JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Ideals and Quotients, Revisited

You already met ideals as kernels. Now we treat them as the right generalization of "multiples of n", learn to read the quotient ring R/I as a place where I becomes zero, and meet the operations — sum, product, intersection — that make ideals a calculus of their own.

From kernels to a calculus of ideals

In Volume I you proved that the ideals of a commutative ring R are exactly the kernels of ring homomorphisms out of R. That single fact is the whole motivation: an ideal I is a subset closed under addition and absorbing under multiplication (r·I ⊆ I for every r in R), and it is precisely the data you need to form the quotient ring R/I. Think of nZ ⊂ Z. The defining feature is not that nZ is a subgroup but that multiplying any integer by any multiple of n stays a multiple of n. That absorption is what lets the product of cosets be well defined.

The reframe for Volume II is this: stop thinking of an ideal as a static subset and start thinking of R/I as a new ring where the elements of I have been forced to zero. Every relation you want to impose on R — "set x²+1 = 0", "identify 2 with 0" — is the act of quotienting by the ideal those expressions generate. Ring theory becomes the study of which collapses are possible and what survives them.

Operations on ideals

Ideals form a structure rich enough to compute in. Given I and J you can form their sum I+J = {a+b : a∈I, b∈J}, the smallest ideal containing both; their intersection I∩J; and their product IJ, the ideal generated by all products ab. These are not random — they generalize gcd, lcm, and multiplication of integers. In Z, write I = (m) and J = (n); then I+J = (gcd(m,n)), I∩J = (lcm(m,n)), and IJ = (mn). The familiar identity gcd·lcm = mn becomes the ideal identity (I+J)(I∩J) ⊆ IJ, with equality exactly when the two are coprime.

Work inside R = Z. Take I = (12) and J = (18).

  I + J  = (gcd(12,18)) = (6)       # 12u + 18v ranges over all multiples of 6
  I ∩ J  = (lcm(12,18)) = (36)
  I · J  = (12·18) = (216)

Check the containment (I+J)(I∩J) ⊆ IJ:
  (6)(36) = (216) = IJ.    # here equality, since (I+J)(I∩J) = IJ in a PID

Coprime case: I = (4), J = (9).
  I + J = (gcd(4,9)) = (1) = R   # 4 and 9 are coprime
  I ∩ J = (36),  IJ = (36)       # so I ∩ J = IJ exactly when I + J = R
Ideal arithmetic in Z recovers gcd, lcm, and the coprimality condition.

When I+J = R we call I and J comaximal, and this is the hypothesis of the Chinese Remainder Theorem in its modern dress: if I and J are comaximal then R/(I∩J) ≅ R/I × R/J. The classical statement about congruences mod coprime moduli is exactly this isomorphism for R = Z.

Reading a quotient ring

The skill to internalize is computing in R/I by working in R and remembering "I is zero". The isomorphism theorems are your bookkeeping: a surjection R → S with kernel I gives R/I ≅ S, and the ideals of R/I correspond bijectively to the ideals of R that contain I. So the lattice of ideals above I in R is literally the lattice of ideals of the quotient.

  1. Identify R/I as a known ring: R[x]/(x²+1) ≅ C when R = R, because setting x² = −1 makes x behave like i.
  2. Detect collapses: in Z[x]/(2, x) every element reduces to its constant term mod 2, so the quotient is the field F₂ — both relations were needed.
  3. Watch for zero divisors you create: Z[x]/(x²−1) has (x−1)(x+1) = 0 with neither factor zero, so the quotient is not an integral domain.