Skip to Main Content

Geometric Group Theory

Matt Clay

Communicated by Notices Associate Editor Chikako Mese

Article cover

Introduction

Groups and spaces go hand in hand. For a given space, there are many groups associated to it. We can consider the group of symmetries, that is, the group of structure preserving bijections. Additionally, there is the fundamental group and also the homology and cohomology groups to name a few more. As pointed out by Hermann Weyl, these groups can give “a deep insight” into a given space. An example of this phenomenon is in the study of knots. Algebraic invariants in the form of groups show that the trefoil knot cannot be unknotted for instance. See Figure 1.

Figure 1.

Groups show that these knots are distinct.

Graphic without alt text

Geometric group theory takes a different perspective on this relationship between groups and spaces. Rather than using the algebraic structure and properties of groups to study spaces, the main philosophy of geometric group theory is the following.

Study groups using the topology and geometry of the spaces they act on.

That is, groups are the central objects of study and the techniques and tools used to investigate them are dynamical, geometrical, and topological in nature.

In name, geometric group theory is quite new in relation to other mathematical fields⁠Footnote1. The foundational essays by Gromov Gro87Gro93 introducing the notion of hyperbolic groups and initiating the study of finitely generated groups as metric spaces sparked an enormous amount of research and established lines of investigation that are still very active today. Prior to the emergence of geometric group theory, there were geometrical ideas present in group theory in the works of Dehn, Whitehead, van Kampen, and others. Additionally, Thurston’s work on 3–manifolds showed how the geometry of a manifold influences algebraic and algorithmic properties of its fundamental group. It is Gromov’s essays though that mark the beginning of the perspective where these ideas are at the forefront.

1

The earliest use of the “geometric group theory” I could find was in reference to a symposium at Sussex University in the summer of 1991.

This article is intended to give an idea about how the topology and geometry of a space influences the algebraic structure of groups that act on it and how this can be used to investigate groups. As you will see, I take the approach I learned from my advisor Mladen Bestvina of favoring illustrative examples over general theory. As is true of any survey of a mathematical field, many aspects and areas of geometric group theory are not mentioned at all. The final section includes a short list of books on geometric group theory for further reading.

Groups and Spaces

As mentioned above, geometric group theory uses group actions on spaces to understand the group’s structure. What type of information could one hope to glean from an action? Are there always interesting actions to study? We will take a look at both of these questions now.

An example:

To give an illustration of how the topology of a space that a group acts on influences the group’s structure, let’s take a look at an example of a group action that appears in many areas of mathematics. We will consider the group of matrices with integer entries and determinant equal to 1. This group is called the special linear group:

Is finitely generated? That is, are there finitely many matrices , …, such that any matrix in can be expressed as a product ? (Note, each may appear multiple times.) The answer is “yes” and there is an algebraic approach to this problem, but let’s take a geometric perspective and consider an action of on a metric space.

The space we will consider is the Farey complex which is constructed as follows. First, we start with a graph whose vertex set is the set of rational numbers —always expressed in lowest terms—along with an additional point we denote . Edges join two vertices and if . Figure 2 shows a portion of this graph, known as the Farey graph.

Figure 2.

The Farey graph and Farey complex.

Graphic without alt text

As seen in Figure 2, the edges in the Farey graph naturally form triangles. In fact, the vertices of any such triangle always have the form , and . For instance, and , and also and . There is an action of on the Farey graph defined by permuting the vertices using the rule:

It is easy to check that two vertices and are connected by an edge only if their images and are. Hence, this defines an action on the Farey graph and by extension on the Farey complex, which is the space we get by filling in the triangles in the Farey graph.

You have most likely seen this space and action before but under a different guise. Indeed, the Farey complex gives a tessellation of the hyperbolic plane by ideal triangles whose vertices in the upper half plane model are either rational or . Moreover, the action described above is none other than the usual action of matrices with real entries and positive determinant by fractional linear transformations of the upper half plane, in particular, by conformal maps. The conformal maps:

relate the two pictures. See Figure 3.

Figure 3.

The Farey tessellation of the upper half plane by ideal triangles.

Graphic without alt text

Now it is time to examine this action. Let denote the triangle in the Farey complex with vertices , , and . We record the key properties of the action in two claims.

Claim 1.

For any triangle in the Farey complex, there is matrix such that .

Indeed, suppose the vertices of are , , and where . Take and observe that .

Let . Notice that , , and so that and acts on the triangle by a rotation.

Claim 2.

If , then for some integer .

Indeed, if fixes , then it must cyclically permute the vertices , , and . Hence fixes the vertices , , and for some . As the only conformal map that fixes three points is the identity, we see that . (Note, acts as the identity map.) The claim follows once we check that .

Let and let be the triangle that shares an edge with and has the vertex . These are labeled in Figure 2. We observe that . As rotates , we also find that and .

We are now in the position to show that is finitely generated by the matrices and . That is, any matrix in can be expressed as a product of ’s and ’s:

for some integers . Given we want to consider paths in the Farey complex from to . What do we mean by path? Specifically, we mean a sequence of triangles , …, where the triangles and share an edge.

Now we proceed via induction on the length of shortest path to . Claim 2 handles the case that this length is 0. Next, using a path , …, of minimal length we observe by Claim 1 and induction that where can be expressed as product of ’s and ’s. Let’s hit the whole picture with : the triangle is sent to and the triangle is sent to an adjacent triangle, i.e., one of , , or . Assuming for simplicity that , which is equal to , we find that . Claim 2 now shows that and hence . Since can be expressed as a product of ’s and ’s, so can , showing that is finitely generated.

A theorem: characterizing finite generation

What did we actually use to prove finite generation? The important topological property we used was the path-connectedness of the Farey complex so that we had a path from to to apply induction on. The important dynamical property we used was the existence of a transitive tiling for which the stabilizer of a tile is finite and for which one tile meets only finitely many other tiles. These dynamical considerations naturally lead to the following definition.

Definition 1.

An action of a group on a metric space by isometries is geometric if it satisfies the following two conditions:

(1)

(cocompact) there exists a compact set such that ; and

(2)

(properly discontinuous) for any compact set , the set is finite.

The requirement of a transitive tiling is captured by the cocompact condition. The properly discontinuous condition captures both requirements that the stabilizer of a tile is finite and that a tile meets only finitely many tiles.

Technical Sidenote (i.e., feel free to ignore): The actions of on the Farey complex and on the upper half plane are not geometric. For the Farey complex the action is cocompact, but a triangle intersects infinitely many other triangles at a vertex, so the action is not properly discontinuous. We got around this problem by only considering triangles that meet along an edge—there are only finitely many such triangles. In the upper half plane the action is properly discontinuous, but the action is not cocompact. We can get around this by removing an equivariant collection of disjoint open disks tangent to the rational points. In either setting, the crucial point is that our notion of path ignores the vertices/ideal points. There is a geometric action lurking in the background here on the Farey tree that will be explored later.

Here are some examples of geometric actions.

(1)

The group acting by linearly independent translations on equipped with the Euclidean metric.

(2)

More generally, any group of isometries of equipped with the Euclidean metric that leaves a lattice invariant and whose action on the lattice has finitely many orbits.

(3)

The fundamental group of a compact Riemannian manifold , possibly with boundary, acting by deck transformations on its universal cover equipped with the pull-back metric.

Arguing as we did for , we can prove the “if” direction of a geometric characterization of finite generation.

Theorem 1.

A group is finitely generated if and only if it acts geometrically on a path-connected metric space.

For the “only if” direction, we need to introduce an important concept in geometric group theory: the Cayley graph.

A space for every group

For a finitely generated group we need to produce a path-connected metric space that admits a geometric action by . This is similar to what is required to prove Cayley’s theorem from classical group theory: Every group is isomorphic to a permutation group. In the classical setting, we need to produce a set that admits a permutation action by our group. There is only one natural choice, the set is the group and the action is left multiplication.

In our current setting, the idea is similar. The metric space is built on top of the group, the extra parts of the space come from a finite generating set. The result is called a Cayley graph. Here are the details.

Definition 2.

Let be a finitely generated group and let be a finite generating set. The Cayley graph, denoted , is the graph whose vertex set is and where there is an edge joining vertices if , i.e.,  for some generator .

The group acts on by permuting the vertices via left multiplication. Indeed, if the vertices are adjacent, then so are the vertices as and so the permutation action on the vertices extends to the entire graph.

As generates , the Cayley graph is path-connected. Figure 4 illustrates the path connecting the identity element of the group to the element where each belongs to . The key point is that is adjacent to .

Figure 4.

A path in the Cayley graph.

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=0.7] \begin{scope} \tikzstyle{every node}=[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] \tikzstyle{every path}=[thick,blue] \draw(0,0) node(a) {} -- (1.5,1) node(b) {} node[fill=none,pos=0.5](e){}; \draw(b) -- (3,0) node(c) {} node[fill=none,pos=0.5](f){}; \draw(c) -- (3.75,0.5); \filldraw[black] (4.5,0.5) circle [radius=0.03cm]; \filldraw[black] (5,0.5) circle [radius=0.03cm]; \filldraw[black] (5.5,0.5) circle [radius=0.03cm]; \draw(7,1) node(d) {} -- (8.5,0) node(g) {} node[fill=none,pos=0.5](h){}; \draw(6.25,0.5) -- (d); \end{scope} \node[left] at (a) {$1_G$}; \node[above] at (b) {$s_1$}; \node[below] at (c) {$s_1s_2$}; \node[above] at (d) {$s_1s_2\cdots s_{k-1}$}; \node[below] at (g) {$g = s_1s_2\cdots s_k$}; \end{tikzpicture}

Here are some examples of Cayley graphs.

(1)

and : For , we can use and for we can use . These graphs are pictured in Figure 5. Other generating sets are possible too, try drawing the graph . You can find this graph in the essay by Margalit and Thomas CM17, Office Hour 7.

Figure 5.

Cayley graphs for and .

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=0.7] \begin{scope} \foreach\n[remember=\n as \lastn(initially 0)] in {1,2,3} { \draw[thick,blue] (1.3*\lastn,0) -- (1.3*\n,0); } \draw[thick,blue] (-0.5,0) -- (0,0) (3.9,0) -- (4.4,0); \foreach\n in {0,...,3} { \filldraw[red] (1.3*\n,0) circle [radius=0.075cm]; } \node at (-0.2,-0.45) {$-1$}; \node at (1.3,-0.45) {$0$}; \node at (2.6,-0.45) {$1$}; \node at (03.9,-0.45) {$2$}; \end{scope} \begin{scope}[xshift=6cm,yshift=-1.9cm] \foreach\x[remember=\x as \lastx(initially 0)] in {1,2,3} { \draw[thick,blue] (1.3*\lastx,0) -- (1.3*\x,0); \draw[thick,green] (0,1.3*\lastx) -- (0,1.3*\x); \foreach\y[remember=\y as \lasty(initially 0)] in {1,2,3} { \draw[thick,blue] (1.3*\lastx,1.3*\y) -- (1.3*\x,1.3*\y); \draw[thick,green] (1.3*\x,1.3*\lasty) -- (1.3*\x,1.3*\y); }} \foreach\x in {0,...,3} { \draw[thick,green] (1.3*\x,-0.5) -- (1.3*\x,0) (1.3*\x,3.9) -- (1.3*\x,4.4); \draw[thick,blue] (-0.5,1.3*\x) -- (0,1.3*\x) (3.9,1.3*\x) -- (4.4,1.3*\x); } \foreach\x in {0,...,3} { \foreach\y in {0,...,3} { \filldraw[red] (1.3*\x,1.3*\y) circle [radius=0.075cm]; }} \node at (1.7,1.7) {$\left[\begin{smallmatrix} 0 \\ 0\end{smallmatrix}\right]$}; \node at (3,1.7) {$\left[\begin{smallmatrix} 1 \\ 0\end{smallmatrix}\right]$}; \node at (1.7,3) {$\left[\begin{smallmatrix} 0 \\ 1\end{smallmatrix}\right]$}; \node at (3,3) {$\left[\begin{smallmatrix} 1 \\ 1\end{smallmatrix}\right]$}; \end{scope} \end{tikzpicture}
(2)

: For the symmetric group on three elements, we can use the generating sets or . These graphs are pictured in Figure 6 where elements in are listed using cycle notation and the composition is computed right to left.

Figure 6.

Cayley graphs for .

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=0.9] \tikzstyle{every path}=[thick] \begin{scope} \node[regular polygon,regular polygon sides=6,minimum size=2.25cm] (p) at (0,0) {}; \draw[blue] (p.corner 3) -- (p.corner 2); \draw[blue] (p.corner 4) -- (p.corner 5); \draw[blue] (p.corner 6) -- (p.corner 1); \draw[green] (p.corner 2) -- (p.corner 1); \draw[green] (p.corner 3) -- (p.corner 4); \draw[green] (p.corner 5) -- (p.corner 6); \node[above right] at (p.corner 1) {\small$(1 \; 3 \; 2)$}; \node[above left] at (p.corner 2) {\small$(1 \; 2)$}; \node[left] at (p.corner 3) {\small$(\;)$}; \node[below left] at (p.corner 4) {\small$(1 \; 3)$}; \node[below right] at (p.corner 5) {\small$(1 \; 2 \; 3)$}; \node[right] at (p.corner 6) {\small$(2 \; 3)$}; \foreach\a in {1,...,6} { \node[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] at (p.corner \a) {}; } \end{scope} \begin{scope}[xshift=4.5cm,yshift=-0.5cm] \node[regular polygon,regular polygon sides=3,minimum size=4cm] (p) at (0,0) {}; \node[regular polygon,regular polygon sides=3,minimum size=2.5cm] (q) at (0,0) {}; \draw[green] (p.corner 1) -- (p.corner 2) -- (p.corner 3) -- cycle; \draw[green] (q.corner 1) -- (q.corner 2) -- (q.corner 3) -- cycle; \node[above] at (p.corner 1) {\small$(1 \; 3 \; 2)$}; \node[label={[shift={(0.6,-0.2)}]\small$(1 \; 2)$}] at (q.corner 2) {}; \node[below] at (p.corner 2) {\small$(\;)$}; \node[label={[shift={(0.0,-0.9)}]\small$(2 \; 3)$}] at (q.corner 1) {}; \node[below] at (p.corner 3) {\small$(1 \; 2 \; 3)$}; \node[label={[shift={(-0.6,-0.2)}]\small$(1 \; 3)$}] at (q.corner 3) {}; \foreach\a in {1,...,3} { \draw[blue] (q.corner \a) -- (p.corner \a); \node[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] at (p.corner \a) {}; \node[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] at (q.corner \a) {}; } \end{scope} \end{tikzpicture}
(3)

: For the free group of rank two, we can use a basis . Recall that elements in are in one-to-one correspondence to words in the alphabet that are reduced in the sense that they do not contain , , , or . For example, and represent elements in . The identity in is represented by the empty word. The group operation is concatenation followed by deletion of forbidden terms. As the reduced word representing an element is unique and as paths in the Cayley graphs read out a word representing an element as shown in Figure 4, there is a unique non-backtracking path from to any given element. Hence, the Cayley graph is a tree. A portion of this graph is pictured in Figure 7.

Figure 7.

A Cayley graph for .

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=1] \begin{scope} \tikzstyle{every node}=[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] \tikzstyle{every path}=[thick] \draw[blue] (-3.5,-0.2) -- (-3.5,0.2) (3.5,-0.2) -- (3.5,0.2); \draw[green] (-0.2,3.5) -- (0.2,3.5) (-0.2,-3.5) -- (0.2,-3.5); \draw[green] (-3.2,0.5) -- (-2.8,0.5) (2.8,0.5) -- (3.2,0.5); \draw[green] (-3.2,-0.5) -- (-2.8,-0.5) (2.8,-0.5) -- (3.2,-0.5); \draw[blue] (-0.5,2.8) -- (-0.5,3.2) (0.5,2.8) -- (0.5,3.2); \draw[blue] (-0.5,-2.8) -- (-0.5,-3.2) (0.5,-2.8) -- (0.5,-3.2); \draw[green] (1.8,1.5) -- (2.2,1.5) (1.8,-1.5) -- (2.2,-1.5); \draw[green] (-1.8,1.5) -- (-2.2,1.5) (-1.8,-1.5) -- (-2.2,-1.5); \draw[blue] (1.5,1.8) -- (1.5,2.2) (-1.5,1.8) -- (-1.5,2.2); \draw[blue] (1.5,-1.8) -- (1.5,-2.2) (-1.5,-1.8) -- (-1.5,-2.2); \draw[blue] (1.5,0.8) -- (1.5,1.2) (2.5,0.8) -- (2.5,1.2); \draw[blue] (1.5,-0.8) -- (1.5,-1.2) (2.5,-0.8) -- (2.5,-1.2); \draw[blue] (-1.5,0.8) -- (-1.5,1.2) (-2.5,0.8) -- (-2.5,1.2); \draw[blue] (-1.5,-0.8) -- (-1.5,-1.2) (-2.5,-0.8) -- (-2.5,-1.2); \draw[green] (0.8,2.5) -- (1.2,2.5) (0.8,1.5) -- (1.2,1.5); \draw[green] (-0.8,2.5) -- (-1.2,2.5) (-0.8,1.5) -- (-1.2,1.5); \draw[green] (0.8,-2.5) -- (1.2,-2.5) (0.8,-1.5) -- (1.2,-1.5); \draw[green] (-0.8,-2.5) -- (-1.2,-2.5) (-0.8,-1.5) -- (-1.2,-1.5); \draw[blue] (3,-0.7) -- (3,-0.5) node {} -- (3,0.5) node {} -- (3,0.7); \draw[blue] (-3,-0.7) -- (-3,-0.5) node {} -- (-3,0.5) node {} -- (-3,0.7); \draw[green] (-0.7,3) -- (-0.5,3) node {} -- (0.5,3) node {} -- (0.7,3); \draw[green] (-0.7,-3) -- (-0.5,-3) node {} -- (0.5,-3) node {} -- (0.7,-3); \draw[green] (1.3,1) -- (1.5,1) node {} -- (2.5,1) node {} -- (2.7,1); \draw[green] (1.3,-1) -- (1.5,-1) node {} -- (2.5,-1) node {} -- (2.7,-1); \draw[green] (-1.3,1) -- (-1.5,1) node {} -- (-2.5,1) node {} -- (-2.7,1); \draw[green] (-1.3,-1) -- (-1.5,-1) node {} -- (-2.5,-1) node {} -- (-2.7,-1); \draw[green] (-2,0) node(A) {} -- (0,0) node(p) {}; \draw[blue] (1,1.3) -- (1,1.5) node {} -- (1,2.5) node {} -- (1,2.7); \draw[blue] (-1,1.3) -- (-1,1.5) node {} -- (-1,2.5) node {} -- (-1,2.7); \draw[blue] (1,-1.3) -- (1,-1.5) node {} -- (1,-2.5) node {} -- (1,-2.7); \draw[blue] (-1,-1.3) -- (-1,-1.5) node {} -- (-1,-2.5) node {} -- (-1,-2.7); \draw[green] (p) -- (2,0) node(a) {}; \draw[blue] (0,-2) node(B) {} -- (p); \draw[blue] (p) -- (0,2) node(b) {}; \draw[green] (a) -- (3,0) node(aa) {} -- (3.5,0) node {} -- (3.7,0); \draw[blue] (a) -- (2,1) node(ab) {} -- (2,1.5) node {} -- (2,1.7); \draw[blue] (2,-1.7) -- (2,-1.5) node {} -- (2,-1) node(aB) {} -- (a); \draw[green] (-3.7,0) -- (-3.5,0) node {} -- (-3,0) node(AA) {} -- (A); \draw[blue] (A) -- (-2,1) node(Ab) {} -- (-2,1.5) node {} -- (-2,1.7); \draw[blue] (-2,-1.7) -- (-2,-1.5) node {} -- (-2,-1) node(AB) {} -- (A); \draw[blue] (b) -- (0,3) node(bb) {} -- (0,3.5) node {} -- (0,3.7); \draw[green] (b) -- (1,2) node(ba) {} -- (1.5,2) node {} -- (1.7,2); \draw[green] (-1.7,2) -- (-1.5,2) node {} -- (-1,2) node(bA) {} -- (b); \draw[blue] (0,-3.7) -- (0,-3.5) node {} -- (0,-3) node(BB) {} -- (B); \draw[green] (B) -- (1,-2) node(Ba) {} -- (1.5,-2) node {} -- (1.7,-2); \draw[green] (-1.7,-2) -- (-1.5,-2) node {} -- (-1,-2) node(BA) {} -- (B); \end{scope} \node[above right] at (p) {\small$1_{F_2}$}; \node[above right] at (A) {\small$a^{-1}$}; \node[above right] at (a) {\small$a$}; \node[above right] at (B) {\small$b^{-1}$}; \node[above right] at (b) {\small$b$}; \node[above right] at (aa) {\small$a^2$}; \node[above right] at (aB) {\small$ab^{-1}$}; \node[above right] at (ab) {\small$ab$}; \node[above right] at (AA) {\small$a^{-2}$}; \node[above right] at (AB) {\small$a^{-1}b^{-1}$}; \node[above right] at (Ab) {\small$a^{-1}b$}; \node[above right] at (bA) {\small$ba^{-1}$}; \node[above right] at (ba) {\small$ba$}; \node[above right] at (bb) {\small$b^2$}; \node[below left] at (BA) {\small$b^{-1}a^{-1}$}; \node[above right] at (Ba) {\small$b^{-1}a$}; \node[above right] at (BB) {\small$b^{-2}$}; \end{tikzpicture}

There is a metric on the vertices of defined as the minimum number of edges in an edge-path between a given pair of vertices. This metric can be extended to the points lying in edges by identifying (in an equivariant way) each edge with the unit interval . However for most applications in geometric group theory, having a metric only on the vertices suffices. The action of on the Cayley graph with this metric is by isometries.

The only item left to verify in Theorem 1 is that the action of on is geometric. We can easily check these properties in turn.

(1)

(cocompact) Let be the union of the vertices together with the edges incident on and for each . As is finite, is compact and clearly .

(2)

(properly discontinuous) Suppose that is a finite subgraph and let denote the number of vertices in . If then for a pair of vertices in and hence . Thus the cardinality of is at most .

Groups and Spaces with Negative Curvature

In the previous section, we used a path-connected space and a geometric action to derive an algebraic consequence: finite generation. Path-connectivity is a fairly weak topological property, however the notion of a geometric action is quite restrictive. For instance, by proper discontinuity the subgroup fixing a given point must be finite. What can be gained from actions on spaces with more requirements on the topology and geometry, but perhaps fewer requirements on the dynamics of the action?

One geometric property that is particularly useful is the notion of negative curvature. We will look at two instances of negative curvature in geometric group theory: trees and –hyperbolic spaces.

Actions on trees

Negative curvature, say in the hyperbolic plane, influences the geometry in several ways: uniqueness of geodesics, exponential growth in the volume of balls, and a uniform bound on the diameter of an inscribed circle to a triangle to name a few. To discuss the familiar notion of curvature from differential geometry, a space requires more structure than just an ordinary metric, but several researchers have given notions of negative curvature expressed solely in terms of a distance function on an arbitrary set. Before discussing such a notion of negative curvature, let’s consider a simple example of a metric space that has the properties listed above for the hyperbolic plane: a tree.

To see an example of the usefulness of group actions on trees, let’s go back to the example of and think about its finite-order elements, i.e., matrices for which some positive power is equal to the identity. We can quickly compute that , thus and and so and have finite order. Are there any others? There are obvious ones of course. Powers of and powers of clearly have finite order, as do their conjugates, and for any and . But is that it? The answer to this last question is “yes” and we will see why using the action of on the Farey tree, which we now describe.

Divide each triangle in the Farey complex into three quadrilaterals that meet pairwise along one leg of a tripod. Taken collectively these tripods form a tree, which is called the Farey tree. See Figure 8.

Figure 8.

The Farey tree.

Graphic without alt text

There are two types of vertices in the Farey tree: (red) degree three coming from the center of a triangle, and (green) degree two coming from an edge of a triangle. Let denote the vertex that corresponds to the center of the triangle and let denote the vertex that corresponds to the edge in the Farey complex between and . These are labeled in Figure 8.

From our study of the action of on the Farey complex, we conclude that every vertex in the Farey tree is a translate of or . This follows from Claim 1 and the fact that cyclically permutes the edges of and hence all of the vertices adjacent to . Additionally, we can conclude from Claim 2 that the stabilizer of is the cyclic subgroup of order 6 generated by . In a similar manner, we can conclude that the stabilizer of is the cyclic subgroup of order 4 generated by .

An important property of an action on a tree is the following claim.

Claim 3.

Suppose that a group acts on a tree. If has finite order, then has a fixed point.

The key fact here is that a finite set of points , …, in a tree has a unique center, i.e., a point that minimizes the quantity

The center is easy to characterize. Suppose that and maximize for , , …, . One can show that the center is the unique point with . Now fix a point in the tree and let be the center of the set where is the order of . Since the action is by isometries, we must have that is the center of the set . But permutes the points in , i.e., , and so .

Applying Claim 3 to the action of on the Farey tree, we see if has finite order, then for some point in this tree. If fixes a point in the interior of an edge, then it must fix one of the incident vertices as well since these vertices have different degrees and cannot be interchanged by . So we may assume that is a vertex of the Farey tree. As every vertex is a translate of or , we have that or for some matrix . In the former, we observe that and so for some . Similarly, in the latter, we conclude that for some . Hence every finite-order element in is conjugate to a power of or . This is exactly what we desired to show.

The action of on the Farey tree is geometric. The argument we gave shows that if a group acts geometrically on a tree, then there are only finitely many conjugacy classes of finite-order elements. Indeed, by Claim 3 and since the action is cocompact, any finite order element is conjugate into one of finitely many stabilizer subgroups. Since the action is properly discontinuous, each of these subgroups is finite and so the result follows.

We can replace the assumption of proper discontinuity of the action with the assumption that each point stabilizer subgroup has finitely many conjugacy classes of finite-order elements and reach the same conclusion.

Theorem 2.

Suppose acts cocompactly on a tree. If every point stabilizer has finitely many conjugacy classes of finite-order elements, then so does .

Theorem 2 illustrates a common paradigm in geometric group theory. If some property holds for groups acting geometrically on a certain type of metric space, then the same should be true for a group acting on this same type of metric space so long as certain subgroups (e.g., point stabilizers) have property . In other words, we should be able to promote a property from a collection of subgroups to the whole group if we can find the appropriate space where these subgroups are the point stabilizers.

This idea suggests a useful strategy. Suppose you have some family of groups that fit into a hierarchy: , , , …where the groups in act geometrically on a certain type of metric space and the groups in also act on this same type of metric space with point stabilizers belonging to . If we can verify the above paradigm for this type of metric space, this gives an inductive way to show that all the groups in this family have some particular property or structure. In the next section, we will mention an instance where this strategy has been particularly fruitful: the mapping class group of an orientable surface.

Actions on –hyperbolic spaces

Actions on trees are nice to work with, but they form a fairly restrictive class of groups. There are many interesting and natural groups in which every action on a tree has a global fixed point. For example, this is true for when . Surely, not much can be gained in general from actions with a global fixed point.

Gromov’s influential essay Gro87 introduced a notion of negative curvature that unifies essential properties of the hyperbolic plane, trees, and small cancellation groups—a thoroughly studied class of groups explored in the latter half of the 20th century in which geometric notions and techniques were starting to gain traction. The idea behind Gromov’s definition of a –hyperbolic space is to take one of the useful consequences of negative curvature from the hyperbolic plane and use it as a definition for a metric space. Gromov gave such a definition solely using a metric on an arbitrary set , but the most common formulation used—and one that applies to almost all the spaces one comes across in geometric group theory—requires a geodesic metric space, which is defined as follows. A geodesic in a metric space is a function where is a connected subset of such that for all . A geodesic metric space is a metric space such that for all , there is a geodesic with and . A connected graph, in particular the Cayley graph of a finitely generated group, is a geodesic metric space.

There are many equivalent formulations of a –hyperbolic metric space using geodesic triangles, divergence of geodesics, or nearest point projections to geodesics. We will state the most common formulation using geodesic triangles, which Gromov attributed to Rips. In the statement, represents the image of any geodesic in from to .

Definition 3.

Let be a geodesic metric space. A geodesic triangle is –thin if the –neighborhood of any two of the edges contains the third. That is, for all there is an such that . A –hyperbolic space is a geodesic metric space where every geodesic triangle is –thin.

The key point in the definition is that the same works for every geodesic triangle, no matter how long the sides are. See Figure 9.

Figure 9.

A –thin triangle.

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=0.8,yscale=-1] \renewcommand{\c}{black!50!white} \draw[\c,opacity=0.4,line width=2.1cm,line cap=round] (0,0) .. controls (4.4,-0.5) and (4.6,-1) .. (5,-3); \draw[\c,opacity=0.4,line width=2.1cm,line cap=round] (5,-3) .. controls (5.1,-0.3) and (5.8,-0.1) .. (8,1); \begin{scope} \tikzstyle{every path}=[thick,blue] \tikzstyle{every node}=[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] \draw(0,0) .. controls (4.4,-0.5) and (4.6,-1) .. (5,-3); \draw(5,-3) .. controls (5.1,-0.3) and (5.8,-0.1) .. (8,1); \draw(0,0) .. controls (4,-0.2) and (5.5,0) .. (8,1); \node at (0,0) (a) {}; \node at (5,-3) (b) {}; \node at (8,1) (c) {}; \end{scope} \draw[thin,<->] (1.31,-1.46) -- (1.5,-0.24); \node at (1.6,-0.9) {$\delta$}; \node[above] at (a) {$a$}; \node[left] at (b) {$b$}; \node[right] at (c) {$c$}; \end{tikzpicture}

Here are some examples of –hyperbolic spaces.

(1)

A tree is –hyperbolic since every geodesic triangle is a tripod and so any side is contained in the union of the other two. See Figure 10. We think of thinner triangles indicating the space being more negatively curved—this is true for scalar curvature in Riemannian geometry—and so in this sense, trees are negatively curved in the extreme.

Figure 10.

A typical geodesic triangle in a tree.

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=0.7,yscale=-1] \tikzstyle{every path}=[thick] \begin{scope} \tikzstyle{every node}=[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] \draw[blue] (0,0) node(a) {} -- (5,0) -- (5,2) node(c) {}; \draw[blue] (2,0) -- (2,-2) node(b) {}; \end{scope} \draw[green] (0.1,0.1) -- (4.9,0.1) -- (4.9,1.9); \draw[green] (0.1,-0.1) -- (1.9,-0.1) -- (1.9,-1.9); \draw[green] (2.1,-1.9) -- (2.1,-0.1) -- (5.1,-0.1) -- (5.1,1.9); \node[above] at (a) {$a$}; \node[left] at (b) {$b$}; \node[right] at (c) {$c$}; \end{tikzpicture}
(2)

The hyperbolic plane is –hyperbolic. As every geodesic triangle is contained in an ideal triangle, we only have to compute for an ideal triangle, which is a fun exercise. See Figure 11.

Figure 11.

Ideal triangles in the the hyperbolic plane are –thin.

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture}[scale=0.7] \tikzstyle{every path}=[thick] \draw[green] (0.5,1.5) arc [radius=2.121320344, start angle = 45, end angle = 90]; \node at (0,2.2) {$\delta$}; \draw[black] (-4,0) -- (4,0); \draw[blue] (-1,0) -- (-1,4) (2,0) -- (2,4); \draw[blue] (2,0) arc [radius=1.5, start angle = 0, end angle = 180]; \draw[blue] (2,0) arc [radius=1.5, start angle = 0, end angle = 180]; \begin{scope} \tikzstyle{every node}=[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] \node at (0.5,1.5) (a) {}; \node at (-1,2.121320344) (b) {}; \end{scope} \node[below] at (a) {$\frac{1 + i}{2}$}; \node[left] at (b) {$\frac{i}{\sqrt{2}}$}; \node[below] at (-1,0) {0}; \node[below] at (2,0) {1}; \end{tikzpicture}
(3)

The Farey graph is –hyperbolic. Indeed, suppose that lies on a geodesic between the vertices and . Let and be the vertices adjacent to along this geodesic and assume that . As we are dealing with a geodesic, we must have since otherwise there is an edge between and . Hence there is some vertex adjacent to such that . As the removal of the vertices and and also the edge connecting these two vertices disconnects the Farey graph, we see that any path from to must pass through either or .

For contrast, with the Euclidean metric is not –hyperbolic for any . Indeed, the geodesic triangle with vertices , and is –thin only for . To see this, consider the point .

The typical questions one may try to answer using actions on –hyperbolic spaces often fit into the following categories.

(1)

Algorithmic: When do two words in a generating set represent the same element or conjugate elements?

(2)

Local-to-global: Are paths in the Cayley graph that are locally geodesics globally geodesics as well?

(3)

Rigidity: If two groups have geometrically similar Cayley graphs, are the groups algebraically similar? Can we characterize homomorphisms to and from the group?

We will discuss in turn geometric actions and other types of actions on –hyperbolic spaces.

Geometric actions on –hyperbolic spaces

A metric space is proper if closed balls are compact. A group is hyperbolic if it acts geometrically on a proper –hyperbolic space⁠Footnote2. Free groups and fundamental groups of closed hyperbolic manifolds are hyperbolic groups. It is fair to ask how common hyperbolic groups are given that we started this section noticing that useful tree actions do not always exist. Gromov introduced a model of a “random finitely presented group” that includes a parameter called the “density” that controls the number of relators in terms of the number of generators Gro93, Chapter 9. When , Gromov showed that a random group is infinite and hyperbolic. (For those curious, when a random group has at most two elements.) Thus, it is fair to say that hyperbolic groups are quite ubiquitous.

2

In the literature, these groups are sometimes referred to as negatively curved, word hyperbolic, or Gromov hyperbolic.

An equivalent definition of a hyperbolic group is that is finitely generated and the Cayley graph is –hyperbolic for some finite generating set . Moreover, “some” in the previous sentence can be replaced with “every.” Hyperbolic groups satisfy a long list of useful properties and besides Gromov’s original essay, there are many comprehensive works focused on these groups. See for instance the notes edited by Short ABC91, the chapters by Bridson and Haefliger BH99, Chapters III.H and III., and the references within these works.

As hyperbolic groups are defined by a geometric condition (in several equivalent ways), from their inception researchers have wondered if there is an algebraic characterization. It is not too difficult to find algebraic obstructions. One of the first usually encountered involves the centralizer of an infinite-order element. If is a hyperbolic group and has infinite order, then , the cyclic subgroup generated by , has finite index in , the centralizer of . Recall, the centralizer of is the subgroup of consisting of elements with . The idea behind this fact nicely illustrates a typical geometric argument using the –thin triangle condition.

Suppose that and consider the four vertices , , , and in the Cayley graph for a large . The fact that implies that these four points lie on a rectangle. The horizontal sides are formed by a geodesic and its translate by , the geodesic . To get the vertical sides, use a geodesic and its translate by . The translate by gives a geodesic from to , but this latter point is exactly by the commutivity assumption. See Figure 12.

Figure 12.

A commuting rectangle in .

\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \begin{tikzpicture} \tikzstyle{every node}=[black,inner sep=0pt,outer sep=0pt] \tikzstyle{every path}=[thick] \node[label={[shift={(-0.4,-0.2)}]$1_G$}] at (0,0) (1) {}; \node[label={[shift={(0.4,-0.2)}]$g^k$}] at (6,0) (g) {}; \node[label={[shift={(-0.3,0)}]$h$}] at (0,3) (h) {}; \node[label={[shift={(0.4,0)}]$hg^k$}] at (6,3) (hg) {}; \node[label={[shift={(0,-0.5)}]$g^m$}] at (3.6,0) (gm) {}; \node[label={[shift={(0,0.2)}]$hg^n$}] at (2.4,3) (gn) {}; \draw[blue] (1) .. controls (4,0.95) .. (g) node[pos=0.35,label={[shift={(-0.1,-0.4)}]$x$}] (x) {}; \draw[blue] (h) .. controls (2,2.05) .. (hg) node[pos=0.65,label={[shift={(0,0.18)}]$y$}] (y) {}; \draw[blue] (1) .. controls (0.75,1.5) .. (h); \draw[blue] (g) .. controls (5.25,1.5) .. (hg); \draw[blue] (1) .. controls (2,1.75) and (4,1.25) .. (hg) node[pos=0.45] (z) {}; \draw[thin] (gm) -- (x) -- (z) -- (y) -- (gn); \begin{scope} \tikzstyle{every node}=[fill,red,circle,minimum size=0.15cm,inner sep=0pt,outer sep=0pt] \node at (1) {}; \node at (g) {}; \node at (h) {}; \node at (hg) {}; \node at (x) {}; \node at (y) {}; \node at (gm) {}; \node at (gn) {}; \end{scope} \node at (3.2,1.1) {\footnotesize$\leq\delta$}; \node at (3.2,1.9) {\footnotesize$\leq\delta$}; \node at (3.7,0.4) {\footnotesize$\leq L$}; \node at (2.35,2.6) {\footnotesize$\leq L$}; \end{tikzpicture}

Now an important property of hyperbolic groups is that infinite cyclic subgroups are undistorted, that is, the distance from to is approximately . This fact, plus a stability result about paths that coarsely resemble geodesics, imply that there is a constant so that any point on the geodesic is within of for some , …, . Likewise, any point on the geodesic is within of for some , …, . Now let be the midpoint of the geodesic . By considering the two geodesic triangles and pictured in Figure 12, we see that is within of a point that lies on one of other three sides of the rectangle. By choosing large enough, we can ensure that lies on the geodesic as shown in Figure 12. We have and for some which gives

Hence the coset has an element whose distance from is at most . As there are only finitely many such elements and as distinct cosets are always disjoint, there are only finitely many cosets.

As a consequence, no subgroup of a hyperbolic group can be isomorphic to . In several classes of geometrically defined groups, this turns out to be the only obstruction to hyperbolicity. For instance, this is true for the class of fundamental groups of closed 3–manifolds. In general, there are other algebraic obstructions to consider. Hyperbolic groups cannot contain a subgroup isomorphic to one of the Baumslag–Solitar groups: