Introduction to the Ideal Class Group

Core Idea

The ideal class group measures how far a number ring departs from unique factorization. In rings of algebraic integers where elements may not factor uniquely, ideals always factor uniquely into prime ideals. Two ideals are equivalent if they differ by multiplication by a principal ideal. The class group is the quotient of fractional ideals by principal ideals, and its order—the class number h(K)—equals 1 precisely when the ring is a principal ideal domain with unique factorization. Computing class numbers reveals the arithmetic complexity of number fields and connects to deep results in algebraic number theory.

How It's Best Learned

Work through ℤ[√−5], where 6 = 2 · 3 = (1+√−5)(1−√−5) shows factorization failure. Then verify that ideal factorization restores uniqueness and compute that h = 2, making the class group ℤ/2ℤ.

Common Misconceptions

The class group is not about individual elements failing to factor—it is about the global structure of ideals. Students sometimes think unique factorization fails "everywhere" when h > 1, but many elements still factor uniquely; it is the exceptions that the class group quantifies.

Explainer

From your study of the failure of unique factorization, you know that some rings of algebraic integers do not behave like ℤ. In ℤ[√−5], the equation 6 = 2 · 3 = (1 + √−5)(1 − √−5) gives two genuinely distinct factorizations into irreducibles — the ring is not a unique factorization domain. The ideal class group is the algebraic object that measures exactly how badly unique factorization fails. It does not just say "factorization is broken" — it quantifies the structural obstruction and organizes it into a group.

The key insight from your study of ideals is that while elements may not factor uniquely, ideals always do in a Dedekind domain (which every ring of algebraic integers is). In ℤ[√−5], the ideals (2), (3), (1 + √−5), and (1 − √−5) are not prime ideals, but each can be factored uniquely into products of prime ideals. For instance, (2) = 𝔭₁² where 𝔭₁ = (2, 1 + √−5) is a prime ideal. The passage from elements to ideals restores unique factorization — the problem is not that factorization is impossible, but that it happens at the level of ideals rather than elements.

Two fractional ideals I and J are declared equivalent if I = αJ for some nonzero element α of the field — that is, they differ by multiplication by a principal ideal. The equivalence classes form a group under ideal multiplication, called the ideal class group Cl(K). The identity element is the class of principal ideals (those of the form (α) for some element α). The class number h(K) = |Cl(K)| counts how many equivalence classes there are. The critical fact is: h(K) = 1 if and only if every ideal is principal, which happens if and only if the ring is a PID, which for Dedekind domains is equivalent to being a UFD. So h(K) = 1 is the precise algebraic condition for unique factorization to hold.

For ℤ[√−5], the class group is ℤ/2ℤ, so h = 2. There are exactly two ideal classes: the principal ideals and one non-trivial class represented by 𝔭₁ = (2, 1 + √−5). The non-unique factorization of 6 is a direct consequence: the prime ideal factorization of (6) passes through non-principal ideals, so the factorization at the ideal level cannot be "lifted" to a unique factorization at the element level. Computing class numbers for specific number fields — using Minkowski's bound to reduce the computation to finitely many ideals — is one of the central practical tasks of algebraic number theory and connects to deep results including the analytic class number formula involving L-functions.