4
$\begingroup$

Let $G$ be an infinite finitely generated group. Fix a finite generating set for $G$.

Define $\mathrm{len}_G:G\to\mathbb{Z}_{\geq 0}$ by sending $g$ to the minimum length of a word in the generators and their inverses equal to $g$.

Let $H\subset G$ is an infinite finitely generated subgroup. Fix a finite generating set for $H$.

Question. Under what conditions is there a computable function $m \colon \mathbb{Z}_{\geq 0}\to\mathbb{Z}_{\geq 0}$ such that for all $h\in H$ the inequality $$ \mathrm{len}_H(h)\leq m(\mathrm{len}_G(h)) $$ holds?

$\endgroup$
4
  • 3
    $\begingroup$ If $G$ has solvable word problem, I suspect this holds iff the membership problem in $H$ is solvable. $\endgroup$
    – YCor
    Commented Apr 26, 2021 at 12:16
  • $\begingroup$ This is the notion of (actual) distortion function, which appears in a paper Sunik, Margolis & Meakin. @YCor's suspicion is correct, and can be strengthened slightly: if the problem of comparison of $H$-words to $G$-words is decidable in $G$, then this holds iff the membership problem in $H$ is solvable. $\endgroup$ Commented Apr 26, 2021 at 16:42
  • $\begingroup$ Specifically, it is Proposition~1.1 of their paper. $\endgroup$ Commented Apr 26, 2021 at 16:42
  • $\begingroup$ @Carl-FredrikNybergBrodda thanks! I think the notion of distortion of a subgroup appears in papers by Gromov and Gersten in the early 90s. $\endgroup$
    – YCor
    Commented Apr 26, 2021 at 16:46

1 Answer 1

4
$\begingroup$

I'll put my comment here, so that the question has an answer.

The function $\operatorname{len}_H$ is what is called an actual distortion function (for $H$ in $G$) by Margolis, Meakin & Šuniḱ (see [1]). This is a notion that has been studied before in various forms in some papers by e.g. Gromov and Gersten (as @YCor points out). In their paper, MMS prove (Proposition 1.1) the rather straightforward result that as long as the problem of comparing $H$-words with $G$-words is decidable (which is no harder than the word problem for $G$), then a recursive actual distortion function for $H$ in $G$ exists if and only if the membership problem for $H$ in $G$ is decidable. The proof is very short and elementary.

[1] Margolis, Stuart W.; Meakin, John; Šuniḱ, Zoran, Distortion functions and the membership problem for submonoids of groups and monoids., Geometric methods in group theory. Contemporary Mathematics 372, 109-129 (2005). ZBL1103.20049.

$\endgroup$

You must log in to answer this question.