Let $G$ be a finitely generated residually finite group with positive rank gradient, and let $F_2$ be the free group on $2$ elements. Must there be an embedding $i \colon F_2 \to G$ ?

A group $G$ is called residually finite if the intersection of all of its finite index subgroups is trivial.

A group $G$ is said to have positive rank gradient if the rank of finite index subgroups grows linearly with the index. More formally, the Rank Gradient (RG) of a finitely generated group $G$ is defined to be:

$$\mathrm{RG}(G) = \inf_{H} \frac{\mathrm{rank}(H) - 1}{[G : H]}$$ where $H$ runs over all subgroups of finite index in $G$, and the rank of a group is the smallest cardinality of a generating set for the group.

The notion of rank gradient is also important in the theory of $3$-manifolds.

  • 1
    $\begingroup$ Silly remark: "residually finite" (RF) is unnecessary, since if it's true in the RF case, letting $N$ be the intersection of finite index subgroups of $G$, we can apply the result to $G/N$ (noting that RG$(G)=$RG$(G/N)$) and lift $F_2$ to $G$. $\endgroup$ – YCor Sep 1 '15 at 13:44


Denis Osin, Rank gradient and torsion groups. Bull. Lond. Math. Soc. 43 (2011), no. 1, 10–16,

the following theorem is proved

Theorem For every prime $p$, there exists a finitely generated infinite residually finite $p$-group with positive rank gradient.



  • $\begingroup$ I am not completely sure how to parse the statement of the theorem. Should "residually finite $p$-group" be read together, meaning that the intersection of subgroups of index a power of $p$ is trivial ? $\endgroup$ – Joël Sep 1 '15 at 14:37
  • $\begingroup$ @Joel But I think "(residually finite) $p$-group" is equivalent to "residually (finite $p$-group)". $\endgroup$ – Derek Holt Sep 1 '15 at 14:46
  • $\begingroup$ I am still confused. What is a $p$-group? For me it is a finite group of order a power of $p$. $\endgroup$ – Joël Sep 1 '15 at 14:56
  • $\begingroup$ p-group means each element has order a p-power. It need not be finite. Like the Grigorchuk group $\endgroup$ – Benjamin Steinberg Sep 1 '15 at 15:11
  • 1
    $\begingroup$ It is worth mentioning that Jan-Christoph Schlage-Puchta proved a similar result independently, see his paper: A p-group with positive rank gradient. J. Group Theory 15 (2012), no. 2, 261–270. You might also like to look at a joint paper with me: On p-deficiency in groups. J. Group Theory 16 (2013), no. 4, 497–517. (Both can be found on the Arxiv.) $\endgroup$ – Yiftach Barnea Sep 1 '15 at 18:33

Lackenby ( http://people.maths.ox.ac.uk/lackenby/lg070105.ps ) proved that a finitely presented group, which has a pro-$p$ completion of positive rank gradient, is large, i.e. it contains a subgroup of finite index that projects onto a non-abelian free group. So while the answer to the original question is no, if you add some extra conditions it becomes yes in a very strong sense.

  • $\begingroup$ I see that in the paper there are two more conditions (apart from positivity of the rank gradient) needed to conclude that the group is large. Is your claim correct? $\endgroup$ – Pablo Sep 6 '15 at 9:56
  • 1
    $\begingroup$ Sorry, when thinking about rank gradient I am usually only interested in the pro-$p$-case, so I remembered Lackenby's result n a pro-$p$ way. I corrected my answer accordingly. $\endgroup$ – Jan-Christoph Schlage-Puchta Sep 10 '15 at 20:04

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.