A group captures a single reversible operation. But the number systems you actually compute with — the integers, the reals, the polynomials, matrices — carry two operations at once, addition and multiplication, laced together by the familiar distributive law a(b + c) = ab + ac. Strip that structure down to its bare essentials and you get a ring: the abstract home of "arithmetic," the setting in which "add" and "multiply" both make sense and cooperate.

The word sounds mysterious, but you have been living inside rings your whole life. The integers \mathbb{Z} are the founding example — Richard Dedekind and David Hilbert coined Ring (a "collection" that cycles back on itself) precisely to name what \mathbb{Z} and its cousins have in common. The hours on a clock form a ring. The 2 \times 2 matrices form a ring. Every polynomial you have ever factored lives in a ring. Once you spot the axioms, the same skeleton appears across arithmetic, algebra, geometry and number theory — and, exactly as with groups, a theorem proved for "rings in general" pays off everywhere at once.

In this page we pin down the ring axioms, watch them hold in five very different systems, meet the two fault lines that split rings into families — whether multiplication commutes, and whether there is a multiplicative identity — and, just as usefully, watch a would-be ring fail. By the end you should be able to take any set with a "plus" and a "times" and decide, axiom by axiom, whether it is a ring.

Two operations on a clock

Take the numbers \{0, 1, 2, 3, 4, 5\} and both add and multiply them the way a 6-hour clock would — wrapping around whenever you pass 6. This is modular arithmetic modulo 6, written \mathbb{Z}_6. So 4 + 5 = 9 \equiv 3 and 4 \times 5 = 20 \equiv 2. Both operations stay inside the set, and both tables tell a story.

The left table is a group table — every row a reshuffle of \{0,\dots,5\}, the fingerprint of a group (this is (\mathbb{Z}_6, +), an abelian group). The right table is looser: multiplication has an identity — the row and column headed by 1 copy the headers straight back — but rows now repeat values and the row headed by 0 is all zeros. Multiplication need not be reversible, and that asymmetry between the two operations is exactly what a ring is built to allow.

The formal definition

Notice the deliberate asymmetry. Addition is the strong operation: it is a full abelian group, so you can always subtract. Multiplication is the weak one: associative and (usually) with an identity, but not required to commute and not required to have inverses. The distributive law is the glue — the only axiom mentioning both operations — and it is what lets multiplication "see" the additive structure. From it alone you can already prove the reassuring fact that 0 \cdot a = 0 in every ring.

Five rings in the wild

The definition earns its keep only when you see how many different objects it captures. Here are five genuine rings, deliberately unalike.

The integers (\mathbb{Z}, +, \times). The prototype: addition is an abelian group, multiplication is associative and commutative with identity 1, and distributivity is the ordinary law you learned as a child. A commutative ring with unity.

The integers mod n, (\mathbb{Z}_n, +, \times). A finite ring — the clock above. Everything is inherited from \mathbb{Z} by reducing mod n. Commutative, with unity 1.

Polynomials \mathbb{R}[x]. All polynomials with real coefficients, added and multiplied as usual. Commutative, with unity the constant polynomial 1. We give this its own page later.

The 2 \times 2 real matrices M_2(\mathbb{R}). Matrices add entrywise and multiply by the row-times-column rule. Here is a ring where multiplication does not commute: AB \neq BA in general. A non-commutative ring with unity (the identity matrix I).

The even integers 2\mathbb{Z}. Even numbers are closed under both + and \times, so they form a ring — but there is no 1 inside (it is odd!). A commutative ring without unity ("rng," pun intended: a ring missing its identity).

Two questions that sort every ring

Two optional properties carve the world of rings into families, and naming them tells you almost everything about how a ring behaves.

Does multiplication commute? If ab = ba always, the ring is commutative (\mathbb{Z}, \mathbb{Z}_n, polynomials). If it can fail, the ring is non-commutative (matrices). Most of a first course lives among the commutative rings, where multiplication is as friendly as addition.

Is there a 1? A ring with unity (or "with identity") has an element 1 \neq 0 fixing everything under multiplication. Almost every important ring has one; conventions differ, but in this module — unless we say otherwise — "ring" means commutative ring with unity, the natural stage for number theory and algebraic geometry. When we need the extra room we will say "non-commutative" or "without unity" out loud.