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

Let $G$ be an at most countable discrete group acting freely on a standard probability measure space $X$ in a measure preserving way.

It is well known that if $G$ is a finite group then this action admits a fundamental domain. As pointed out by Andreas below, by Rokhlin lemma, if $G$ contains an element of infinite order we can find an $(\varepsilon, N)$-fundamentalish domain $U$, where the latter is defined as follows:

Call a set $U\subset X$ an $(\varepsilon, N)$-fundamentalish domain iff there exist $N$ elements $g_1, \ldots, g_N$ of $G$ such that the sets $g_i(U)$ are pairwise disjoint and the measure of their union is at least $1-\varepsilon$.

Question: If $G$ is an infinite group, $N_0$ is a natural number, $\varepsilon_0$ is a positive real number, does there exist an $(\varepsilon, N)$-fundamentalish domain with $\varepsilon<\varepsilon_0$ and $N>N_0$?

For example when the action is profinite and "transitive on each level", then clearly answer is positive: there exist $(0,N)$-fundamentalish domains for arbitrary large $N$.

share|cite|improve this question
Exactly :-) Is there a Rokhlin lemma for action of an arbitary group (i.e. a Tarski monster)? I'll reformulate the question accordingly – Łukasz Grabowski Nov 11 '11 at 13:51

If you fix $N$ and group elements $g_1$, ..., $g_N \in G$, then your question becomes closely related to tilings of groups. Specifically, in Chapter 2, section 2 of "Entropy and Isomorphism Theorems for Actions of Amenable Groups," Ornstein and Weiss prove:

Let $G$ be a countable group acting freely and measure preservingly on a standard probability space $(X, \mu)$. Fix a finite set $T \subseteq G$. If for every $\epsilon > 0$ there is a measurable set $U \subseteq X$ such that the $T$-translates of $U$ are disjoint and $\mu(T \cdot U) > 1 - \epsilon$, then $T$ tiles $G$ in the sense that there is a set of centers $C \subseteq G$ such that the sets $Tc$ ($c \in C$) partition $G$.

They also prove that if $G$ is amenable then the converse holds. So if $G$ is amenable and $T$ is a tile for $G$, then for every free probability measure preserving action of $G$ and every $\epsilon > 0$ there is a $(\epsilon, |T|)$-fundamentalish domain. Thus a natural question is: which amenable groups admit arbitrarily large finite tiles?

Weiss called a group $G$ MT (mono-tileable) if for every finite set $F \subseteq G$ there is a finite tile $T \subseteq G$ containing $F$. In "Monotileable amenable groups," Weiss proved that all solvable groups and all residually finite groups are MT. In "Elementary Amenable Groups," Chou proved that all elementary amenable groups and all free products of non-trivial groups are MT. So in particular, your question has a positive answer whenever the group $G$ is elementary amenable. A stronger tiling condition, called ccc, is studied in chapter 4 of "Groups Colorings and Bernoulli Subflows" (this paper is in preparation). Weaker properties of poly-MT and poly-ccc are studied in "Burnside's Problem, spanning trees, and tilings." To the best of my knowledge these are the only papers which study tilings of countable groups.

share|cite|improve this answer
Thanks Brandon, curiously, this question was more-or-less an outcome of my reading the arxiv preprint you updated a couple of days ago... – Łukasz Grabowski Nov 14 '11 at 22:11

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.