Loading field map
Low dimensional topology is the study of topological spaces of dimension three and lower, and of the ways higher-dimensional spaces can be built from or understood through these low-dimensional pieces. Its subject matter includes knots and links, surfaces, and three-dimensional manifolds, along with the maps and deformations between them. The field is distinguished from general topology by its focus on spaces that are close to human geometric intuition, yet whose classification and analysis require deep and often unexpected mathematics.
The primary objects are manifolds of dimension 1, 2, and 3. A manifold is a space that locally resembles ordinary Euclidean space of a fixed dimension. A one-dimensional manifold is a curve; the only connected compact examples are the circle and the interval. Two-dimensional manifolds, or surfaces, include the sphere, the torus, the projective plane, and their higher-genus generalizations. Three-dimensional manifolds are spaces that locally look like ordinary three-dimensional space; examples include the three-dimensional sphere, the three-dimensional torus, and spaces obtained by gluing together polyhedra along their faces.
Knots and links are central objects. A knot is an embedding of a circle into three-dimensional space, considered up to continuous deformation (ambient isotopy). A link is a disjoint union of several such circles. The fundamental question is classification: when are two knots or links equivalent? This question is approached through invariants—algebraic or combinatorial quantities that are unchanged under deformation—such as the knot group (the fundamental group of the complement), polynomial invariants, and more recent homological invariants.
Three-dimensional manifolds are the other central objects. They arise naturally in many contexts: as the boundaries of four-dimensional spaces, as configuration spaces of mechanical systems, and as the spatial universe in cosmological models. The classification of three-manifolds is a major achievement of the field, but it is not a simple list. Instead, it is a structural theorem that describes how every compact three-manifold can be decomposed into pieces of a few well-understood types.
The deepest questions in low dimensional topology concern classification and recognition. For surfaces, classification has been complete since the late nineteenth century: every compact surface is determined by its orientability, its number of boundary components, and its Euler characteristic (or genus). No such simple list exists for three-manifolds, and the recognition problem—deciding whether two given three-manifolds are the same—is algorithmically undecidable in full generality, though it is solvable for many important classes.
A second central question is the relationship between topology and geometry. The geometrization conjecture, proved by Grigori Perelman in the early 2000s, states that every compact three-manifold can be cut along spheres and tori into pieces that each admit one of eight geometric structures, with the hyperbolic structure being the most common and most flexible. This theorem provides a classification scheme: instead of listing manifolds, it describes the possible geometric models and how they are glued together.
A third question concerns the relationship between dimension three and dimension four. Many phenomena in four-dimensional topology are controlled by three-dimensional data. For example, the boundary of a four-dimensional manifold is a three-manifold, and understanding which three-manifolds arise as boundaries of which four-manifolds is a rich and active area. Conversely, some questions about three-manifolds are best answered by embedding them in four dimensions.
The field has roots in nineteenth-century work on the classification of surfaces and in the study of knots, which began as an attempt to understand the structure of the luminiferous ether. Early knot tables were compiled by Peter Guthrie Tait and others in the 1880s, motivated by physical speculation that atoms might be knotted vortex tubes. These tables were purely combinatorial, listing diagrams of knots and attempting to distinguish them by eye.
The modern subject took shape in the early twentieth century with the introduction of algebraic tools. Henri Poincaré's work on the fundamental group provided a powerful invariant for distinguishing spaces, and his conjecture that every simply connected closed three-manifold is the three-sphere became a central challenge. Max Dehn contributed key techniques, including the study of knot groups and the operation of Dehn surgery, which builds three-manifolds by cutting out a solid torus and gluing it back differently.
The mid-twentieth century saw the development of combinatorial and geometric methods. The theory of Heegaard splittings, which describes a three-manifold as two handlebodies glued along their boundary surfaces, provided a way to present and study three-manifolds. The work of Herbert Seifert on fibered spaces and of Kurt Reidemeister on knot invariants established the field's modern foundations. The introduction of the Alexander polynomial and later the Jones polynomial gave increasingly sensitive tools for distinguishing knots.
A major conceptual shift came with the work of William Thurston in the 1970s and 1980s. Thurston proposed that most three-manifolds admit a hyperbolic structure—a metric of constant negative curvature—and developed a program for proving this. His geometrization conjecture unified the classification problem with geometry, and his work on hyperbolic Dehn surgery showed that many three-manifolds are obtained by deforming hyperbolic structures. This geometric viewpoint transformed the field, making it possible to think about three-manifolds as geometric objects rather than purely combinatorial ones.
Perelman's proof of the geometrization conjecture, completed in 2003, was a landmark. It used the Ricci flow, a partial differential equation that smooths out the geometry of a manifold, to decompose any three-manifold into geometric pieces. The proof was long and technically demanding, and it resolved not only the geometrization conjecture but also the Poincaré conjecture, which had been open for a century.
The field is organized around several distinct but interacting approaches, each with its own questions, methods, and strengths.
The oldest approach works with explicit presentations: knot diagrams, triangulations, and handle decompositions. A knot is represented by a projection to the plane with over/under information at crossings; a three-manifold is represented by a triangulation into tetrahedra or by a Heegaard splitting. These presentations make the objects concrete and computable, and they are the basis for many invariants. The Reidemeister moves—local changes to a diagram that do not change the underlying knot—provide a complete set of moves relating any two diagrams of the same knot. This approach is algorithmic in spirit, and it underlies the software used to compute invariants and to search for examples.
The limits of this approach are practical: the number of diagrams or triangulations needed to represent a given object can grow rapidly, and many invariants are hard to compute from these presentations. Nevertheless, combinatorial methods remain essential for constructing examples and for proving theorems by explicit calculation.
The geometric approach, initiated by Thurston, studies three-manifolds through their possible metrics. The key idea is that a manifold may admit a complete, finite-volume metric of constant curvature: spherical, Euclidean, or hyperbolic. The geometrization theorem says that every compact three-manifold can be decomposed into pieces that admit one of eight such geometries, with hyperbolic geometry being the generic case. This viewpoint provides powerful tools: hyperbolic manifolds have a rich rigidity theory, and their volumes and other geometric invariants are topological invariants.
The geometric approach also includes the study of surfaces in three-manifolds. An incompressible surface is one that cannot be simplified by surgery; the existence of such surfaces is a key structural feature. The theory of normal surfaces, developed by Wolfgang Haken, provides an algorithmic way to find them in triangulated manifolds. This theory is both geometric and combinatorial, and it underlies many algorithms for recognizing three-manifolds.
Algebraic topology provides invariants that are often computable and powerful. The fundamental group of a knot complement is a complete invariant for prime knots, though it is not easy to use in practice. Homology groups, especially the first homology group, are easier to compute and are often used to distinguish manifolds. More sophisticated invariants, such as the Alexander polynomial and its generalizations, are derived from algebraic structures associated to the manifold.
The Jones polynomial, discovered in 1984, was a surprise: it was defined through a skein relation—a linear relation among the polynomials of knots that differ at a single crossing—and it turned out to be related to quantum groups and statistical mechanics. This discovery opened a new area, quantum topology, which produces invariants from algebraic structures such as Hopf algebras and categories. These invariants are often more sensitive than classical ones, and they have connections to physics, especially to topological quantum field theory.
A more recent approach, developed by Peter Ozsváth and Zoltán Szabó in the early 2000s, is Heegaard Floer homology. This theory assigns to a three-manifold a collection of homology groups, built from a Heegaard splitting and the symplectic geometry of a certain space of maps. The theory is powerful: it detects the genus of a knot, the unknot, and many other subtle properties. It is related to earlier invariants such as the Alexander polynomial, which appears as its Euler characteristic, and it has connections to four-dimensional topology through the theory of contact structures.
Heegaard Floer homology is technically demanding, but it has become a central tool. It is one of several related theories—including instanton Floer homology and monopole Floer homology—that are conjecturally equivalent and that provide deep invariants of three-manifolds. These theories are not merely computational tools; they encode geometric information about the manifold and its submanifolds.
These approaches are not rival schools but complementary tools that are often used together. A typical proof in the field might use a geometric decomposition to reduce a problem to a hyperbolic piece, then use an algebraic invariant to distinguish that piece from others, and finally use a combinatorial algorithm to verify the result. The geometrization theorem itself is proved using analytic methods (the Ricci flow) but is stated in geometric terms and has combinatorial consequences.
The relationships among approaches are often deep. For example, the Jones polynomial was originally defined combinatorially, but it was later understood through quantum groups and through the geometry of the complement. Heegaard Floer homology is defined through symplectic geometry but recovers classical algebraic invariants. The geometric and algebraic approaches are connected by the fact that many invariants are computable from a hyperbolic metric, and conversely, the existence of certain algebraic structures often forces geometric properties.
The field today is characterized by a rich interplay of techniques and by the availability of powerful computational tools. The geometrization theorem provides a structural backbone: many questions can be reduced to the hyperbolic case, where geometric tools are most effective. The theory of hyperbolic three-manifolds is mature, with deep results on volume, rigidity, and the structure of the deformation space.
Heegaard Floer homology and its relatives remain active areas of research, with ongoing work on their structure, their computational complexity, and their applications to knot theory and four-dimensional topology. The relationship between these theories and other invariants, such as the Jones polynomial and its categorifications, is an active topic.
Knot theory has expanded beyond the classical setting to include virtual knots, Legendrian and transverse knots in contact manifolds, and knots in arbitrary three-manifolds. These generalizations require new invariants and new geometric ideas. The study of surfaces in three-manifolds, including the theory of Heegaard splittings and the classification of incompressible surfaces, continues to develop, with connections to the study of mapping class groups and to the geometry of the curve complex.
The field also maintains strong connections to other areas of mathematics and to physics. Topological quantum field theory, which originated in the study of invariants of three-manifolds, has become a subject in its own right, with applications to quantum computation and to the study of topological phases of matter. The relationship between low dimensional topology and theoretical physics, especially through Chern–Simons theory and string theory, has been fruitful on both sides.
Low dimensional topology is unusual among mathematical fields in that its central objects are concrete and visual, yet its methods draw on nearly every branch of modern mathematics: algebraic topology, differential geometry, partial differential equations, symplectic geometry, and category theory. The field is not finished; many fundamental questions remain open, including the classification of knots, the complexity of distinguishing three-manifolds, and the full understanding of the relationship between the various Floer homologies. But the structural framework provided by geometrization and the algebraic invariants developed over the past several decades give the field a coherence that is rare in mathematics.