1 Foundations of knot theory

Knot theory studies closed loops embedded in three-dimensional space and classifies them up to deformation. The central questions ask when two drawings represent the same knot, how knots can be distinguished, and which structures arise from combining knots with links and tangles. Although the subject is part of topology, it draws on ideas from algebra, geometry, and combinatorics.

1.1 Mathematical definition of a knot

A mathematical knot is a smooth or piecewise smooth embedding of a circle into three-dimensional space, often taken in ordinary Euclidean space or in the 3-sphere. Because the object has no endpoints, it differs from an everyday knotted string. Two knots are regarded as equivalent if one can be transformed into the other by a continuous deformation that preserves the embedding.

1.2 Ambient isotopy

The standard notion of equivalence in knot theory is ambient isotopy, meaning a deformation of the entire surrounding space that carries one knot to another. This avoids cutting the knot or forcing strands through each other. Ambient isotopy preserves the topological type of the embedded loop and serves as the basis for most classification problems in the field.

1.3 Knot diagrams

Knot diagrams are planar projections of knots with extra information added at each crossing to indicate the spatial arrangement. They provide a practical way to study knots using two-dimensional pictures. Many theoretical results begin with a diagram and then analyze how it changes under allowable moves.

1.3.1 Crossings and over- and under-strands

At a crossing, one strand passes over another in the projection. This over- and under-strand information is essential, since different spatial knots can share the same underlying planar curve. The collection of crossings in a diagram encodes much of the knot's structure and makes systematic comparison possible.

1.3.2 Reidemeister moves

Reidemeister moves are local changes to a knot diagram that do not alter the knot type. There are three basic moves: twisting or untwisting a loop, moving one strand over another, and sliding one crossing past another. Two diagrams represent the same knot if and only if they are related by a sequence of these moves, together with planar isotopy.

Links generalize knots by allowing several closed loops in the same space, while tangles consist of strands with endpoints arranged in a fixed boundary region. Both notions broaden the range of objects studied in low-dimensional topology. They also provide useful building blocks for knot constructions and invariants.

A link is a collection of knots that may be interlaced or separated, with the components considered jointly. The simplest nontrivial examples include linked circles with nonzero linking number. Link theory studies how the components interact and how their pairwise and higher-order relationships can be measured.

1.4.2 Rational tangles

Rational tangles are tangles that can be described using continued fractions and simple twisting operations. They form a particularly tractable class and connect knot theory with number-theoretic ideas. Rational tangles also appear in the study of 2-bridge knots and related link families.

2 Basic classes of knots

Certain families of knots recur frequently because they have clear geometric descriptions and often admit explicit calculations. These classes help organize knot tables and serve as test cases for invariants. They also illustrate the variety of ways a knot can be embedded in space.

2.1 The unknot

The unknot is the simplest knot, equivalent to a round circle. It has no genuine entanglement and serves as the baseline against which all other knots are compared. Deciding whether a complicated diagram represents the unknot is a central problem in the subject.

2.2 Torus knots

Torus knots lie on the surface of a torus and wrap around it in two directions. They are usually described by two integers that record the longitudinal and meridional windings. These knots are highly structured and are among the best understood families in knot theory.

2.3 Twist knots

Twist knots are formed by introducing a chain of twists into a simple clasp-like configuration. They provide an accessible infinite family that includes several familiar examples. Their diagrams and invariants can often be described explicitly, making them useful in examples and computations.

2.4 Satellite knots

Satellite knots are constructed by placing one knot pattern inside a tubular neighborhood of another knot. The outer knot, often called the companion, influences the overall structure. This category includes many complicated knots that cannot be reduced to simpler torus or twist forms.

2.5 Composite and prime knots

A composite knot can be decomposed as a connected sum of two nontrivial knots, while a prime knot cannot. This distinction parallels prime factorization in arithmetic, though the decomposition theory is more subtle. Prime knots play a foundational role because many classification results are organized around them.

3 Knot invariants

Knot invariants are quantities or algebraic objects that remain unchanged under ambient isotopy. They are the main tools used to distinguish knots and to prove that two knots are not equivalent. Different invariants detect different features, so the field often combines several of them.

3.1 Classical invariants

Classical invariants arise from early algebraic and geometric approaches to knots. They include fundamental groups, surface data, and linking phenomena. Many later invariants build on these older constructions.

3.1.1 Knot group

The knot group is the fundamental group of the knot complement. It encodes how loops can wind around the knot and often contains rich information about the embedding. Although not a complete invariant, it distinguishes many knots and links.

3.1.2 Seifert surface genus

A Seifert surface is an orientable surface whose boundary is the knot, and its genus measures the surface's complexity. The minimal genus among all such surfaces is an important invariant. It is closely related to knot complexity and to questions about surfaces in three-dimensional space.

3.1.3 Linking number

The linking number measures the degree of entanglement between two components of a link. It can be computed from a diagram by counting signed crossings between components. As one of the oldest invariants, it gives a first approximation to how link components interact.

3.2 Polynomial invariants

Polynomial invariants associate algebraic expressions to knots and links. They are often easier to compute than many geometric invariants and have strong discriminatory power. Their discovery transformed knot theory into a more computationally accessible subject.

3.2.1 Alexander polynomial

The Alexander polynomial is one of the earliest polynomial invariants of knots. It is derived from algebraic data associated with the knot complement. While it does not completely classify knots, it remains an important and influential invariant.

3.2.2 Jones polynomial

The Jones polynomial assigns a Laurent polynomial to a knot or link and is sensitive to many distinctions missed by older invariants. Its discovery revealed deep connections between knot theory and statistical mechanics. It became a central object in modern low-dimensional topology.

3.2.3 HOMFLY-PT polynomial

The HOMFLY-PT polynomial generalizes both the Alexander and Jones polynomials. It provides a two-variable invariant that captures a broader range of knot and link information. This flexibility makes it valuable in both theoretical and computational work.

3.3 Homological invariants

Homological invariants refine polynomial invariants by replacing numerical outputs with graded algebraic structures. They often contain more information and can lead to stronger classification results. Their construction typically involves chain complexes built from knot diagrams or complements.

3.3.1 Khovanov homology

Khovanov homology categorifies the Jones polynomial by replacing it with a graded homology theory. The polynomial emerges as an Euler characteristic of the homology groups. This refinement has led to new results about knot detection and slice properties.

3.3.2 Knot Floer homology

Knot Floer homology is a powerful invariant derived from Floer theory. It relates knot theory to symplectic geometry and Heegaard splittings. The theory detects several important geometric features, including genus and fibering behavior in many cases.

3.4 Finite-type invariants

Finite-type invariants form a hierarchy based on how they respond to singularities or crossing changes. They are organized by degree and are closely tied to combinatorial structures. This framework has become a major area in the theory of knot invariants.

3.4.1 Vassiliev invariants

Vassiliev invariants are finite-type invariants defined through extensions to singular knots. They are central to the modern study of perturbative knot theory. Many other invariants can be interpreted or approximated within this framework.

4 Knot operations and constructions

Knot operations produce new knots from old ones and help reveal structural relationships. Some operations preserve particular properties, while others transform knots in ways that are useful for classification. These constructions also connect knot theory with 3-manifold topology.

4.1 Connected sum

The connected sum combines two knots by cutting each and joining the open ends in a compatible way. The operation is associative up to equivalence and underlies the notion of composite knots. It is one of the main ways to build more complicated knots from simpler pieces.

4.2 Mirroring and reversal

A mirror image of a knot is obtained by switching all crossings in a diagram, while reversal changes the orientation of the knot. These operations may or may not produce equivalent knots, depending on the knot type. They are important in the study of chirality and symmetry.

4.3 Cabling

Cabling replaces a knot by a new pattern winding around a tubular neighborhood of the original knot. The result is a family of related knots or links with controlled topology. Cabling often increases complexity and is used in the construction of satellite knots.

4.4 Dehn surgery

Dehn surgery modifies a three-dimensional manifold by removing a tubular neighborhood of a knot and gluing it back differently. In knot theory, it provides a bridge between knots and the manifolds obtained from them. This operation is central to much of modern three-dimensional topology.

4.5 Mutation

Mutation is a transformation that cuts a knot diagram along a sphere or tangle region and reglues it after a symmetry. It can produce knots that are difficult to distinguish by many classical invariants. Mutation illustrates the limits of certain classification methods.

5 Classification and tabulation

Classification seeks to organize knots into families and distinguish them systematically. Tabulation records knots up to a chosen complexity, usually measured by crossing number. These efforts provide reference data for both theoretical work and computation.

5.1 Knot tables

Knot tables list knots in order of increasing crossing number and often include standard diagrams, names, and invariant data. They are widely used as a reference for researchers and students. Tables help identify knots encountered in calculations or applications.

5.2 Crossing number

The crossing number is the smallest number of crossings in any diagram of a knot. It gives a basic measure of complexity, though it is often difficult to determine exactly. Many classification results are organized by crossing number.

5.3 Alternating knots

Alternating knots admit diagrams in which over- and under-crossings alternate as one travels along the knot. These knots are often easier to analyze and display strong structural properties. Many classical classification results were first established for alternating knots.

5.4 Chirality and amphicheirality

A chiral knot is not equivalent to its mirror image, while an amphicheiral knot is equivalent to its mirror image. These properties describe a knot's symmetry under reflection. Detecting chirality is an important test for many invariants.

5.5 Slice and ribbon knots

Slice knots bound smoothly embedded disks in four-dimensional space, and ribbon knots form a notable subclass with especially tractable embeddings. These notions connect knot theory to four-dimensional topology. They also play a major role in understanding how knots can bound surfaces in higher dimensions.

6 Connections with other areas of mathematics

Knot theory interacts deeply with several other branches of mathematics. Its problems often translate into questions about manifolds, groups, and quantum structures. These connections have expanded the field far beyond its original setting.

6.1 Three-dimensional manifolds

The complement of a knot is a three-dimensional manifold, so knot theory naturally studies the geometry and topology of manifolds. Many properties of knots are best understood through the spaces that surround them. This perspective has led to powerful classification techniques.

6.1.1 Lens spaces

Lens spaces are simple families of three-manifolds that often arise from surgery on knots or links. They provide useful examples in surgery theory and manifold classification. Their appearance in knot theory reflects the strong link between embedded loops and ambient spaces.

6.1.2 Branched covers

A branched cover is a covering space that is allowed to branch along a knot or link. Such constructions convert knot problems into questions about related manifolds. Branched covers are especially useful for extracting algebraic and geometric information.

6.2 Braid groups

Braid groups describe collections of strands interwoven in time and offer a dynamic viewpoint on knots. Many knots can be represented as closures of braids. This relationship makes braid theory an essential companion to knot theory.

6.2.1 Alexander's theorem

Alexander's theorem states that every knot or link can be represented as the closure of a braid. This result provides a uniform algebraic encoding of knots. It is a cornerstone of the braid-knot correspondence.

6.2.2 Markov's theorem

Markov's theorem characterizes when two braids have equivalent closures. It describes the moves on braids that preserve the knot or link type. Together with Alexander's theorem, it gives a complete bridge between braids and knots.

6.3 Quantum topology

Quantum topology studies knot theory using ideas inspired by quantum physics and representation theory. It led to new invariants and new conceptual links between topology and physics. The field has had a major influence on modern low-dimensional topology.

6.3.1 Quantum invariants

Quantum invariants arise from quantum groups, state-sum models, and related algebraic structures. They generalize classical polynomial invariants and often come with deep categorical interpretations. These invariants have proved effective in distinguishing knots and links.

6.3.2 Topological quantum field theory

Topological quantum field theory assigns algebraic data to manifolds and cobordisms in a way that is insensitive to smooth deformation. In knot theory, it provides a conceptual framework for constructing and understanding quantum invariants. The theory has become a central organizing principle in the subject.

7 Applications

Knot theory appears in several scientific disciplines where entanglement, looping, or spatial constraints matter. Its applications often use mathematical knots as models rather than literal ropes. The field has become especially relevant in molecular and physical systems.

7.1 Molecular biology

In biology, knot theory helps describe the geometry of long molecules such as DNA. These molecules can become entangled during replication, recombination, or packaging. Topological methods offer a language for analyzing such configurations.

7.1.1 DNA topology

DNA topology studies the looping, twisting, and supercoiling of DNA molecules. Knot-theoretic concepts help classify the spatial forms that DNA can adopt. This is useful in understanding how biological processes affect molecular structure.

7.1.2 Enzymatic knotting and unknotting

Certain enzymes can change the topological state of DNA by creating or removing crossings. These actions may produce knotting or unknotting as part of normal cellular activity. Knot theory provides models for describing the resulting transformations.

7.2 Chemistry

Chemistry uses knot theory to analyze the structure of complex molecules and molecular assemblies. Some compounds are intentionally designed to have knotted or linked forms. These structures can influence stability, reactivity, and self-assembly.

7.2.1 Molecular knots and catenanes

Molecular knots are single molecules with knotted backbones, while catenanes consist of interlinked molecular rings. Both are examples of topological architecture in chemistry. Their study combines synthetic methods with topological classification.

7.3 Physics

Physics uses knot-theoretic ideas in contexts ranging from statistical models to field configurations. Knot invariants can emerge from physical observables, and geometric entanglement can influence measurable behavior. This has made knots relevant in several theoretical settings.

7.3.1 Statistical mechanics

In statistical mechanics, knot theory appears in models of polymers, lattice systems, and partition functions. The Jones polynomial, in particular, arose from work connected to statistical physics. Such links have encouraged the use of knot invariants in physical modeling.

7.3.2 Knot-like defects in field theory

Some field theories admit defect structures or solitonic configurations that resemble knots or links. These objects are studied using topological methods because their stability can depend on global rather than local features. Knot theory supplies useful language for describing their classification.

8 Historical development

The history of knot theory reflects its evolution from a physical curiosity into a sophisticated mathematical discipline. Early work focused on listing knots and interpreting them as models of atoms or forces. Later developments placed the subject firmly within topology and broadened its scope.

8.1 Origins in 19th-century physics

Knot theory began partly in nineteenth-century attempts to model matter and forces through vortices or atomic structures. Although these physical theories were later abandoned, they motivated systematic study of knots. The initial interest left a lasting mathematical legacy.

8.2 Early classification efforts

Early researchers compiled tables of knots and sought methods to distinguish them. These efforts emphasized diagrams, crossing counts, and primitive invariants. The resulting catalogues laid the groundwork for later classification by more sophisticated tools.

8.3 Rise of modern topology

As topology matured in the twentieth century, knots were reinterpreted as embeddings and studied through complement spaces. This shift introduced algebraic topology, manifold theory, and geometric methods into the subject. Knot theory then became a major part of low-dimensional topology.

8.4 Discovery of polynomial invariants

The introduction of polynomial invariants, especially the Alexander and Jones polynomials, revolutionized the field. These tools made it possible to distinguish many knots efficiently and revealed unexpected links to other areas of mathematics. Their success stimulated the development of homological and quantum refinements.

8.5 Recent advances

Recent work has expanded knot theory through categorification, Floer-theoretic methods, and interactions with physics. Computational approaches have also improved the ability to tabulate knots and test conjectures. The field continues to grow through connections with geometry, algebra, and applied sciences.