Artin group

An Artin group is a group defined by a presentation that generalizes the standard presentation of the braid group. Given a Coxeter matrix \(m = (m_{ij})\), where \(m_{ii} = 1\) and \(m_{ij} = m_{ji} \geq 2\) for \(i \neq j\), the corresponding Artin group has generators \(s_1, \dots, s_n\) and relations: \[\underbrace{s_i s_j s_i \cdots}_{m_{ij} \text{ factors}} = \underbrace{s_j s_i s_j \cdots}_{m_{ij} \text{ factors}}\] for all \(i \neq j\). When \(m_{ij} = 2\), the generators commute. When \(m_{ij} = 3\), the generators satisfy the braid relation. The Artin group is obtained from the Coxeter group by simply removing the relation \(s_i^2 = 1\); in this sense, Artin groups are to Coxeter groups as braid groups are to symmetric groups.

The classification of Artin groups parallels the classification of Dynkin diagrams and finite Coxeter groups. The finite-type Artin groups — those whose associated Coxeter group is finite — include the braid groups (type A_n), the hyperoctahedral Artin groups (type B_n), and the exceptional Artin groups of types E_6, E_7, E_8, F_4, and G_2. These are the most studied and best understood.

From Braid Groups to Artin Groups

The braid group B_n is the Artin group of type A_{n-1}. This is not merely a special case; it is the prototype. Every theorem about braid groups raises a natural question about general Artin groups. The word problem for braid groups was solved by Garside in 1969. The word problem for finite-type Artin groups was solved by Brieskorn and Saito in 1972, and for all Artin groups by Charney in 2022. The conjugacy problem, solved for braid groups by Garside, remains open for arbitrary Artin groups — a striking reminder that generalization does not always preserve tractability.

The topological interpretation of braid groups as fundamental groups of configuration spaces also generalizes. The Salvetti complex provides a classifying space for Artin groups: a CW complex whose fundamental group is the Artin group and whose universal cover is contractible (for finite-type Artin groups). The Salvetti complex is constructed from the associated hyperplane arrangement, and its geometry encodes the algebraic structure of the Artin group in ways that are still being explored.

Artin Groups and Geometric Group Theory

Artin groups occupy a central position in geometric group theory — the study of groups through their actions on geometric and topological spaces. They are not hyperbolic in the sense of Gromov (except in trivial cases), but they exhibit many hyperbolic-like properties: automaticity, biautomaticity, and the existence of combing geodesics. The Garside structure on finite-type Artin groups provides a normal form that is a geodesic in the word metric, making these groups algorithmically tractable in ways that general Artin groups are not.

The geometry of Artin groups is intimately connected to the geometry of the associated Coxeter group. The Coxeter group acts on a convex polyhedral cone (the Tits cone), and the Artin group acts on a complexified version of this space. The quotient of this action by the Artin group is the classifying space. This connection between discrete group actions and continuous geometry is one of the most productive themes in modern geometric group theory.

The Open Frontier

Much of the theory of Artin groups remains conjectural. The \(K(\pi, 1)\) conjecture — that the Salvetti complex is a classifying space for all Artin groups, not just finite-type ones — has been proven for many cases but remains open in general. The center of an Artin group, the structure of its outer automorphism group, and its cohomology are all active areas of research.

What makes Artin groups remarkable is not just their mathematical structure but their ubiquity. They appear in the study of singularities, in configuration spaces, in the topology of hyperplane complements, and in the representation theory of quantum groups. They are the algebraic shadow of a geometric phenomenon — the arrangement of reflecting hyperplanes — and they carry that geometry into domains where the original reflection symmetry is no longer visible.

The Artin group is what remains when you remove the mirrors from a reflection group but keep the choreography. It is the dance without the dancers, the pattern without the pattern-maker. That such an abstraction should retain so much structure — that the braid relations alone, without the involution, should still encode the geometry of the Coxeter arrangement — is evidence that symmetry is not a property of objects but a property of relations. The Artin group does not describe what is symmetric. It describes what symmetry leaves behind when it departs.