Prime spectra and the Zariski topology
Treat primes as points, calculate algebraic closed sets and see why topology alone cannot retain nilpotent structure.
On this page
The question for today
What you’ll understand
- Calculate V(I), D(f), closures and specializations of prime points.
- Explain the correspondence between a basic open and a localization.
- Distinguish the topological spectrum from the ring information lost by removing nilpotents.
A prime ideal can be regarded as a point where the ring has a local algebra and a residue field. The resulting prime spectrum is usually unlike a Euclidean point cloud: some points are generic for whole irreducible pieces, and distinct points need not have disjoint neighborhoods. The Zariski topology is designed so that imposing equations gives closed sets and inverting a function gives an open set. It encodes how local algebraic viewpoints fit together, but its underlying topology cannot by itself distinguish a reduced ring from a nilpotent thickening.
Equations define closed sets of prime points
Spec R is the set of proper prime ideals of a commutative unital ring. For an ideal I, define V(I) as the primes containing I. Arbitrary intersections correspond to sums of ideals, and the union of two such closed sets corresponds to their product. The prime condition is exactly what makes that finite-union rule work: a prime containing IJ must contain I or J. Also V(0) is the whole spectrum and V(R) is empty. These properties establish a topology without assuming R Noetherian or a finitely generated algebra over a field.
The complement D(f) consists of primes not containing f, and these principal open sets form a basis. Prime correspondence for localization identifies Spec R_f with D(f), including its topology. This is the precise algebraic meaning of examining the region where f is invertible. An open set need not be the complement of one equation, but it is a union of principal opens. Ring homomorphisms induce continuous maps in the reverse direction by contracting primes; the reversal reflects how equations pull back along a map of geometric objects.
Generic points specialize to larger primes
The closure of a prime point p is V(p), so q lies in that closure exactly when p⊆q. Maximal ideals give closed points, whereas a smaller prime can be generic for a larger closed subset. In the spectrum of a domain, the zero ideal is a point whose closure is the entire space. A nonempty open set therefore contains that generic point. This is incompatible with the usual Hausdorff separation picture when the spectrum has more than one point, but it is natural for a topology expressing algebraic specialization.
For a finite-variable polynomial ring over an algebraically closed field, maximal ideals correspond to ordinary coordinate points by the Nullstellensatz. Without algebraic closedness, closed points can have larger residue fields, so restricting attention only to real coordinate evaluations can miss prime points. The spectrum also contains nonmaximal primes describing irreducible subvarieties or generic behavior. Those extra points are not numerical approximations to closed ones; their inclusion relations organize the structure used in dimension theory and local algebra.
Compactness and invisible nilpotents
Spec R is quasi-compact: every open cover has a finite subcover. Reduce to a cover by principal opens D(f_i). If their generators lay in a proper ideal, a maximal ideal containing that ideal would miss the whole cover. Hence one belongs to the ideal generated by the f_i, and its expression uses finitely many generators, giving a finite subcover. This argument uses neither Noetherianity nor a metric. Noetherian rings give the stronger Noetherian topological property, but quasi-compactness is already present for every ring spectrum.
Every prime contains every nilpotent element. Consequently R and its quotient by the nilradical have homeomorphic spectra. The dual numbers k[ε]/(ε²) and the field k each have a one-point underlying spectrum, although the first has a nonzero element squaring to zero. To retain this extra algebra one equips the spectrum with its local rings and structure sheaf, leading to affine schemes. This lesson concerns the underlying topological space; a one-point picture alone does not establish equality of the rings, their modules or their infinitesimal behavior.
Work it through
Closed and open sets in Spec Z
Calculate V(30), D(6), V(6)∩V(10) and V(6)∪V(10) in Spec Z. Include the generic point.
The prime ideals are (0) and (p) for prime integers p. The ideals containing 30 are (2),(3),(5), so V(30) is those three closed points and excludes (0).
D(6) contains (0) and every prime point except (2) and (3), because exactly those prime integers divide 6.
The intersection is V((6)+(10))=V(2)={(2)}. The union is V(60)={(2),(3),(5)}. The generic point belongs to neither finite closed set, while its own closure is all of Spec Z.
Algebraic ideal sums and products compute the topology directly. The point (0) is present even though it is not a closed point corresponding to a finite residue field.
The living notebook
Explore the model
Interactive lab · modular
Residues are not the prime points of a spectrum
Pause and Reset, then set modulus m=6 and additive step=1; predict six residue labels in the orbit. For a fresh comparison Reset and set m=6,step=2; predict the three labels 0,2,4. Spec(Z/6Z) instead has only two prime points, corresponding to 2 and 3. Use the orbit to list ring elements, then distinguish those elements from the ideals that form spectrum points; this instrument does not draw the Zariski space.
- Residue r0
- 0
- Orbit period · m/gcd(m,a)
- 12
- Distinct residues visited
- 1 of 12
Addition modulo m starts at zero: rₙ = na mod m. Playback advances two additions per second. The orbit has period m/gcd(m,a), including period one for a = 0. This is not a cryptographic security test or a primality algorithm. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.
Look forA spectrum’s points are prime ideals, not one point for every element or residue class.
Pause & reason
Check your understanding
Your turn
Try it yourself
Describe the underlying spectrum of A=k[ε]/(ε²), its unique residue field and its nilradical. Why does its one-point topology not identify A with k?
Take a moment to make a prediction or work it out on paper. Open a hint when you need one.
Give me a hint
Every prime must contain ε because ε²=0, then pass to the quotient by (ε).
Compare with the solution
Every prime contains ε, and A/(ε)=k has only the zero prime. Thus Spec A has one point, the maximal ideal (ε), with residue field k. Its nilradical is (ε), which is nonzero although its square is zero. The reduced quotient A/(ε) has the same one-point topology, but A still contains a nonzero nilpotent and is two-dimensional as a k-vector space. The ring structure, not the point set alone, retains this infinitesimal distinction.
Keep the idea
Take it with you
- Prime ideals form points, and ideal equations form Zariski closed sets.
- Point closure follows prime inclusion, with generic and closed points playing different roles.
- Removing nilpotents preserves spectrum topology but changes local algebra.
Read further
Original sources and deeper explanations to follow at your own pace.
Optional progress, saved in this browser when available. Revisit any lesson at any time.
Lesson not marked complete.