Commutative Algebra

Commutative algebra is the rigorous study of commutative rings and their ideals — the algebraic backbone that underlies both algebraic geometry and modern number theory. Where abstract algebra asks what rings and modules are, commutative algebra asks how they behave: how prime ideals chain together to measure dimension, how ideals decompose into irreducible constituents, and how localization lets us zoom in on the local behavior of a ring at a single point. The subject reaches from the classical ideal theory of Richard Dedekind and David Hilbert in the nineteenth century through the sweeping structural theorems of Emmy Noether and Wolfgang Krull in the twentieth, and today it forms the algebraic language in which schemes, moduli spaces, and arithmetic geometry are written.

Rings, Ideals, and Prime Spectra

A commutative ring $R$ is a set equipped with two binary operations — addition and multiplication — satisfying the usual axioms (associativity, commutativity, distributivity, and the existence of additive and multiplicative identities), with the additional requirement that $ab = ba$ for all $a, b \in R$. The most fundamental examples are the integers $\mathbb{Z}$, polynomial rings $k[x_1, \ldots, x_n]$ over a field $k$, and quotients of these by ideals. The commutativity hypothesis might seem mild, but it has profound consequences: it allows the theory of prime ideals to model geometric spaces in a way that non-commutative rings cannot.

An ideal $I \subseteq R$ is a subgroup of $(R, +)$ that is closed under multiplication by arbitrary ring elements: if $a \in I$ and $r \in R$, then $ra \in I$. Ideals play the role of “generalized divisors.” The sum of two ideals $I + J = \{a + b : a \in I, b \in J\}$ and their product $IJ$, generated by all products $ab$ with $a \in I$ and $b \in J$, are again ideals. The radical of an ideal is

$\sqrt{I} = \{r \in R : r^n \in I \text{ for some } n \geq 1\},$

which captures all elements that are “eventually in $I$.” An ideal is radical if it equals its own radical.

A prime ideal $\mathfrak{p} \subseteq R$ is a proper ideal with the property that $ab \in \mathfrak{p}$ implies $a \in \mathfrak{p}$ or $b \in \mathfrak{p}$ — precisely generalizing the notion of a prime number. Equivalently, $\mathfrak{p}$ is prime if and only if the quotient ring $R/\mathfrak{p}$ is an integral domain. A maximal ideal $\mathfrak{m}$ is a proper ideal not contained in any strictly larger proper ideal; the quotient $R/\mathfrak{m}$ is then a field. Every maximal ideal is prime, but not conversely.

The prime spectrum $\operatorname{Spec}(R)$ is the set of all prime ideals of $R$, equipped with the Zariski topology: the closed sets are the sets $V(I) = \{\mathfrak{p} \in \operatorname{Spec}(R) : I \subseteq \mathfrak{p}\}$ for ideals $I \subseteq R$. This construction, developed systematically by Alexander Grothendieck in the 1950s and 1960s, identifies commutative rings with the rings of functions on geometric spaces. A ring homomorphism $\varphi: R \to S$ induces a continuous map $\varphi^*: \operatorname{Spec}(S) \to \operatorname{Spec}(R)$ by $\mathfrak{q} \mapsto \varphi^{-1}(\mathfrak{q})$, making $\operatorname{Spec}$ a contravariant functor from commutative rings to topological spaces — the bridge between algebra and geometry.

The nilradical $\mathfrak{N}(R) = \sqrt{(0)}$ is the intersection of all prime ideals of $R$; it consists of all nilpotent elements. The Jacobson radical $\mathfrak{J}(R)$ is the intersection of all maximal ideals and captures the “unobservable” part of the ring. In a Noetherian ring (as we will see), these structures are finitely controlled.

Modules and Module Theory

A module $M$ over a ring $R$ is the natural generalization of a vector space: it is an abelian group $(M, +)$ with a scalar multiplication $R \times M \to M$ satisfying the obvious bilinearity and unit axioms. When $R$ is a field, modules and vector spaces coincide, but for a general ring the theory is far richer. Modules are the correct setting for studying linear algebra over rings and provide the language for all homological methods.

The essential building blocks are free modules: a module $M$ is free of rank $n$ if $M \cong R^n$ as $R$-modules, meaning it has a basis $\{e_1, \ldots, e_n\}$ such that every element of $M$ is uniquely a finite linear combination $\sum r_i e_i$. Over a principal ideal domain (PID) such as $\mathbb{Z}$ or $k[x]$, the structure theorem for finitely generated modules gives a complete classification: every such module decomposes as

$M \cong R^r \oplus R/(d_1) \oplus R/(d_2) \oplus \cdots \oplus R/(d_k)$

where $d_1 \mid d_2 \mid \cdots \mid d_k$ are the elementary divisors and $r$ is the free rank. Applied to $\mathbb{Z}$-modules this recovers the classification of finitely generated abelian groups; applied to $k[x]$-modules it gives the Jordan and rational canonical forms from linear algebra.

The tensor product $M \otimes_R N$ of two $R$-modules is the universal target for bilinear maps $M \times N \to P$: a map from $M \times N$ is bilinear if and only if it factors through a unique $R$-linear map from $M \otimes_R N$. Concretely, $M \otimes_R N$ is generated by symbols $m \otimes n$ subject to the relations $m \otimes (n_1 + n_2) = m \otimes n_1 + m \otimes n_2$ and $(rm) \otimes n = m \otimes (rn)$. An $R$-module $F$ is flat if the functor $- \otimes_R F$ is exact — it preserves short exact sequences. Free modules and projective modules are flat, and flatness is a key hypothesis in many theorems about base change and fiber products.

Exact sequences organize the relations between modules. A sequence $0 \to A \xrightarrow{f} B \xrightarrow{g} C \to 0$ is short exact if $f$ is injective, $g$ is surjective, and $\ker g = \operatorname{im} f$. Such sequences model “extension problems”: $C$ is a quotient of $B$ by a submodule isomorphic to $A$. The snake lemma and the five lemma are the essential diagram-chasing tools that relate different exact sequences, and they underpin the homological machinery developed in the later sections.

Localization and Local Rings

Localization is the algebraic analogue of restricting a function to an open set. Given a multiplicative subset $S \subseteq R$ — a subset closed under multiplication and containing $1$ — the localization $S^{-1}R$ is the ring of formal fractions $r/s$ with $r \in R$ and $s \in S$, subject to the equivalence $r/s = r'/s'$ whenever $t(rs' - r's) = 0$ for some $t \in S$. When $S = R \setminus \mathfrak{p}$ for a prime ideal $\mathfrak{p}$, the localization $R_\mathfrak{p} = (R \setminus \mathfrak{p})^{-1}R$ is called the localization at $\mathfrak{p}$. It has a unique maximal ideal $\mathfrak{p}R_\mathfrak{p}$, making it a local ring.

A ring is local if it has exactly one maximal ideal. Local rings are the algebraic models for the neighborhood of a point: in algebraic geometry, the local ring of an algebraic variety at a point $p$ captures everything about the variety in an infinitesimal neighborhood of $p$. The unique maximal ideal $\mathfrak{m}$ consists of “functions vanishing at $p$,” and the residue field $k(p) = R_\mathfrak{p}/\mathfrak{m}R_\mathfrak{p}$ is the field of values at the point.

Localization interacts beautifully with module theory. If $M$ is an $R$-module, the localization $S^{-1}M = S^{-1}R \otimes_R M$ is an $S^{-1}R$-module, and localization is an exact functor. A fundamental principle is that many properties of rings and modules can be checked locally: an $R$-module homomorphism $f: M \to N$ is injective (resp. surjective, zero) if and only if $f_\mathfrak{p}: M_\mathfrak{p} \to N_\mathfrak{p}$ is injective (resp. surjective, zero) for every prime $\mathfrak{p}$.

Discrete valuation rings (DVRs) are the simplest local rings beyond fields: a DVR is a PID with exactly one nonzero prime ideal $(\pi)$, and every nonzero element has a unique representation as $u \cdot \pi^n$ where $u$ is a unit and $n \geq 0$. The integer $n$ is the valuation $v(x)$. Classic examples include $\mathbb{Z}_{(p)}$ (the integers localized at the prime $p$) and the formal power series ring $k[[t]]$. DVRs model the local behavior of curves at smooth points, and their theory is the algebraic foundation for ramification in number fields.

Noetherian and Artinian Rings

A ring $R$ is Noetherian if every ascending chain of ideals $I_1 \subseteq I_2 \subseteq I_3 \subseteq \cdots$ eventually stabilizes — there exists $N$ such that $I_n = I_N$ for all $n \geq N$. This ascending chain condition (ACC) is equivalent to requiring that every ideal of $R$ is finitely generated. Named for Emmy Noether, who isolated this condition in her landmark 1921 paper Idealtheorie in Ringbereichen, the Noetherian property is the single most important finiteness hypothesis in commutative algebra.

The central result is Hilbert’s Basis Theorem (1888): if $R$ is Noetherian, then so is the polynomial ring $R[x]$, and by induction, so is $R[x_1, \ldots, x_n]$. This theorem, which David Hilbert proved by a nonconstructive existence argument that shocked his contemporaries (Paul Gordan famously called it “theology, not mathematics”), guarantees that every ideal in a polynomial ring over a Noetherian ring is finitely generated. It is the starting point for the entire theory of algebraic varieties and Gröbner bases.

An Artinian ring satisfies the dual descending chain condition (DCC): every descending chain of ideals $I_1 \supseteq I_2 \supseteq I_3 \supseteq \cdots$ stabilizes. Artinian rings are more constrained than Noetherian ones. The Hopkins-Levitzki theorem states that every Artinian ring is Noetherian, and the structure theorem for Artinian rings asserts that every Artinian ring is isomorphic to a finite product of Artinian local rings. In particular, a commutative Artinian ring has only finitely many prime ideals, all of which are maximal.

Krull’s intersection theorem is a fundamental finiteness result for Noetherian local rings: if $(R, \mathfrak{m})$ is a Noetherian local ring and $M$ is a finitely generated $R$-module, then

$\bigcap_{n=1}^\infty \mathfrak{m}^n M = 0.$

This says that elements lying in arbitrarily high powers of the maximal ideal must be zero — a precise algebraic statement of the intuition that “higher-order vanishing” forces vanishing outright. It underlies the theory of completions and formal power series.

Primary Decomposition and Unique Factorization

An ideal $Q \subseteq R$ is primary if $ab \in Q$ implies $a \in Q$ or $b^n \in Q$ for some $n \geq 1$. Equivalently, $Q$ is primary if every zero-divisor in $R/Q$ is nilpotent. The radical $\sqrt{Q}$ of a primary ideal is always prime, and we say $Q$ is $\mathfrak{p}$-primary when $\sqrt{Q} = \mathfrak{p}$.

The Lasker-Noether theorem (named for Emanuel Lasker, who proved it for polynomial rings in 1905, and Emmy Noether, who gave the general algebraic proof in 1921) states that in a Noetherian ring, every ideal $I$ has a primary decomposition:

$I = Q_1 \cap Q_2 \cap \cdots \cap Q_r$

where each $Q_i$ is a $\mathfrak{p}_i$-primary ideal. The primes $\mathfrak{p}_i = \sqrt{Q_i}$ are called the associated primes of $I$, and the set $\operatorname{Ass}(R/I) = \{\mathfrak{p}_1, \ldots, \mathfrak{p}_r\}$ is an invariant of $I$. While the individual primary components $Q_i$ can vary, the first uniqueness theorem asserts that the set of associated primes is unique. The minimal (or isolated) primes among the $\mathfrak{p}_i$ — those not containing any other $\mathfrak{p}_j$ — are uniquely determined, as are their primary components; the non-minimal (embedded) primes may have non-unique components.

Primary decomposition generalizes prime factorization: in $\mathbb{Z}$, the primary decomposition of the ideal $(n)$ recovers the prime factorization $n = p_1^{a_1} \cdots p_k^{a_k}$ via $(n) = (p_1^{a_1}) \cap \cdots \cap (p_k^{a_k})$. In a Dedekind domain (such as the ring of integers of a number field), every nonzero ideal factors uniquely into prime ideals — an exact analogue of the fundamental theorem of arithmetic, capturing the “ideal” factorization that Dedekind introduced in the 1870s to restore unique factorization in rings like $\mathbb{Z}[\sqrt{-5}]$ where it fails for elements.

Gauss’s lemma states that if $R$ is a unique factorization domain (UFD), then the polynomial ring $R[x]$ is also a UFD. This, combined with induction, shows that $k[x_1, \ldots, x_n]$ is a UFD for any field $k$ — a fact essential to algebraic geometry. The Eisenstein criterion provides a practical irreducibility test: if a prime $p \in R$ divides every coefficient of a polynomial $f = a_n x^n + \cdots + a_0$ except the leading one, and $p^2 \nmid a_0$, then $f$ is irreducible over the fraction field of $R$.

Integral Extensions and Dimension Theory

An element $\alpha$ in an extension ring $S \supseteq R$ is integral over $R$ if it satisfies a monic polynomial equation $\alpha^n + r_{n-1}\alpha^{n-1} + \cdots + r_0 = 0$ with coefficients $r_i \in R$. The set of all elements of $S$ integral over $R$ forms a subring called the integral closure of $R$ in $S$; when this equals $R$ itself, $R$ is integrally closed. Every UFD is integrally closed, and Dedekind domains — the rings of integers of number fields — are precisely the Noetherian integrally closed domains of Krull dimension one.

Integral extensions satisfy three fundamental theorems. The lying-over theorem guarantees that for an integral extension $R \hookrightarrow S$ and any prime $\mathfrak{p} \subseteq R$, there exists a prime $\mathfrak{q} \subseteq S$ with $\mathfrak{q} \cap R = \mathfrak{p}$. The going-up theorem states that if $\mathfrak{p} \subseteq \mathfrak{p}'$ are primes in $R$ and $\mathfrak{q}$ lies over $\mathfrak{p}$, then there exists $\mathfrak{q}' \supseteq \mathfrak{q}$ lying over $\mathfrak{p}'$. The going-down theorem, which requires $R$ to be integrally closed, gives the analogous result for descending chains. These theorems are essential for comparing the structure of a ring with that of its integral extensions.

The Krull dimension of a ring $R$ is the supremum of the lengths of chains of prime ideals:

$\mathfrak{p}_0 \subsetneq \mathfrak{p}_1 \subsetneq \cdots \subsetneq \mathfrak{p}_d.$

A field has dimension $0$, the ring $\mathbb{Z}$ has dimension $1$, the polynomial ring $k[x_1, \ldots, x_n]$ has dimension $n$, and a local Noetherian ring $(R, \mathfrak{m})$ has dimension equal to the minimum number of generators of an $\mathfrak{m}$-primary ideal — a deep result known as Krull’s principal ideal theorem implies more precisely that $\operatorname{ht}(\mathfrak{p}) \leq n$ whenever $\mathfrak{p}$ is a prime minimal over an ideal generated by $n$ elements.

Noether normalization is a fundamental structural result: every finitely generated $k$-algebra $A = k[x_1, \ldots, x_n]/I$ contains a polynomial subring $k[y_1, \ldots, y_d]$ such that $A$ is integral over it, where $d = \dim A$. This identifies the Krull dimension with the transcendence degree over $k$ and provides the algebraic foundation for Hilbert’s Nullstellensatz: if $I \subsetneq k[x_1, \ldots, x_n]$ is a proper ideal and $k$ is algebraically closed, then $V(I) \neq \emptyset$ (every proper ideal has a common zero). The strong Nullstellensatz further asserts $I(V(I)) = \sqrt{I}$, making the correspondence $I \mapsto V(I)$ between radical ideals and algebraic varieties a genuine bijection.

Homological Methods

Homological algebra entered commutative algebra through the work of David Auslander, Maurice Buchsbaum, and Jean-Pierre Serre in the 1950s, who showed that classical properties like regularity and smoothness have purely homological characterizations.

The projective dimension $\operatorname{pd}_R(M)$ of an $R$-module $M$ is the length of the shortest projective resolution

$0 \to P_n \to \cdots \to P_1 \to P_0 \to M \to 0.$

The injective dimension $\operatorname{id}_R(M)$ is defined dually. The global dimension of $R$ is $\sup\{\operatorname{pd}_R(M) : M \text{ an } R\text{-module}\}$.

The Ext and Tor groups are the derived functors of $\operatorname{Hom}$ and $\otimes$ respectively. Concretely, $\operatorname{Ext}^n_R(M, N)$ is computed by taking a projective resolution of $M$ and applying $\operatorname{Hom}_R(-, N)$, while $\operatorname{Tor}_n^R(M, N)$ is computed by taking a projective resolution of $M$ and applying $- \otimes_R N$. These groups measure the failure of $\operatorname{Hom}$ and $\otimes$ to be exact: $M$ is projective if and only if $\operatorname{Ext}^1_R(M, -) = 0$, and $M$ is flat if and only if $\operatorname{Tor}_1^R(M, -) = 0$.

The Auslander-Buchsbaum formula is one of the jewels of homological commutative algebra: for a finitely generated module $M$ of finite projective dimension over a Noetherian local ring $(R, \mathfrak{m})$,

$\operatorname{pd}_R(M) + \operatorname{depth}(M) = \operatorname{depth}(R),$

where depth measures the length of maximal regular sequences in $\mathfrak{m}$ acting on $M$. This formula has sweeping consequences: a Noetherian local ring $R$ has finite global dimension if and only if it is a regular local ring (one where $\dim R$ equals the minimal number of generators of $\mathfrak{m}$, equivalently where $\mathfrak{m}$ can be generated by a regular sequence). Serre’s characterization — $R$ is regular if and only if it has finite global dimension — gave the first algebraic proof that the localization of a regular local ring is again regular, resolving a question that had resisted all classical approaches.

The Koszul complex $K_\bullet(x_1, \ldots, x_n; R)$ associated to elements $x_1, \ldots, x_n \in R$ provides a canonical free resolution when the sequence is regular. Its homology groups $H_i(K_\bullet)$ vanish for $i > 0$ precisely when $x_1, \ldots, x_n$ is a regular sequence in $R$, giving an efficient homological criterion for regularity and providing the foundation for the theory of Cohen-Macaulay rings — those Noetherian local rings where $\operatorname{depth}(R) = \dim(R)$.

Computational Commutative Algebra

The algorithmic side of commutative algebra was transformed by Bruno Buchberger’s 1965 Ph.D. thesis, which introduced Gröbner bases and gave a complete algorithm for computing them.

A monomial order on $k[x_1, \ldots, x_n]$ is a total order on monomials that is compatible with multiplication and has $1$ as the minimum. The two most common are lexicographic order (lex) and graded reverse lexicographic order (grevlex). Given a monomial order, the leading term $\operatorname{LT}(f)$ of a nonzero polynomial is its largest monomial with coefficient. A Gröbner basis for an ideal $I$ is a finite generating set $\{g_1, \ldots, g_t\}$ such that the leading terms of the $g_i$ generate the ideal of all leading terms of elements of $I$:

$\langle \operatorname{LT}(g_1), \ldots, \operatorname{LT}(g_t) \rangle = \langle \operatorname{LT}(f) : f \in I \rangle.$

Gröbner bases solve the ideal membership problem: $f \in I$ if and only if the remainder of $f$ upon division by the Gröbner basis is zero. They also allow computation of intersections, quotients, and syzygies of ideals, and they give algorithms for elimination: projecting a variety by eliminating variables. The elimination theory built from Gröbner bases provides a complete algorithmic framework for solving polynomial systems, computing dimension and degree, and performing primary decomposition.

Buchberger’s algorithm computes a Gröbner basis by repeatedly computing S-polynomials $S(f, g) = \frac{L}{\operatorname{LT}(f)} f - \frac{L}{\operatorname{LT}(g)} g$ (where $L = \operatorname{lcm}(\operatorname{LT}(f), \operatorname{LT}(g))$) and reducing them modulo the current basis. The algorithm terminates because the ascending chain condition on the ideal of leading terms must stabilize in $k[x_1, \ldots, x_n]$ (by Hilbert’s basis theorem). The resulting basis depends on the chosen monomial order, but the ideal it generates does not.

The Hilbert function $H_I(d) = \dim_k (k[x_1, \ldots, x_n]/I)_d$ counts the number of independent polynomials of degree $d$ in the quotient ring. For large $d$, $H_I(d)$ agrees with a polynomial — the Hilbert polynomial — whose degree is the dimension of the variety $V(I)$ and whose leading coefficient encodes its degree. Computing the Hilbert function from a Gröbner basis is mechanical, and the result gives intrinsic geometric invariants from purely algebraic data.

Systems such as Macaulay2 (developed by Daniel Grayson and Michael Stillman), Singular, and CoCoA implement these algorithms and have become essential tools for research in algebraic geometry, commutative algebra, and related fields. They can perform primary decomposition, compute syzygies and free resolutions, check Cohen-Macaulay and Gorenstein properties, and verify membership in thousands of ideals — computations that would be infeasible by hand. The interplay between theoretical developments and computational experiments has been especially productive in combinatorial commutative algebra, where monomial ideals associated to graphs and simplicial complexes encode topological and combinatorial data in a form that computer algebra can explore.