Rings ↔ spaces, ideals ↔ subvarieties, primes ↔ irreducibles — the Rosetta Stone of algebraic geometry
Algebraic geometry is built on a profound correspondence: every geometric concept has an algebraic counterpart, and vice versa. This dictionary is the Rosetta Stone of the subject — once you internalize it, you can translate freely between the language of rings and ideals and the language of curves and surfaces.
The correspondence is contravariant: bigger ideals correspond to smaller varieties, and a map of varieties going one way induces a ring homomorphism going the other way. This reversal is initially confusing but ultimately powerful — it lets you use the full machinery of commutative algebra to study geometry.
Click an algebraic concept on the left to see its geometric counterpart on the right, or click a geometric concept to see its algebraic translation. Each entry includes a concrete example.
The generators of I are the equations; their common zero set is the variety.
The polynomial ring k[x, y] is the ring of all polynomial "functions" on the affine plane. Each polynomial f(x, y) assigns a value to every point (a, b) in A². The ring structure (addition and multiplication) corresponds to pointwise operations on these functions.
More generally, k[x₁, ..., xₙ] is the ring of polynomial functions on affine n-space Aⁿ. This is the starting point: the ambient ring encodes the ambient space. (See: Rings & Fields → Polynomial Rings)
An ideal I = (f₁, ..., fₘ) determines a variety V(I) = { p ∈ A² : f₁(p) = ··· = fₘ(p) = 0 }. The key reversal: bigger ideal ⇒ smaller variety. Adding generators to I imposes more equations, so V(I) shrinks.
(0) ⊂ (y - x²) ⊂ (x, y)
↕ ↕ ↕
A² ⊃ parabola ⊃ { origin }
The containment reverses! The zero ideal (0) imposes no constraints, so V((0)) = A² (the whole plane). The maximal ideal (x, y) imposes two constraints, cutting down to a single point. (See: Rings & Fields → Ideals)
The quotient ring k[V] = k[x, y] / I(V) is the coordinate ring of V — the ring of polynomial functions restricted to V. Two polynomials f and g that agree on V (meaning f - g ∈ I(V)) are identified in the quotient.
# On the circle V(x² + y² - 1):
k[V] = k[x, y] / (x² + y² - 1)
# In this ring, x² + y² = 1, so:
x² = 1 - y² (geometry constrains algebra)
The coordinate ring encodes the variety completely: you can recover V from k[V] as the set of maximal ideals of k[V]. This is the spectrum construction, which is the foundation of scheme theory. (See: Rings & Fields → Quotient Rings)
An ideal I is prime if fg ∈ I implies f ∈ I or g ∈ I. Geometrically, V(I) is irreducible — it cannot be written as a union of two proper subvarieties. This is the algebraic version of "connected in one piece."
Prime: (y - x²) → parabola (irreducible)
Not prime: (xy) → x-axis ∪ y-axis (reducible: xy = 0 ⟹ x = 0 or y = 0)
Prime: (x) → y-axis (irreducible)
Maximal ideals are prime ideals that are as large as possible — they correspond to individual points, the smallest irreducible varieties. (See: Rings & Fields → Ideals, specifically prime vs. maximal)
A polynomial map φ: V → W between varieties induces a ring homomorphism φ*: k[W] → k[V] going the other way — by composition (pullback). If g is a function on W, then φ*(g) = g ∘ φ is a function on V.
# Map φ: A¹ → parabola, t ↦ (t, t²)
# Pullback φ*: k[x, y]/(y - x²) → k[t]
φ*(x) = t, φ*(y) = t²
# Direction reverses! φ goes A¹ → V, but φ* goes k[V] → k[A¹]
This contravariance is the signature feature of algebraic geometry. It means that the category of affine varieties is equivalent to the opposite of the category of finitely generated reduced k-algebras. (See: Rings & Fields → Homomorphisms)
The local ring at a point p is obtained by localizing the coordinate ring — allowing division by functions that don't vanish at p. It captures the geometry of V in an infinitesimal neighborhood of p, discarding information about the rest of the variety.
A point is smooth (non-singular) if and only if its local ring is a regular local ring — isomorphic to a power series ring k[[x₁, ..., xₙ]]. This is the algebraic version of "the variety looks like affine space near p." Singularities are precisely the points where the local ring is not regular.
| Algebra | Geometry |
|---|---|
| k[x₁, ..., xₙ] | Affine n-space Aⁿ |
| Ideal I ⊂ k[x₁, ..., xₙ] | Subvariety V(I) ⊂ Aⁿ |
| Radical ideal √I | Variety V(I) (same as V(√I)) |
| Prime ideal | Irreducible variety |
| Maximal ideal (x₁-a₁, ..., xₙ-aₙ) | Point (a₁, ..., aₙ) |
| Quotient ring k[V] = k[x]/I(V) | Coordinate ring (functions on V) |
| Ring homomorphism k[W] → k[V] | Morphism V → W (contravariant!) |
| Localization at p | Infinitesimal neighborhood of p |
| Regular local ring | Smooth point |
| Dimension (Krull dimension) | Dimension (geometric) |
| Integral domain | Irreducible variety |