Ideals and Quotient Rings

In this lesson you’ll learn what an ideal is and why a subring isn’t enough, how to build the quotient ring $R/I$, what makes an ideal maximal, and the theorem that turns an irreducible polynomial into a field.

For groups, the subobjects that let you form quotients are the normal subgroups. For rings there’s an analogous restriction, and an ordinary subring doesn’t qualify.

A subring is a subset closed under addition, negation, and multiplication. That’s enough to be a ring in its own right, but not enough to build a quotient.

An ideal $I \subseteq R$ satisfies two conditions:

1. $(I, +)$ is a subgroup of $(R, +)$.
2. Absorption: $ra \in I$ and $ar \in I$ for every $r \in R$ and $a \in I$.

Condition 2 is much stronger than requiring $I$ closed under its own multiplication. It says multiplying by anything at all in $R$, not just by elements of $I$, keeps you inside.

$3\mathbb{Z}$ is an ideal of $\mathbb{Z}$: any integer times a multiple of 3 is again a multiple of 3. But $\mathbb{Z}$ is a subring of $\mathbb{Q}$ that is not an ideal, since $\frac12 \cdot 1 = \frac12 \notin \mathbb{Z}$.

The analogy to remember: ideal is to ring as normal subgroup is to group. Both are exactly the subobjects you can quotient by.

For commutative $R$ with unity, the principal ideal generated by $a$ is

$$\langle a \rangle = \{ra : r \in R\}$$

all multiples of $a$. It’s the smallest ideal containing $a$.

In $\mathbb{Z}$, every ideal is principal: $\langle n \rangle = n\mathbb{Z}$. A ring where this always happens is a principal ideal domain, and $\mathbb{Z}$ and $F[x]$ for a field $F$ are both examples. In $\mathbb{Z}[x]$ it fails: the ideal of polynomials with even constant term needs two generators, $\langle 2, x \rangle$.

Two special cases: $\langle 0 \rangle = \{0\}$ and $\langle 1 \rangle = R$. So $\{0\}$ and $R$ are always ideals, and any ideal containing a unit is all of $R$, since $u \in I$ forces $1 = u^{-1}u \in I$ and then $r = r \cdot 1 \in I$.

Consequence: a field has only the two trivial ideals. There’s nothing to quotient by, which is part of what makes fields the end of the line.

Given an ideal $I \trianglelefteq R$, the cosets $a + I$ form a ring under

$$(a + I) + (b + I) = (a+b) + I, \qquad (a+I)(b+I) = ab + I$$

Absorption is exactly what makes the multiplication well-defined. Check it: if $a' = a + i$ and $b' = b + j$ with $i, j \in I$, then

$$a'b' = ab + aj + ib + ij$$

The three extra terms all lie in $I$ - $aj$ and $ib$ by absorption, $ij$ because $I$ is closed. So $a'b' \in ab + I$, the same coset. Without absorption, $aj$ escapes and the product is meaningless.

The model example: $\mathbb{Z}/n\mathbb{Z} = \mathbb{Z}_n$. Quotienting the integers by the ideal $n\mathbb{Z}$ gives modular arithmetic, and now both operations carry over, not just addition.

Two conditions on an ideal $I \ne R$ in a commutative ring with unity:

• $I$ is prime if $ab \in I$ implies $a \in I$ or $b \in I$.
• $I$ is maximal if no ideal sits strictly between $I$ and $R$.

They connect to the quotient in a clean way:

$$I \text{ prime} \iff R/I \text{ is an integral domain}$$ $$I \text{ maximal} \iff R/I \text{ is a field}$$

In $\mathbb{Z}$ the prime ideals are $\{0\}$ and $\langle p \rangle$ for prime $p$, and the maximal ones are exactly $\langle p \rangle$. That matches what we know: $\mathbb{Z}/p\mathbb{Z}$ is a field, $\mathbb{Z}/\{0\} \cong \mathbb{Z}$ is a domain but not a field.

The word “prime” is not a coincidence. For $\langle n \rangle$ to satisfy the prime-ideal condition means $n \mid ab$ implies $n \mid a$ or $n \mid b$, which is Euclid’s lemma and holds exactly for primes.

Here’s the payoff, and it’s the engine of the next two lessons.

If $F$ is a field and $f \in F[x]$ is irreducible, then $\langle f \rangle$ is a maximal ideal and

$$F[x]/\langle f \rangle$$

is a field.

Its elements are the possible remainders on division by $f$, so it has as many elements as there are polynomials of degree less than $\deg f$.

This is how the complex numbers get built. Take $\mathbb{R}[x]$, note $x^2 + 1$ is irreducible over $\mathbb{R}$, and form

$$\mathbb{R}[x]/\langle x^2+1 \rangle \cong \mathbb{C}$$

The coset of $x$ becomes a square root of $-1$, because in the quotient $x^2 + 1 = 0$. The complex numbers are a quotient ring, and $i$ is not a mysterious invented object but the name of a coset.

The group version carries over verbatim. A ring homomorphism satisfies

$$\varphi(a+b) = \varphi(a)+\varphi(b), \qquad \varphi(ab) = \varphi(a)\varphi(b)$$

Its kernel is always an ideal, its image is a subring, and

$$R/\ker\varphi \cong \operatorname{im}\varphi$$

Just as every normal subgroup is a kernel, every ideal is the kernel of a ring homomorphism - the projection $R \to R/I$.

Example 1: Is the set of even integers an ideal of $\mathbb{Z}$?

Solution. Additive subgroup ✓. Absorption: for any integer $r$ and even $a = 2m$,

$$ra = 2(rm)$$

which is even ✓

Yes, $2\mathbb{Z} = \langle 2 \rangle$ is an ideal. And $\mathbb{Z}/2\mathbb{Z} \cong \mathbb{Z}_2$, a field, consistent with $\langle 2 \rangle$ being maximal since 2 is prime.

Example 2: Show $\mathbb{Z}$ is not an ideal of $\mathbb{Q}$.

Solution. Additive subgroup ✓. Absorption fails: take $r = \frac13 \in \mathbb{Q}$ and $a = 1 \in \mathbb{Z}$.

$$ra = \tfrac13 \notin \mathbb{Z}$$

Not an ideal, though it is a subring.

In fact $\mathbb{Q}$ is a field, so its only ideals are $\{0\}$ and $\mathbb{Q}$ - there was no chance.

Example 3: Compute $\mathbb{Z}[x]/\langle x \rangle$.

Solution. $\langle x \rangle$ is all polynomials with zero constant term. Two polynomials lie in the same coset exactly when their constant terms agree.

So a coset is named by its constant term, and

$$\mathbb{Z}[x]/\langle x \rangle \cong \mathbb{Z}$$

Formally, the evaluation map $f \mapsto f(0)$ is a surjective ring homomorphism $\mathbb{Z}[x] \to \mathbb{Z}$ with kernel $\langle x \rangle$, so the first isomorphism theorem gives the result.

Since $\mathbb{Z}$ is a domain but not a field, $\langle x \rangle$ is prime but not maximal in $\mathbb{Z}[x]$.

Example 4: Construct $\mathbb{C}$ as a quotient.

Solution. $x^2 + 1$ has no real root, and it’s degree 2, so it’s irreducible over $\mathbb{R}$.

Elements of $\mathbb{R}[x]/\langle x^2+1\rangle$ are remainders on division by a quadratic, so they have the form $a + bx$ with $a, b \in \mathbb{R}$.

Multiplication: reduce using $x^2 \equiv -1$.

$$(a+bx)(c+dx) = ac + (ad+bc)x + bdx^2 \equiv (ac - bd) + (ad+bc)x$$

Compare complex multiplication:

$$(a+bi)(c+di) = (ac-bd) + (ad+bc)i$$

Identical. So $\mathbb{R}[x]/\langle x^2+1 \rangle \cong \mathbb{C}$, with $x \leftrightarrow i$.

Example 5: A quotient that is not a domain.

Compute $\mathbb{R}[x]/\langle x^2 - 1 \rangle$.

Solution. $x^2 - 1 = (x-1)(x+1)$ is reducible, so $\langle x^2-1\rangle$ is not maximal and the quotient shouldn’t be a field.

Indeed, in the quotient, $(x-1)(x+1) = x^2 - 1 \equiv 0$ while neither factor is zero. Zero divisors, so not even a domain.

In fact $\mathbb{R}[x]/\langle x^2-1\rangle \cong \mathbb{R} \times \mathbb{R}$, by the Chinese Remainder Theorem applied to the coprime factors. Irreducibility of the modulus is exactly what prevents this splitting.

Example 6: Maximal ideals of $\mathbb{Z}$.

Solution. Every ideal of $\mathbb{Z}$ is $\langle n \rangle$ for some $n \ge 0$, and $\langle m \rangle \subseteq \langle n \rangle$ exactly when $n \mid m$.

So $\langle n \rangle$ is maximal when the only divisors of $n$ are 1 and $n$ - that is, when $n$ is prime.

$$\text{maximal ideals of } \mathbb{Z} = \{\langle p \rangle : p \text{ prime}\}$$

Confirmed by the quotient test: $\mathbb{Z}/\langle p \rangle = \mathbb{Z}_p$ is a field exactly when $p$ is prime.

Note $\langle 0 \rangle$ is prime (since $ab = 0$ forces $a=0$ or $b=0$ in $\mathbb{Z}$) but not maximal, since $\langle 0 \rangle \subsetneq \langle 2 \rangle \subsetneq \mathbb{Z}$. Prime does not imply maximal, though the converse holds in a commutative ring with unity.

Example 7: The kernel of evaluation.

Find the kernel of $\varphi: \mathbb{R}[x] \to \mathbb{R}$ given by $\varphi(f) = f(2)$.

Solution. $\varphi$ is a ring homomorphism, since evaluating a sum or product at a point respects both operations.

$$\ker\varphi = \{f : f(2) = 0\} = \langle x - 2 \rangle$$

by the factor theorem: $f(2) = 0$ exactly when $x - 2$ divides $f$.

$\varphi$ is surjective (constants hit everything), so

$$\mathbb{R}[x]/\langle x-2 \rangle \cong \mathbb{R}$$

$\mathbb{R}$ is a field, so $\langle x-2\rangle$ is maximal - consistent with $x - 2$ being irreducible.

AES encryption. The AES S-box computes multiplicative inverses in $GF(2^8) = \mathbb{F}_2[x]/\langle x^8+x^4+x^3+x+1 \rangle$. That specific irreducible polynomial is written into the standard, and the whole cipher is arithmetic in this quotient ring.

Reed–Solomon codes. Encoding is multiplication in a quotient ring $F[x]/\langle x^n - 1 \rangle$, whose elements are cyclic codes. The cyclic shift structure that makes hardware encoders cheap is a consequence of quotienting by $x^n - 1$.

CRC checksums. A CRC is the remainder of the message polynomial modulo a fixed generator over $\mathbb{F}_2$, which is to say the message’s image in a quotient ring. The generator polynomial’s factorization determines exactly which burst errors are caught.

Complex numbers in software. Implementing complex arithmetic as pairs with the rule $(a,b)(c,d) = (ac-bd, ad+bc)$ is implementing the quotient ring construction. The same pattern extends to quaternions and to dual numbers, which are $\mathbb{R}[x]/\langle x^2 \rangle$ and are used for automatic differentiation in machine-learning libraries.

Algebraic geometry. A geometric shape corresponds to an ideal of polynomials vanishing on it, and the quotient ring is the ring of functions on the shape. Points correspond to maximal ideals. This dictionary between ideals and geometry is the foundation of the whole field.

Computer algebra. Gröbner bases compute canonical forms for ideals, which is how systems solve polynomial systems, do geometric theorem proving, and check whether one polynomial lies in the ideal generated by others.