Geometric Group Theory

Geometric group theory studies groups by treating them as geometric objects.

13 frameworks, from 1910 to today · Geometry · Mathematics

still a working view ended · each leads to its entry below

Overview

Geometric group theory studies groups by treating them as geometric objects1. The central move is to attach a space to a group—a Cayley graph, a manifold, a complex—and then read off algebraic properties from the geometry of that space, or conversely use algebra to understand the space1. What is at stake is a translation principle: questions about infinite groups that look purely algebraic (Is this group finitely presented? Does it contain a free subgroup? Is it amenable?) often become transparent when the group acts on a space with the right kind of curvature or combinatorial structure. The field's history is the story of building that dictionary, and of discovering that the dictionary is richest when the spaces involved have negative or non-positive curvature.

From Cayley Graphs to Combinatorial Group Theory

The founding object is the Cayley graph, introduced by Arthur Cayley in 1878 in a paper on group diagrams. Given a group generated by a set \(S\), the Cayley graph has one vertex per group element and an edge for each multiplication by a generator2. The graph encodes the group's multiplication table geometrically, and it was used early on to study finite groups and group presentations3.

Through the early twentieth century, the study of groups given by generators and relations—combinatorial group theory—developed largely as a branch of topology. Max Dehn, working on the fundamental groups of surfaces and knots around 1910–1912, posed three decision problems that shaped the field for decades: the word problem (given a word in the generators, does it represent the identity?), the conjugacy problem, and the isomorphism problem. Dehn also introduced Dehn's algorithm, a rewriting procedure that solves the word problem for surface groups by repeatedly cancelling more than half of a relator45. This was the seed of small cancellation theory, developed systematically in the 1960s by Roger Lyndon and others: if a presentation's relators overlap only in short pieces, then the group has strong algorithmic and algebraic properties, and its Cayley 2-complex is "negatively curved" in a combinatorial sense. Small cancellation theory produced the first broad families of groups with solvable word problems and, later, the first explicit finitely presented infinite groups with no finite quotients (the Novikov–Boone theorem, 1950s, showed the word problem is unsolvable in general, so positive results required hypotheses like small cancellation).

Hyperbolic Groups and the Geometric Turn

The decisive turning point came from Mikhail Gromov1. In a 1987 essay, "Hyperbolic Groups," Gromov isolated a class of finitely generated groups whose Cayley graphs look like hyperbolic space: they satisfy a linear isoperimetric inequality, or equivalently, geodesic triangles are uniformly thin165. These are now called Gromov-hyperbolic or word-hyperbolic groups16. The definition is coarse—it depends only on the large-scale geometry of the Cayley graph, not on the choice of generating set—and it covers free groups, surface groups, and fundamental groups of negatively curved manifolds, while excluding groups like \(\mathbb{Z}^2\).

Gromov's insight was that hyperbolicity is a robust geometric condition with powerful algebraic consequences. Hyperbolic groups are finitely presented, have solvable word and conjugacy problems, and have a well-behaved boundary at infinity—a compact space on which the group acts by homeomorphisms, generalizing the circle at infinity of the hyperbolic plane. The boundary became a central tool: it converts group-theoretic questions into topological dynamics, and it was used to prove rigidity results and to distinguish groups up to quasi-isometry.

The same period saw the emergence of quasi-isometry as the right notion of equivalence17. A quasi-isometry is a map between metric spaces that distorts distances by a bounded multiplicative and additive error, with a quasi-inverse7. In 1968, Albert Schwarz and, independently, John Milnor observed that any two Cayley graphs of the same finitely generated group are quasi-isometric, so quasi-isometry classes are group invariants. This raised the quasi-isometric rigidity question: when does the quasi-isometry class of a group determine the group up to commensurability? Milnor's 1968 paper "A note on curvature and fundamental group" showed that groups of polynomial growth are virtually nilpotent, and Gromov's 1981 polynomial growth theorem (proved using the asymptotic cone, a scaling limit of the Cayley graph) completed the picture: polynomial growth is exactly virtual nilpotency. The asymptotic cone, introduced by Gromov and developed by van den Dries and Wilkie in the 1980s, became a standard tool for studying large-scale geometry.

Rigidity, Outer Space, and Boundaries

The 1970s and 1980s produced the field's landmark rigidity theorems. In 1973, Gromov and, independently, Pierre Pansu and others studied lattices in semisimple Lie groups; Mostow's rigidity theorem (1968) had shown that lattices in rank-one groups are determined by their fundamental group, and Margulis's superrigidity (1970s) extended this. The quasi-isometric rigidity of lattices in higher-rank semisimple groups was established by Bruce Kleiner and Bernhard Leeb in the 1990s, building on work of Gromov and others. These results showed that the large-scale geometry of a group can pin down its algebraic structure completely.

A parallel development was the study of Outer Space, introduced by Marc Culler and Karen Vogtmann in 19868. Outer Space is a contractible space on which the outer automorphism group of a free group acts, analogous to the symmetric space on which a lattice acts8. It gave a geometric handle on \(\mathrm{Out}(Fn)\), leading to the proof that \(\mathrm{Out}(Fn)\) has finite virtual cohomological dimension and to a rich theory of automorphisms of free groups. The analogy with Teichmüller theory—where the mapping class group acts on Teichmüller space—became a guiding theme.

Boundaries of hyperbolic groups were developed extensively by Gromov, and later by Brian Bowditch, who proved in the 1990s that a hyperbolic group with a connected, locally connected boundary is a Kleinian group, and by others who studied the boundary as a topological invariant. The boundary is a quasi-isometry invariant, and its topology constrains the group: for instance, a hyperbolic group with a sphere boundary is a Poincaré duality group.

Automatic Groups and the Computational Turn

In the late 1980s, Jim Cannon, David Epstein, Derek Holt, Silvio Levy, Michael Paterson, and William Thurston developed automatic groups, a class defined by the existence of a finite-state automaton that recognizes which words represent the same group element. Automatic groups have solvable word problems and are closed under free products and direct products with free groups. The theory, published in the 1992 book Word Processing in Groups, gave a computational handle on hyperbolic groups (which are automatic) and on many other classes, and it connected group theory to formal language theory. The automatic structure is a coarse condition—it is preserved by quasi-isometry—and it became a standard tool for proving algorithmic properties.

CAT(0) Cube Complexes and Hierarchical Hyperbolicity

The 1990s and 2000s saw the rise of CAT(0) cube complexes, spaces built by gluing Euclidean cubes so that they satisfy a non-positive curvature condition in the sense of comparison geometry. These complexes had been studied in the 1980s by Gromov and others, but their power for group theory became clear with the work of Sageev, who introduced hyperplanes and used them to construct actions of groups on cube complexes. The theory of special cube complexes, developed by Frédéric Haglund and Daniel Wise in 2008, gave a combinatorial criterion for a cube complex to be a classifying space, and Wise's 2013 proof that one-relator groups with torsion are virtually special resolved a long-standing question. The cube complex machinery also produced the first examples of hyperbolic groups that are not residually finite, answering a question of Gromov.

A related development was the theory of hierarchically hyperbolic groups, introduced by Jason Behrstock, Mark Hagen, and Alessandro Sisto around 2017. These groups, which include mapping class groups and right-angled Artin groups, have a system of hyperbolic spaces—one for each "direction"—that captures their large-scale geometry. The theory unifies many earlier results and provides a framework for studying groups that are not themselves hyperbolic but have hyperbolic-like structure.

Coarse Geometry and the Novikov Conjecture

The coarse perspective—studying metric spaces up to large-scale equivalence—became a field in its own right through the work of Gromov, who introduced coarse geometry in the 1980s, and of John Roe, who developed coarse cohomology and the coarse Baum–Connes conjecture in the 1990s. The coarse Baum–Connes conjecture predicts that a certain assembly map from coarse K-homology to the K-theory of the Roe algebra is an isomorphism; it implies the Novikov conjecture on the homotopy invariance of higher signatures for finitely presented groups. The conjecture is known for groups with finite asymptotic dimension and for hyperbolic groups, but remains open in general. This line of work connected geometric group theory to operator algebras and index theory, and it produced the notion of coarse embedding and the study of which groups embed coarsely into Hilbert space1.

Random Walks and Growth

Random walks on groups, studied since the 1950s by Kesten and others, became a geometric tool in the 1980s and 1990s. The growth rate of a random walk—whether it is recurrent or transient—is a quasi-isometry invariant for nilpotent groups, and the Poisson boundary, a measure-theoretic boundary, generalizes the hyperbolic boundary. The growth of groups—the function counting elements of word length at most \(n\)—was studied by Milnor and Gromov, and the polynomial growth theorem (Gromov 1981) was a landmark. The growth rate is a quasi-isometry invariant, and the study of growth functions connects to algebraic geometry and to the theory of algebraic groups.

The Present

Today geometric group theory is a broad field with several active fronts1. The study of hyperbolic groups and their boundaries continues, with open questions about which hyperbolic groups are residually finite and which have the property that their boundary is a sphere. The theory of CAT(0) cube complexes and hierarchical hyperbolicity is being extended to new classes of groups, including Artin groups and mapping class groups. Coarse geometry and the Baum–Connes conjecture remain central, with recent progress on the coarse Baum–Connes conjecture for groups with finite asymptotic dimension and for certain expanders. The quasi-isometric classification of groups is still largely open: for instance, it is not known whether all finitely generated groups quasi-isometric to a free group are virtually free. The field's founding insight—that groups are geometric objects—continues to generate new dictionaries between algebra, topology, and geometry1.

References

  1. Geometric group theory (introduction, History, Modern themes and developments), Wikipedia
  2. Thin group (combinatorial group theory), Wikipedia
  3. Cayley graph, Wikipedia
  4. Small cancellation theory (History, Dehn's algorithm), Wikipedia
  5. Dehn function (History), Wikipedia
  6. Hyperbolic group, Wikipedia
  7. Quasi-isometry, Wikipedia
  8. Outer space (mathematics) (introduction, History), Wikipedia

The frameworks of Geometric Group Theory

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1910-1980

    Combinatorial Group Theory

    How can groups be studied through generators and relations?

    Fly to its tower in Nousfarer

  2. 1950-Present

    Random Walks and Growth of Groups

    How fast does a group grow under random walks?

    Fly to its tower in Nousfarer

  3. 1960-1990

    Small Cancellation Theory

    When do short overlaps in relators give good groups?

    Fly to its tower in Nousfarer

  4. 1968-Present

    Quasi-isometry

    When are two metric spaces the same at large scale?

    Fly to its tower in Nousfarer

  5. 1968-Present

    Rigidity Theory

    When does large-scale geometry determine algebraic structure?

    Fly to its tower in Nousfarer

  6. 1980-Present

    Coarse Geometry and the Novikov Conjecture

    What large-scale invariants detect the Novikov conjecture?

    Fly to its tower in Nousfarer

  7. 1984-Present

    Asymptotic Cones

    What do scaling limits of metric spaces reveal?

    Fly to its tower in Nousfarer

  8. 1986-Present

    Outer space

    What space does the outer automorphism group act on?

    Fly to its tower in Nousfarer

  9. 1987-Present

    Hyperbolic group

    Which groups have negatively curved Cayley graphs?

    Fly to its tower in Nousfarer

  10. 1988-Present

    Automatic group

    When can a finite automaton recognize group equality?

    Fly to its tower in Nousfarer

  11. 1990-Present

    Boundaries of Hyperbolic Groups

    What does the boundary at infinity of a group look like?

    Fly to its tower in Nousfarer

  12. 1990-Present

    CAT(0) Cube Complexes

    When do groups act on non-positively curved cube complexes?

    Fly to its tower in Nousfarer

  13. 2017-Present

    Hierarchical Hyperbolicity

    How can hyperbolic-like structure be organized hierarchically?

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Combinatorial Group TheoryRandom Walks and Growth of GroupsSmall Cancellation TheoryQuasi-isometryRigidity TheoryCoarse Geometry and the Novikov Conje…Asymptotic ConesOuter spaceHyperbolic groupAutomatic groupBoundaries of Hyperbolic GroupsCAT(0) Cube ComplexesHierarchical Hyperbolicity

Beyond the field

Where Geometric Group Theory's ideas came from in other fields, and what they led to there: the strongest 12 of 31 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Geometric Group Theory is a planet in Nousfarer

Its city has a tower for each of these 13 frameworks. Land, walk in, and climb one concept at a time.

Land on Geometric Group Theory

Related fields

Fields its ideas are linked to

More in Geometry

See where Geometric Group Theory sits on the Map of Human Knowledge