If you hold a snowflake up and rotate it by 60 degrees, it looks the same. Rotate by 60 degrees again, it still looks the same. Do that six times and you’re back where you started. The snowflake has six rotational symmetries. Add in the reflections through its six axes, and you get twelve symmetries total.
That set of twelve operations is more than just a catalogue. It has structure. Doing two symmetries in a row gives you another symmetry in the set. Every symmetry has an inverse that undoes it. There’s a “do-nothing” symmetry that acts as an identity. That structure — closure, identity, inverses, and an associative operation — is the definition of a group.
Group theory is the study of these structures, and it is one of the most surprising unifiers in mathematics. The same abstract framework describes snowflakes, the Rubik’s cube, the integers under addition, the permutations of a deck of cards, the symmetries of the universe, and the error-correcting codes in your phone. Once you start seeing groups, you find them everywhere.
The simplest group
The smallest non-trivial group has just two elements: call them (identity) and . The rule is
You can realise this group concretely in many ways. It’s the symmetries of a rectangular card (flip it over vs. leave it alone). It’s the parity of integers (even + even = even, odd + odd = even). It’s the two elements of under addition mod 2, which is the fundamental building block of digital circuits.
Any two realisations of this group are, in the sense that matters, identical. Group theory is what you get when you stop caring which specific objects the group is acting on and start caring only about the rules of combination.
Symmetries of shapes
The most visual examples of groups come from symmetry. Consider an equilateral triangle. You can rotate it by 0°, 120°, or 240° and it looks the same. You can reflect it through any of three axes. That gives six symmetries total, and they form a group called the dihedral group of order 6, denoted .
Here’s the twist: the composition isn’t commutative. Rotating by 120° and then reflecting through a vertical axis does not give the same result as reflecting first and then rotating. Most groups you’ll meet outside arithmetic are like this — the order of operations matters.
The symmetries of a cube form a group with 48 elements. A dodecahedron has a group of order 120. These are finite groups, and they were classified painstakingly over the twentieth century. The classification of finite simple groups — completed in 2004 after tens of thousands of pages of proof by hundreds of mathematicians — is one of the great collective achievements of the field. It turns out there are exactly 18 infinite families of simple groups, plus 26 sporadic outliers, culminating in the astonishing Monster group of order roughly .
The Rubik’s cube as a group
If you want an example you can hold in your hands: the Rubik’s cube.
Every legal sequence of moves produces a state that can be labelled as “identity applied to cube” (if you undo it all) or some more complicated permutation of the stickers. The collection of all reachable states, with the operation “apply one sequence then another,” forms a group. Its order is
Just over 43 quintillion. Yet any configuration can be solved in at most 20 moves — a result called “God’s number,” proved by computer search in 2010. The fact that such a huge group has such a small diameter (every element reachable from identity in 20 steps or fewer) is deeply non-obvious and required about 35 CPU-years of computation to establish.
Cubers often don’t realise they’re doing group theory, but every clever sequence — “F R U R’ U’ F’” — is an element of the group, and commutators (applying a move, then another, then undoing the first) are the bread and butter of algorithmic cube-solving.
Groups in physics
The reason group theory matters outside of mathematics is that the laws of nature themselves are deeply symmetric.
Noether’s theorem, proved in 1918 by Emmy Noether, says: every continuous symmetry of a physical system corresponds to a conservation law. Symmetry under time translation gives conservation of energy. Symmetry under spatial translation gives conservation of momentum. Symmetry under rotation gives conservation of angular momentum. The deep structure of “things that don’t change” in physics is a structure of groups acting on configuration spaces.
The Standard Model of particle physics is entirely written in the language of group theory. The symmetries of electromagnetism form the group . The weak nuclear force, together with electromagnetism, has symmetry group . The strong force has . The entire theoretical framework is “quantum field theory with gauge group ,” and the reason physicists spend so much time studying Lie groups is that these continuous symmetry groups dictate which particles can exist and how they interact.
Galois, whose name is now attached to an entire branch of algebra, was 18 years old when he began developing the group-theoretic ideas that would eventually explain why the quintic equation has no general solution in radicals. He died in a duel at 20, leaving behind notes that took decades for others to fully understand. Almost two hundred years later, his ideas power everything from elliptic curve cryptography to the mathematics of quantum computing.
Why group theory is beautiful
Mathematicians often describe group theory as “beautiful,” which is a word that tends to turn people off if they haven’t done enough mathematics to see what it means. The beauty is specifically this: an enormous amount of mathematical structure is captured by four simple axioms, and the structure forces consequences that feel inevitable in hindsight but weren’t obvious going in.
Lagrange’s theorem: the order of any subgroup divides the order of the whole group. A four-line proof, but it immediately constrains the possible structure of every finite group in existence.
Cayley’s theorem: every group is a subgroup of a permutation group. Everything we ever called a symmetry is in some sense just a shuffling of things.
Sylow’s theorems: the number of subgroups of a given order is strongly constrained by number-theoretic conditions on the group’s size. Abstract algebra and number theory turn out to be joined at the hip.
These feel like results about the structure of reasoning itself — what must be true given just the rules of combination. And because of that, group theory shows up everywhere something has those rules: shapes, equations, particles, codes, cubes, tilings, molecules.
A takeaway
If you’ve never thought about symmetry systematically, try this exercise: pick an object around you — a chair, a piece of paper, a coffee mug — and enumerate everything you can do to it that leaves it looking the same. Include the “do nothing” operation. Compose two operations and check the result is one of your listed operations.
You’ve just constructed a group. What you’ve constructed has probably been studied for decades, and it almost certainly has something to say about objects you’d never have connected to your chair.
That’s the thing about group theory. Once you have the lens, you can’t stop finding places to use it.
Frequently asked
Why is it called a 'group' and not a 'set of symmetries'?
A group is a set together with an operation satisfying four axioms: closure, associativity, identity, and inverses. Symmetries are the most intuitive example, but many groups (like the integers under addition) don't arise as symmetries of an obvious object. The name 'group' is from Évariste Galois, who used it informally in 1830.
Is group theory useful outside of pure math?
Extremely. Group theory underpins modern physics (every conservation law comes from a symmetry, per Noether's theorem), chemistry (molecular symmetry determines spectra), cryptography (elliptic curve groups), and error-correcting codes (used in every CD, QR code, and deep-space communication).