Knot Theory: When Loops Become Mathematics

Tie a piece of string into a knot, then fuse the two ends together. You now hold one of the central objects of an entire branch of mathematics — a closed loop tangled in space that cannot be untangled without cutting. The question mathematicians started asking in the 1860s — "are these two loops the same knot?" — turned out to be surprisingly, beautifully deep. And the answers they've found connect the shape of DNA to quantum computing, the structure of atoms to Fields Medals, and the way proteins fold to the limits of algorithms.

Welcome to knot theory, where loops become mathematics.

The Concept

A mathematical knot is a closed curve in three-dimensional space — a loop with no loose ends. The simplest possible knot is the unknot: just a plain circle, no crossings, fully unentangled. Every other knot is a closed loop that has been twisted through itself in some way that can't be undone.

The fundamental question of knot theory is deceptively simple: when are two knots the same? If you hold a tangled rubber band and wiggle it around — stretching, rotating, sliding strands over each other — are you changing the knot, or just the knot's appearance? Mathematicians say two knots are equivalent if you can continuously deform one into the other without ever cutting the loop. The technical term is ambient isotopy: you can move the knot through space any way you like, but you can't pass one piece of the string through another.

This sounds easy until you try it. Look at the trefoil knot — the simplest non-trivial knot, with just three crossings where the string passes over or under itself. It genuinely looks like something you might unknot with patience. But it can't be unknotted. No matter how long you fiddle with it, the trefoil stays a trefoil. Proving this requires mathematical machinery — you can't just "see" it's impossible.

That's the core challenge of knot theory: developing the tools to prove two knots are different, not just assume it.

Why It Matters

The Vortex Atom That Started Everything

Knot theory was born from a wrong idea about physics that turned out to be right about mathematics.

In the 1860s, the great Scottish physicist Lord Kelvin (William Thomson) was captivated by the work of Hermann Helmholtz on vortex rings in fluids — the mathematical cousins of smoke rings. Kelvin proposed a bold hypothesis: atoms are knotted vortices in the aether, the medium then believed to permeate all space. Different chemical elements corresponded to different knots. Hydrogen was one knot; carbon another.

His friend Peter Guthrie Tait took this seriously enough to begin cataloguing knots — the first systematic classification project in the history of the subject. By 1877, Tait had classified all knots with seven or fewer crossings. By 1885, he had published tables of knots up to ten crossings. These tables remained the standard reference for over a century.

Kelvin's vortex atom theory was abandoned when the aether turned out not to exist. But the mathematics Tait developed in its service was real and important, and the field it created — knot theory — became a cornerstone of modern topology.

DNA Can Tie Itself in Knots

The most immediately life-or-death application of knot theory isn't physics or computing — it's biology.

DNA in your cells isn't the tidy double-helix of textbook illustrations. Inside the nucleus, roughly two meters of DNA are compressed into a space about 10 micrometers across. Under that kind of topological stress, DNA routinely becomes knotted and tangled. During replication, when a cell copies its genome before dividing, the two newly-made DNA strands become interlinked like chain links — each wrapped around the other in a way that would prevent them from ever being separated.

This is where your cells rely on enzymes called topoisomerases. These molecular machines work by cutting DNA strands, passing other strands through the break, and then resealing the cut — essentially performing surgery on the knot. Type II topoisomerase cuts both strands of the double helix simultaneously, allows another segment to pass through, then reseals the break. The net effect is a change in the topology of the DNA — an untangling.

Knot theory gives biologists the language to describe and quantify what these enzymes are doing. The linking number (a basic knot invariant measuring how many times two loops wind around each other) is a precise way to measure the entanglement that topoisomerase is resolving.

The medical angle: many chemotherapy drugs, including doxorubicin, work by blocking topoisomerases. Cancer cells, which replicate constantly and therefore have high topoisomerase activity, are killed when this enzyme is disabled and their DNA becomes permanently entangled. Understanding DNA topology is directly understanding how those drugs work.

The Details

The Problem of Telling Knots Apart

The central technical challenge of knot theory is distinguishing knots. Given two diagrams — two 2D pictures of crossings — how do you determine whether they represent the same 3D knot or different ones?

In 1926, J.W. Alexander and G.B. Briggs (and independently, Kurt Reidemeister in 1927) solved a crucial sub-problem: they identified the three Reidemeister moves, a complete set of local manipulations that can be performed on a knot diagram without changing the underlying knot. Any two diagrams of the same knot can be transformed into each other through these three moves:

Move 1: Twist a strand, adding or removing a single crossing
Move 2: Slide one loop over another, adding or removing two crossings
Move 3: Move a strand over a crossing point

This gave mathematicians a concrete target: a knot invariant is any property that doesn't change under these three moves. If two knots have different invariants, they must be different knots. The hunt for good invariants became — and remains — one of the central projects of the field.

The Alexander Polynomial: A Fingerprint for Knots

In 1923, mathematician James W. Alexander II discovered the first knot polynomial — a polynomial in a variable t that could be computed from a knot diagram and that stayed constant across all deformations. For the trefoil, the Alexander polynomial is t - 1 + t⁻¹. For the unknot, it's just 1. These are different polynomials, so the trefoil and unknot are provably different knots.

The Alexander polynomial was a revelation — a genuine mathematical fingerprint for knots. It remained the only known knot polynomial for over 60 years.

But it had limits. Most famously, it couldn't distinguish a knot from its mirror image. The trefoil comes in two versions: a right-handed one and a left-handed one, mirror images of each other that cannot be deformed into each other. The Alexander polynomial gives the same value for both. This isn't a minor gap — chirality (handedness) is a fundamental topological feature, and the Alexander polynomial was blind to it.

The Jones Polynomial: When Algebra Met Topology by Accident

In 1984, New Zealand mathematician Vaughan Jones was studying an abstract area of algebra called von Neumann algebras — a tool from quantum mechanics — when he noticed an unexpected connection. The algebraic structures he was working with could be interpreted as knot invariants. He had accidentally discovered a new knot polynomial.

The Jones polynomial was dramatically more powerful than its 60-year-old predecessor. Crucially, it could distinguish left-handed from right-handed knots. The two forms of the trefoil have different Jones polynomials. For the first time, mathematicians had an invariant that could detect chirality.

Jones received the Fields Medal in 1990, the highest honor in mathematics. But the story doesn't stop there. Edward Witten — a physicist, not a mathematician — soon showed that the Jones polynomial could be derived from quantum field theory. Specifically, it emerged from a branch of physics called Chern-Simons theory. This was astonishing: a pure mathematical property of abstract loops had a natural home in theoretical physics. Witten received the Fields Medal in 1990 as well, becoming the first (and still only) physicist to receive it.

The connection between knot theory and quantum field theory opened a torrent of new mathematics. It hinted that knots weren't just a curiosity — they were somehow fundamental to the structure of space, matter, and quantum mechanics.

How Many Knots Are There?

The growth in knot complexity with crossing number is humbling.

At 3 crossings, there's exactly one prime knot: the trefoil. At 4 crossings, exactly one: the figure-eight knot. At 5 crossings, two prime knots. At 6 crossings, three. At 7 crossings, seven.

The count explodes from there. By 20 crossings, there are more than 2 billion distinct prime knots. The universe of knots is combinatorially vast — and yet mathematicians can, in principle, classify all of them.

Knots can also be chiral (distinct from their mirror image, like the trefoil) or amphichiral (identical to their mirror image, like the figure-eight knot). The figure-eight knot has a beautiful symmetry: reflecting it gives you the exact same knot back.

The Unknotting Problem

Here's a puzzle that sounds trivial and turns out to be computationally profound: given an arbitrary knot diagram, is it the unknot in disguise?

This is the unknotting problem, and it remains one of the great challenges at the boundary of topology and computer science. A diagram might have hundreds of crossings that could all potentially be simplified away — or might represent something genuinely knotted. It's been proven that the problem lies in the complexity class NP ∩ co-NP: there are short certificates that can verify "yes, this is the unknot" or "no, it isn't." But no one has found an efficient general algorithm for deciding which case you're in. It's a problem where the difficulty is hidden beneath apparent simplicity.

Quantum Computing: Knots as Logic Gates

The most futuristic application of knot theory is in quantum computing, and the connection is direct.

In certain quantum systems, particularly at the boundary between states of matter, there exist exotic quasiparticles called anyons. Unlike electrons (which are fermions) or photons (which are bosons), anyons follow different statistics — and when two anyons swap positions in 2D space, the quantum state of the system changes in a way that depends on the topology of their paths, not the exact paths themselves.

In topological quantum computing, quantum information is encoded not in the state of individual particles but in the braiding pattern of anyons — their interleaved trajectories through space and time. When anyons braid around each other, their world-lines (paths through spacetime) form literal braids. These braids are closed into knots when you "finish" the computation. And the probability of getting a particular outcome is related to evaluating the Jones polynomial of the resulting knot.

This is remarkable: knot theory is not an analogy or a loose inspiration here. It is the literal mathematical foundation of topological quantum computation. The advantage is error-resistance — small perturbations don't change the topology of the braid, so the quantum state is protected from noise in a way that standard quantum bits are not.

Molecular Knots: Chemistry Gets Topological

Chemists have gotten in on the act. In 2004, Fraser Stoddart's group synthesized the first molecular Borromean rings — three interlocked molecular rings such that cutting any one of them releases the other two. These are not metaphorically Borromean; they are literal macroscopic knot-theory objects realized at the nanoscale.

Stoddart (who received the 2016 Nobel Prize in Chemistry for work on molecular machines) demonstrated that chemists could design molecules whose topology — not just their atomic composition — was a meaningful engineering parameter. A knotted molecule and its unknotted version can have dramatically different mechanical and chemical properties.

Researchers are now exploring knotted proteins as well. Proteins whose amino acid chains fold into knotted configurations show unusual mechanical stability — a knot can make a protein behave like a tightrope that gets stronger when pulled. The topology of the fold, not just the sequence of amino acids, matters for function.

Takeaways

Mathematical knots are closed loops — no loose ends — and two knots are equivalent if one can be continuously deformed into the other without cutting. Proving two knots are different requires algebraic invariants, not intuition.

Knot theory emerged from wrong physics: Kelvin's vortex atom theory (1867) motivated Tait's knot tables, but the mathematical framework built along the way outlasted the physical theory by 150+ years and counting.

The Jones polynomial (1984) was a breakthrough that could detect chirality — the handedness of a knot — and its unexpected connection to quantum field theory revealed knots as fundamental mathematical objects, not just topological curiosities.

DNA, proteins, and chemotherapy all involve knot topology in concrete ways: topoisomerases unknot DNA during replication, knotted proteins have unique mechanical properties, and several cancer drugs work by disrupting the enzymes that solve knot problems in living cells.

Topological quantum computing uses knot theory as its literal mathematical foundation: quantum gates are braids, computations produce knots, and the inherent robustness of topology provides error resistance that conventional qubits lack.

Resources: - The Knot Book by Colin Adams — the definitive accessible introduction to knot theory - KnotInfo — a searchable database of knot invariants from Indiana University - Vaughan Jones's 1987 paper "Hecke Algebra Representations of Braid Groups and Link Polynomials" — the original Jones polynomial paper