $F_2 = \langle a,b|\rangle$ be the free group on 2 generators. Then the subgroup generated by the elements $\{a^nb^n\}_{n\in\mathbb{N}}$ is free since it is a subgroup of a free group. Next, it is easy to check that the generator $a^kb^k$ is not an element of the free group generated by the $a^\ell b^\ell$ for $\ell\neq k$ so one concludes that this subgroup has infinite rank.

share|cite|improve this answer

Even better, no normal subgroup of infinite index of a group of cohomological dimension at most two is finitely presented (for free groups, it must be free, so countable rank).

EDIT Bieri's theorem does state:

If $G$ is a group of cohomological dimension of at most two, while $N$ is a normal subgroup of G of infinite index, then either $N$ is free, or $N$ is not finitely presentable. Sigh. I was thinking of $G$ a free or a surface group, in which case my statement is correct (it is a theorem of Jaco (easy mod a not so easy theorem of Whitehead) that a subgroup of a surface group is free if and only if it is of infinite index...

share|cite|improve this answer
This isn't quite right: $F\times\mathbb{Z}$ is a counterexample! I think Bieri proved that no non-free normal subgroup of infinite index of a group of cohomological dimension two is finitely presented. – HJRW Sep 8 '11 at 19:58
$\mathbb{Z} \times \mathbb{Z}$ is even better. ;) – Richard Kent Sep 8 '11 at 20:00
@HW and @Richard: indeed. – Igor Rivin Sep 8 '11 at 20:37
In any event, it is trivial to prove that a finitely generated subgroup of a free group that is normal must have finite index. The covering space of a wedge of circles associated to a finitely generated subgroup of infinite index has a finite core (the Stallings graph) with a bunch of infinite trees attached to it. The covering space associated to a normal subgroup is a Cayley graph of the quotient group and looks the same at each vertex. Clearly, these two situations cannot hold at the same time. – Benjamin Steinberg Sep 8 '11 at 21:58
Benjamin - I never actually look at the proofs in Lyndon and Schupp. – HJRW Sep 13 '11 at 14:02

There is a very nice theorem which states:

Thm: If $F$ is a free group on $n$ generators and $H$ is a normal subgroup of index $j$ then if both $n$ and $j$ are finite $H$ is a free group on $j(n-1)+1$ generators. If $n$ is infinite and $j$ is finite then $H$ is a free group on infinitely many generators. Finally, if $j$ is infinite, then $H$ may be finitely or infinitely generated; however, if $H$ contains a normal subgroup $N$ of $F$, $N\neq 1$, then $H$ is a free group on infinitely many generators.

This is theorem 2.10 in Magnus, Karrass and Solitar "Combinatorial Group Theory" (and has a name...Schreier's formula?)

Once you have this theorem you can prove, for example, that the commutator subgroup is a free group on infinitely many generators (it contains a normal subgroup and is itself characteristic so you can apply the last line of the theorem). But doing it the way Benjamin Steinberg did it is much neater. I just thought it would be useful to mention this formula!

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.