Take the 2-minute tour ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

An interesting fact was relayed to me in another question of mine that

If $M$ is any closed manifold with universal cover homeomorphic to $R^n$ for $n>1$ then $\pi_1(M)$ is freely indecomposable.

What are some other sufficient conditions for the free-indecomposability of a group? Are there any interesting necessary conditions?

share|improve this question
    
see Henry Wilton's answer to your other question about groups with infinitely many ends. –  Ian Agol Mar 10 '11 at 4:43
1  
As this question doesn't have one correct answer, it should probably be community wiki. –  HJRW Mar 10 '11 at 5:00

1 Answer 1

up vote 3 down vote accepted

There are some more-or-less equivalent conditions:

  • Bass--Serre theory says that a group is freely indecomposable if and only if it acts on a tree with trivial edge stabilizers and no global fixed point.

  • More deeply, Stallings' Ends Theorem asserts that a finitely generated group splits over a finite subgroup if and only if it has more than one end.

As these are equivalent (modulo finite=trivial), you might say that they're not very interesting. Truly interesting necessary conditions are rather hard to write down, since being freely indecomposable is, in a sense, generic.

Here's one, much stronger, sufficient condition:

If a group $\Gamma$ has a finite generating set $S$ such that every element of $S$ is torsion and every element of $S^2$ is torsion then $\Gamma$ doesn't split at all, let alone freely.

This is essentially a consequence of Helly's Theorem for trees.


Here's an outline of a proof. Suppose that $\Gamma$ acts on a tree $T$. We want to prove that $\Gamma$ fixes a point.

  1. Prove that if $\gamma\in\Gamma$ has finite order then $\gamma$ fixes a point.
  2. Prove that if the sets of fixed points of $\gamma,\delta\in\Gamma$ are disjoint then $\gamma\delta$ has infinite order. You can do this by constructing an isometric copy of $\mathbb{R}$ in $T$ that $\gamma\delta$ translates.
  3. Prove Helly's Theorem for trees: if $\{T_i}$ is a finite set of pairwise-intersecting subtrees of $T$ then $\bigcap_i T_i\neq \varnothing$.
  4. Conclude as follows. Because every $s\in S$ is torsion, each such $s$ has a fixed point. Because $st$ is torsion for $s,t\in S$, the sets of fixed points pairwise intersect. By Helly's Theorem, there is a point fixed by every $s\in S$, and hence by $\Gamma$.
share|improve this answer
    
I apologize for the ignorance, but is that a variation/application of Helly's theorem on convex sets? Could you point me to a proof of the sufficient condition above? –  JeremyKun Mar 11 '11 at 3:30
    
One proof is basically Example 6.3.1 in Serre's book Trees. Serre's proof is rather algebraic, but the proof I had in mind is more geometric. See the edit above. –  HJRW Mar 11 '11 at 22:04

Your Answer

 
discard

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.