Chapter 11
Medium

Rings — Definition and Examples

00 · Symbol Glossary

$(R,+,\times)$A ring

A set RR with two operations: addition and multiplication, interacting through the distributive law. The central object of the second half of this course, the way (G,)(G,\ast) was for the first half.

$0$Additive identity

The identity element for ++: a+0=aa+0=a for all aRa\in R. Guaranteed by (R,+)(R,+) being an abelian group.

$1$Multiplicative identity (unity)

An element with 1a=a1=a1\cdot a=a\cdot1=a for all aRa\in R, when one exists. A ring possessing such an element is called a ring with unity; this course's rings are unital by convention unless stated otherwise.

$-a$Additive inverse

The inverse of aa under ++, guaranteed since (R,+)(R,+) is a group. Always exists — unlike multiplicative inverses, which are special.

$R^\times$Group of units

R×={aR:a has a multiplicative inverse}R^\times = \{a\in R : a \text{ has a multiplicative inverse}\}. Always a group under multiplication (Theorem 11.3) — the ring-theoretic analog of "the invertible elements."

$M_n(\mathbb{R})$n by n real matrices

All n×nn\times n real matrices (not just invertible ones — contrast with GLn(R)GL_n(\mathbb{R}) from Chapter 02), under matrix addition and multiplication. The standard example of a noncommutative ring.


01 · The Definition

Groups modeled a single operation. Rings model two operations at once — addition and multiplication — linked by the requirement that multiplication distributes over addition, exactly as ordinary arithmetic behaves.

Definition — Ring

A ring is a set RR with two binary operations ++ and ×\times satisfying:

R1. (R,+)(R,+) is an abelian group (identity 00, every element has an additive inverse, ++ is commutative).

R2. ×\times is associative: (ab)c=a(bc)(ab)c=a(bc).

R3. Distributive laws: a(b+c)=ab+aca(b+c)=ab+ac and (a+b)c=ac+bc(a+b)c=ac+bc, for all a,b,cRa,b,c\in R.

If additionally ab=baab=ba for all a,ba,b, RR is commutative. If there exists 1R1\in R with 101\neq0 and 1a=a1=a1\cdot a=a\cdot1=a for all aa, RR has unity (or is unital). This course's rings are commutative with unity unless stated otherwise.

Multiplication need not have inverses, or even be commutative

Notice how much weaker R2–R3 are compared to the group axioms: multiplication need not have an identity, need not have inverses, and need not commute. A ring guarantees a full group's worth of structure for ++, but only associativity (linked to ++ via distributivity) for ×\times. Most of this chapter is about which of these extra multiplicative properties a given ring happens to have.

Example — The integers, the motivating case

(Z,+,×)(\mathbb{Z},+,\times): (Z,+)(\mathbb{Z},+) is an abelian group (Chapter 02), ×\times is associative and commutative, distributivity is ordinary arithmetic, and 1Z1\in\mathbb{Z} is the multiplicative identity. A commutative ring with unity — the model every other ring in this chapter is compared against.


02 · First Consequences of the Axioms

A few facts hold in every ring, before any specific example is chosen — direct analogs of Chapter 02's first consequences of the group axioms.

Theorem 11.1 — Multiplication by Zero

For every aRa\in R: 0a=a0=00\cdot a = a\cdot0 = 0.

Proof

0a=(0+0)a=0a+0a0\cdot a = (0+0)\cdot a = 0\cdot a + 0\cdot a, using 0+0=00+0=0 (identity law in (R,+)(R,+)) and distributivity. Let x=0ax=0\cdot a; the equation reads x=x+xx=x+x. Add x-x to both sides: x+(x)=(x+x)+(x)x+(-x) = (x+x)+(-x), i.e. 0=x+(x+(x))=x+0=x0 = x+(x+(-x)) = x+0 = x (associativity in (R,+)(R,+), then inverse and identity laws). So 0a=00\cdot a=0. The argument for a0=0a\cdot0=0 is symmetric, using a(0+0)=a0+a0a(0+0)=a0+a0.

Theorem 11.2 — Signs Under Multiplication

For all a,bRa,b\in R: (a) (a)b=a(b)=(ab)(-a)b = a(-b) = -(ab); (b) (a)(b)=ab(-a)(-b) = ab.

Proof

(a) ab+(a)b=(a+(a))b=0b=0ab + (-a)b = (a+(-a))b = 0\cdot b = 0 by distributivity and Theorem 11.1. So (a)b(-a)b satisfies the defining property of the additive inverse of abab; by uniqueness of inverses in the group (R,+)(R,+) (Theorem 1.2), (a)b=(ab)(-a)b=-(ab). The argument for a(b)=(ab)a(-b)=-(ab) is symmetric.

(b) Apply part (a) twice: (a)(b)=(a(b))(-a)(-b) = -(a(-b)) (treating a-a as the "a-a" in part (a) applied to a(b)a(-b)... more directly: (a)(b)=(a(b))(-a)(-b) = -(a \cdot(-b)) by part (a) with bb replaced by b-b, and a(b)=(ab)a\cdot(-b) = -(ab) by part (a) directly. So (a)(b)=((ab))=ab(-a)(-b) = -(-(ab)) = ab, using that the additive inverse of an additive inverse returns the original element (a direct consequence of uniqueness of inverses in (R,+)(R,+)).

Example — Sanity check in Z

Theorem 11.2 predicts (3)(5)=3×5=15(-3)(-5)=3\times5=15 and (3)(5)=(3×5)=15(-3)(5) = -(3\times5)=-15 — exactly ordinary arithmetic. The theorem's content is that these familiar sign rules are forced by the ring axioms alone (distributivity plus additive inverses), not a special feature of the integers.


Example — $\mathbb{Z}_n$ as a ring

(Zn,+,×)(\mathbb{Z}_n, +, \times): additive structure from Chapter 01's Zn\mathbb{Z}_n, multiplication [a][b]:=[ab][a][b]:=[ab] well-defined by Theorem 1.5. Commutative, unity [1][1]. A finite commutative ring with unity, for every n2n\geq2.

Example — $M_n(\mathbb{R})$, a noncommutative ring

All n×nn\times n real matrices under matrix addition and multiplication. (Mn(R),+)(M_n(\mathbb{R}),+) is abelian (componentwise addition). Matrix multiplication is associative and distributes over addition (standard linear algebra facts). Unity: the identity matrix II. Noncommutative for n2n\geq2 (Chapter 02's example), and — unlike GLn(R)GL_n(\mathbb{R}) — this ring includes all matrices, not just invertible ones, so most elements have no multiplicative inverse.

Common mistake — 2Z has no unity

(2Z,+,×)(2\mathbb{Z}, +, \times) (even integers): closed under both operations (even+even=even, even×\timeseven=even), abelian additive group (Chapter 01, Exercise 1.1's style argument), associative and distributive (inherited from Z\mathbb{Z}). But is there u2Zu\in2\mathbb{Z} with ua=aua=a for every even aa? Only u=1u=1 could work, and 12Z1\notin2\mathbb{Z}. 2Z2\mathbb{Z} is a ring without unity — a genuine example showing R1–R3 alone do not force a multiplicative identity to exist.


04 · Units and Zero Divisors

Definition — Unit

An element aRa\in R (ring with unity) is a unit if there exists bRb\in R with ab=ba=1ab=ba=1. Write R×R^\times for the set of all units.

Theorem 11.3 — R× Is a Group

(R×,×)(R^\times, \times) is a group.

Proof

Closure: if a,cR×a,c\in R^\times with inverses a1,c1a^{-1},c^{-1}, then (ac)(c1a1)=a(cc1)a1=a1a1=aa1=1(ac)(c^{-1}a^{-1}) = a(cc^{-1})a^{-1} = a\cdot1\cdot a^{-1}=aa^{-1}=1, and similarly on the other side, so acR×ac\in R^\times (this is exactly Theorem 2.3's argument, transplanted from groups). Associativity: inherited from RR's multiplication (R2). Identity: 1R×1\in R^\times since 11=11\cdot1=1. Inverses: by definition of R×R^\times, every element already has an inverse inside R×R^\times (the inverse a1a^{-1} is itself a unit, with inverse aa). All group axioms hold.

Example — Units of Z

Z×={1,1}\mathbb{Z}^\times = \{1,-1\}: these are the only integers with an integer multiplicative inverse (11=11\cdot1=1, (1)(1)=1(-1)(-1)=1; no other integer nn has 1/nZ1/n\in\mathbb{Z} unless n=±1n=\pm1).

Theorem 11.4 — Units of $\mathbb{Z}_n$

[a]Zn×[a]\in\mathbb{Z}_n^\times if and only if gcd(a,n)=1\gcd(a,n)=1.

Proof

(\Leftarrow) If gcd(a,n)=1\gcd(a,n)=1, Bézout's identity gives integers x,yx,y with ax+ny=1ax+ny=1. Reducing mod nn: [a][x]+[n][y]=[1][a][x] + [n][y] = [1], and [n]=[0][n]=[0], so [a][x]=[1][a][x]=[1]: [x][x] is a multiplicative inverse for [a][a].

(\Rightarrow) If [a][x]=[1][a][x]=[1] for some xx, then ax1(modn)ax\equiv1\pmod n, so ax1=nyax-1=ny for some integer yy, i.e. axny=1ax-ny=1. Any common divisor dd of aa and nn divides the left side axnyax-ny, hence divides 11, forcing d=1d=1: gcd(a,n)=1\gcd(a,n)=1.

Example — Units of $\mathbb{Z}_{12}$

gcd(a,12)=1\gcd(a,12)=1 for a{1,5,7,11}a\in\{1,5,7,11\} (the integers in {1,,11}\{1,\ldots,11\} sharing no common factor with 12=22×312=2^2\times3). So Z12×={1,5,7,11}\mathbb{Z}_{12}^\times = \{1,5,7,11\}, a group of order 4 under multiplication mod 12 — matching Exercise 1.5's direct computation for Z5\mathbb{Z}_5 as a special case where nn is prime (every nonzero remainder is automatically coprime to a prime).

Definition — Zero Divisor

A nonzero aRa\in R is a zero divisor if ab=0ab=0 for some nonzero bRb\in R.

Example — Zero divisors in $\mathbb{Z}_4$ and $M_2(\mathbb{R})$

In Z4\mathbb{Z}_4: [2][2]=[4]=[0][2][2]=[4]=[0], and [2][0][2]\neq[0] — so [2][2] is a zero divisor (Chapter 01's FailBlock on Z4\mathbb{Z}_4, now named). In M2(R)M_2(\mathbb{R}): (1000)(0001)=(0000)\begin{pmatrix}1&0\\0&0\end{pmatrix}\begin{pmatrix}0&0\\0&1\end{pmatrix} = \begin{pmatrix}0&0\\0&0\end{pmatrix}, two nonzero matrices multiplying to the zero matrix — zero divisors exist even in this infinite, familiar ring.

Definition — Integral Domain

A commutative ring with unity (101\neq0) having no zero divisors is called an integral domain.

Theorem 11.5 — Cancellation in Domains

In an integral domain RR, if ab=acab=ac and a0a\neq0, then b=cb=c.

Proof

ab=ac    abac=0    a(bc)=0ab=ac \implies ab-ac=0 \implies a(b-c)=0 (distributivity and Theorem 11.2). Since RR has no zero divisors and a0a\neq0, the factor (bc)(b-c) must be 00 (otherwise aa would be a zero divisor, witnessed by the nonzero bcb-c). So bc=0b-c=0, i.e. b=cb=c.

Common mistake — Cancellation fails outside domains

In Z4\mathbb{Z}_4: [2][1]=[2][2][1]=[2] and [2][3]=[6]=[2][2][3]=[6]=[2], so [2][1]=[2][3][2][1]=[2][3] — but [1][3][1]\neq[3]. Cancellation fails because [2][2] is a zero divisor, so Z4\mathbb{Z}_4 is not an integral domain, and Theorem 11.5 simply doesn't apply.


05 · Subrings

Definition — Subring

A subset SRS\subseteq R is a subring if SS is itself a ring under RR's operations (restricted to SS).

Theorem 11.6 — Subring Test

A nonempty subset SRS\subseteq R is a subring if and only if: (S1) a,bS    abSa,b\in S \implies a-b\in S; (S2) a,bS    abSa,b\in S \implies ab\in S.

Proof

(\Rightarrow) A subring satisfies its own R1 (in particular, closed under ++ and additive inverses, so closed under subtraction) and R2/closure of ×\times.

(\Leftarrow) S1 is exactly the subgroup test (Theorem 3.1) applied to (S,+)(S,+) inside (R,+)(R,+) — recall subtraction closure is equivalent to closure under ++ and inverses simultaneously (if a,bSa,b\in S then 0=aaS0=a-a\in S, then b=0bS-b=0-b\in S, then a+b=a(b)Sa+b=a-(-b)\in S). So (S,+)(S,+) is an abelian subgroup of (R,+)(R,+). S2 gives closure of ×\times. Associativity of ×\times and both distributive laws are inherited from RR, since they hold for all elements of RR, hence for the subset SS. All ring axioms hold for SS.

Example — $\mathbb{Z} \le \mathbb{Q} \le \mathbb{R} \le \mathbb{C}$, a chain of subrings

Each is closed under subtraction and multiplication within the next: integers are closed under both inside Q\mathbb{Q}, and so on up the chain. Each number system in this familiar chain is a subring of the next.

Common mistake — Verifying only closure under multiplication is not enough

Is N={0,1,2,}\mathbb{N}=\{0,1,2,\ldots\} a subring of Z\mathbb{Z}? Closed under multiplication: yes. Closed under subtraction: no25=3N2-5=-3\notin\mathbb{N}. Fails S1, so N\mathbb{N} is not a subring, even though it's closed under ×\times and contains 0,10,1. Both parts of Theorem 11.6 are required.


06 · Exercises

EXERCISE 11.1

Apply Theorem 11.2 directly to a familiar ring, treating 3-3 and 5-5 as (3)-(3) and (5)-(5).

By Theorem 11.2(b), (a)(b)=ab(-a)(-b)=ab for any ring. With a=3,b=5a=3,b=5 in Z\mathbb{Z}: (3)(5)=3×5=15(-3)(-5) = 3\times5 = 15 — forced by the ring axioms (distributivity + additive inverses), not a separate arithmetic fact.

Use Theorem 11.2 to justify why (3)×(5)=15(-3)\times(-5) = 15 in Z\mathbb{Z}, citing the specific theorem rather than "common knowledge."

EXERCISE 11.2

Apply Theorem 11.4: which elements of {1,,7}\{1,\ldots,7\} share no common factor with 8?

8=238=2^3. Elements coprime to 8 among {1,,7}\{1,\ldots,7\}: exclude even numbers (2,4,62,4,6), leaving {1,3,5,7}\{1,3,5,7\}. By Theorem 11.4, Z8×={1,3,5,7}\mathbb{Z}_8^\times = \{1,3,5,7\}.

Find Z8×\mathbb{Z}_8^\times using Theorem 11.4.

EXERCISE 11.3

Check the definition of integral domain directly: does Z6\mathbb{Z}_6 have any zero divisors?

[2][3]=[6]=[0][2][3] = [6] = [0] in Z6\mathbb{Z}_6, with both [2][2] and [3][3] nonzero. So [2][2] (and [3][3]) are zero divisors. Z6\mathbb{Z}_6 is not an integral domain. (This matches Theorem 11.4-style reasoning: 66 is not prime, so some nonzero remainders share a common factor with 66 — namely 22 and 33 — and it is exactly those shared-factor pairs that multiply to [0][0].)

Is Z6\mathbb{Z}_6 an integral domain? Exhibit a zero divisor if not.

EXERCISE 11.4

Apply Theorem 11.6: check closure under subtraction and multiplication for matrices with even integer entries.

Let S={AM2(Z):every entry of A is even}S = \{A \in M_2(\mathbb{Z}) : \text{every entry of } A \text{ is even}\}. Subtraction: entrywise, even minus even is even, so ABSA-B\in S. Multiplication: each entry of ABAB is a sum of products of entries from AA and BB; since every entry of AA (and BB) is even, each such product is a multiple of 44 (hence even), and a sum of even numbers is even, so ABSAB\in S. Both conditions of Theorem 11.6 hold: SS is a subring of M2(Z)M_2(\mathbb{Z}) — notably without unity, since ISI\notin S (its diagonal entries are 11, odd).

Show that the set of 2×22\times2 integer matrices with all-even entries is a subring of M2(Z)M_2(\mathbb{Z}).


07 · Chapter Summary

ConceptStatement
Ring(R,+,×)(R,+,\times): (R,+)(R,+) abelian group, ×\times associative, distributive over ++
0a=a0=00\cdot a = a\cdot0=0Theorem 11.1
Sign rules(a)b=a(b)=(ab)(-a)b=a(-b)=-(ab); (a)(b)=ab(-a)(-b)=ab (Thm 11.2)
Unitaa with a multiplicative inverse; R×R^\times is always a group (Thm 11.3)
Units of Zn\mathbb{Z}_n[a]Zn×    gcd(a,n)=1[a]\in\mathbb{Z}_n^\times \iff \gcd(a,n)=1 (Thm 11.4)
Zero divisorNonzero aa with ab=0ab=0 for some nonzero bb
Integral domainCommutative, unital, no zero divisors
Cancellation in domainsab=ac,a0    b=cab=ac,\,a\neq0 \implies b=c (Thm 11.5)
Subring testClosed under subtraction and multiplication (Thm 11.6)

Next: Chapter 12 — Ideals & Quotient Rings identifies the ring-theoretic analog of a normal subgroup — a subset absorbing multiplication from the whole ring — and uses it to build quotient rings the same way Chapter 05 built quotient groups.