Loading field map
Noncommutative algebra is the study of algebraic structures in which the order of multiplication matters. In the rings and algebras most familiar from elementary mathematics—the integers, the rational numbers, the real numbers—multiplication is commutative: \(ab = ba\) for all elements \(a\) and \(b\). Noncommutative algebra begins with the observation that many natural and useful algebraic systems fail this rule, and that their failure is not a defect but a rich source of structure.
The subject is not a single theory but a family of interconnected approaches to a common problem: how to understand rings and modules when the familiar commutative tools—ideals as kernels, prime ideals, localization, and the geometric intuition of points on a variety—no longer apply directly. Its development has been driven by internal questions about the structure of rings, by examples arising from geometry and physics, and by the need to understand linear transformations, which are inherently noncommutative.
A ring is a set equipped with two operations, addition and multiplication, satisfying the usual axioms: addition forms an abelian group, multiplication is associative and distributes over addition, and there is a multiplicative identity. A ring homomorphism is a map preserving both operations. An algebra over a field \(k\) is a ring that is also a vector space over \(k\), with scalar multiplication compatible with ring multiplication. The field \(k\) is often the complex numbers, the real numbers, or a finite field.
The central object of study is the module: an abelian group on which a ring acts linearly, generalizing the notion of a vector space over a field. For a commutative ring, modules are the algebraic analogue of vector bundles or sheaves on a geometric space. For a noncommutative ring, modules play an even more fundamental role, because the ring itself can be studied through its modules. The ring \(R\) acts on itself by left multiplication, giving the left regular module; its submodules are the left ideals. The structure of these submodules—how the ring decomposes into indecomposable pieces, how many non-isomorphic simple modules exist, and how modules fit together—constitutes much of the subject.
The central question of noncommutative algebra can be stated simply: What are the possible structures of rings and their modules, and how can they be classified? This question fragments into many more specific ones. When does a ring decompose as a direct product of simpler rings? When does every module decompose into indecomposable summands, and is that decomposition unique? What are the irreducible representations of a ring—that is, its simple modules—and how do arbitrary modules build up from them? How do the answers change when the ring is required to satisfy finiteness conditions, such as being finite-dimensional over a field or being Noetherian (every ascending chain of ideals stabilizes)?
The earliest noncommutative structures arose not from abstract algebra but from concrete problems. Matrix rings over a field—the set of \(n \times n\) matrices with entries in a field, under matrix addition and multiplication—are noncommutative for \(n \geq 2\). They were studied throughout the nineteenth century in the context of linear algebra and the theory of determinants, but they were not initially conceived of as rings. The first explicit noncommutative algebra in the modern sense was the quaternions, introduced by William Rowan Hamilton in 1843. Hamilton sought a system of numbers in three dimensions analogous to the complex numbers in two; he found instead a four-dimensional system with basis \(1, i, j, k\) satisfying \(i^2 = j^2 = k^2 = ijk = -1\), which forces \(ij = -ji\). The quaternions were controversial at first—many mathematicians doubted that a "number" system could violate commutativity—but they proved useful in physics and geometry.
The abstract theory of rings emerged around 1900, primarily through the work of Emmy Noether and Emil Artin in the 1920s. Noether's axiomatic approach reframed the subject: instead of studying specific examples like matrix rings or quaternions, she studied the general properties of rings and their ideals. Artin, building on work of Joseph Wedderburn, proved the first major structural theorem of the subject: the Wedderburn–Artin theorem, which classifies rings that satisfy a descending chain condition on left ideals (the Artinian condition). The theorem states that a semisimple Artinian ring—one whose left regular module is a direct sum of simple modules—is isomorphic to a finite direct product of matrix rings over division rings. This result remains the cornerstone of the subject: it shows that the "nice" rings are exactly the ones built from matrix rings over division rings, and it reduces their study to the study of division rings.
The development of noncommutative algebra after the 1930s can be organized around three broad approaches, each addressing a different aspect of the central question. These are not mutually exclusive schools but rather complementary research programmes that often overlap and borrow from one another.
The first approach, which dominated the subject from the 1920s through the 1960s, asks directly: What are the possible structures of a ring? Its methods are internal: one studies the lattice of ideals, the existence of prime and maximal ideals, and the ways a ring can be built from simpler pieces.
The Wedderburn–Artin theorem settled the structure of semisimple Artinian rings. The next major step was the theory of the Jacobson radical, developed by Nathan Jacobson in the 1940s. The Jacobson radical \(J(R)\) is the intersection of all maximal left ideals of \(R\); it is a two-sided ideal that captures the "obstruction" to semisimplicity. The quotient \(R/J(R)\) is always semisimple in a weakened sense, and many questions about \(R\) reduce to questions about this quotient and about the radical itself. A ring is semiprimitive if its Jacobson radical is zero; such rings are exactly those that have a faithful family of simple modules, and they can be studied through their representations.
A parallel line of development concerned prime rings and primitive rings, generalizing the role of prime ideals in commutative algebra. A ring is prime if the product of any two nonzero two-sided ideals is nonzero; it is left primitive if it has a faithful simple left module. The structure theory of prime rings was advanced by Jacobson and by Israel Herstein, who proved that certain conditions on a prime ring force it to be commutative or nearly so. Herstein's work on derivations and identities—polynomial equations satisfied by all elements of a ring—showed that prime rings satisfying a polynomial identity are closely related to matrix rings over commutative rings.
The structure-theoretic approach reached a high degree of sophistication in the theory of Noetherian rings, developed extensively from the 1960s onward. A Noetherian ring is one in which every ascending chain of left ideals stabilizes; this finiteness condition is satisfied by many natural examples, including polynomial rings in finitely many variables over a field and their quotients, and by group algebras of polycyclic-by-finite groups. The theory of Noetherian rings, developed by Paul Cohn, Jacques Dixmier, and especially by the school around J. T. Stafford and J. C. McConnell, produced a deep structural picture: Noetherian rings have a well-behaved theory of prime ideals, a notion of localization at certain prime ideals, and a dimension theory analogous to the Krull dimension of commutative rings. The Goldie theorem, proved by A. W. Goldie in the 1950s, is a landmark: it characterizes the Noetherian rings that have a semisimple Artinian ring of fractions, showing that they are exactly the semiprime Goldie rings.
The second approach shifts attention from the ring itself to its modules. This perspective, which became dominant in the 1970s and 1980s, asks: What do the representations of a ring look like? The methods are external: one studies the category of modules, its indecomposable objects, and the ways modules can be built from one another.
The modern form of this approach began with the work of Maurice Auslander and Idun Reiten in the 1970s. They developed Auslander–Reiten theory, which studies the category of finitely generated modules over an Artinian algebra through its almost split sequences—short exact sequences that are minimal in a precise sense and that organize the way indecomposable modules are linked. The Auslander–Reiten quiver is a directed graph whose vertices are the indecomposable modules and whose arrows record the irreducible maps between them; it provides a combinatorial skeleton for the entire module category.
A central distinction in this approach is between finite representation type (only finitely many indecomposable modules up to isomorphism), tame representation type (the indecomposables can be classified by a finite number of one-parameter families), and wild representation type (the classification problem contains the classification of modules over the free algebra in two variables, and is therefore considered hopeless). This trichotomy, conjectured by Donovan–Freislich and proved by Drozd in the 1970s, is one of the deepest results of the subject. It shows that the representation theory of a finite-dimensional algebra is either completely understood, understood up to a manageable family, or essentially impossible in a precise sense.
The module-theoretic approach also produced the theory of tilting theory, developed by Brenner–Butler and Happel–Ringel in the 1980s. A tilting module is a module whose derived category is equivalent to that of another algebra; tilting theory shows that many different algebras have equivalent derived categories, so that their module categories are "the same" at the level of homological algebra. This led to the notion of derived equivalence and to the modern understanding that the relevant invariant of an algebra is not its module category alone but its derived category.
The third approach, which emerged from the 1950s and became increasingly central, treats noncommutative algebra through the lens of homological algebra. The key idea is to study a ring \(R\) through the derived functors \(\mathrm{Ext}\) and \(\mathrm{Tor}\), which measure the failure of modules to behave like vector spaces. A module is projective if it is a direct summand of a free module; it is injective if it has the extension property dual to projectivity. The projective dimension of a module is the length of the shortest projective resolution; the global dimension of a ring is the supremum of the projective dimensions of all modules.
This approach was pioneered by Henri Cartan and Samuel Eilenberg in their 1956 book Homological Algebra, and developed further by Maurice Auslander, who introduced the notion of the Auslander–Gorenstein condition and the Auslander transpose. The homological approach provides invariants that are invisible to the structure theory: two rings can have the same lattice of ideals but different global dimensions, and the global dimension often reflects geometric properties of the space on which the ring acts.
The categorical turn reached its fullest expression in the theory of derived categories, introduced by Jean-Louis Verdier in the 1960s and popularized in the 1980s. The derived category of a ring \(R\) is obtained from the category of complexes of modules by formally inverting quasi-isomorphisms; it is the natural setting for homological algebra, because it makes the derived functors \(\mathrm{Ext}\) and \(\mathrm{Tor}\) into ordinary Hom and tensor functors. The derived category is the object of study in derived noncommutative geometry, which treats a ring as a "space" whose points are invisible but whose sheaves are its modules.
These three approaches are not rivals in the sense of competing for the same territory; they are complementary lenses on the same objects. The structure theory provides the basic vocabulary—prime ideals, radicals, Goldie rings—that the module-theoretic and homological approaches use as input. The module-theoretic approach, in turn, often reveals structure that the ring-theoretic view misses: two rings can be Morita equivalent (have equivalent module categories) without being isomorphic, and Morita equivalence is invisible to the internal structure theory. The homological approach provides the deepest invariants, but it requires the ring to be reasonably well-behaved—typically Noetherian or Artinian—to yield meaningful results.
A concrete example illustrates the interplay. The Weyl algebra \(An(k)\) is the algebra generated by \(2n\) symbols \(x1, \dots, xn, \partial1, \dots, \partialn\) subject to the relations \(\partiali xj - xj \partiali = \delta{ij}\) and all other commutators zero. It is the algebra of differential operators with polynomial coefficients. The structure theory shows that \(An(k)\) is a simple Noetherian domain (it has no nontrivial two-sided ideals) and that it has a skew field of fractions. The module theory shows that its global dimension is \(n\), and that its simple modules are classified only in low dimensions. The homological approach shows that \(An(k)\) is not isomorphic to its opposite ring for \(n \geq 1\), a fact with deep consequences for the geometry of the space on which it acts. Each approach contributes a different piece of the picture, and none alone suffices.
Since the 1980s, the subject has been reshaped by the influence of noncommutative geometry, a programme initiated by Alain Connes. Connes's idea is that a geometric space can be studied through the algebra of functions on it, and that when the space is "badly behaved"—for example, a quotient of a manifold by a non-free group action—the algebra of functions becomes noncommutative. The resulting noncommutative geometry treats a noncommutative algebra as the algebra of functions on a noncommutative space, and it imports tools from differential geometry, topology, and analysis. This programme has produced the theory of spectral triples, which encode geometric information in a Dirac operator, and the cyclic cohomology of Connes, which provides the analogue of de Rham cohomology for noncommutative spaces.
A parallel development, sometimes called noncommutative algebraic geometry, adapts the tools of algebraic geometry to noncommutative rings. The key idea, due to Artin, Michael Artin, and J. T. Stafford, is that a noncommutative graded algebra can be studied through its category of graded modules, which plays the role of the category of sheaves on a projective variety. The Artin–Schelter regular algebras are the noncommutative analogues of polynomial rings; they are graded algebras with finite global dimension and polynomial growth, and they are classified in low dimensions. These algebras are the building blocks of noncommutative projective geometry, and their study has revealed a rich landscape of "noncommutative spaces" that have no commutative counterpart.
The modern subject also retains strong connections to its historical roots. Group algebras \(k[G]\)—the algebra of finite linear combinations of elements of a group \(G\) with coefficients in \(k\)—remain a central object of study, connecting noncommutative algebra to representation theory of finite groups and to the theory of von Neumann algebras. Enveloping algebras of Lie algebras, which include the Weyl algebra as a special case, connect the subject to Lie theory and to the theory of differential equations. Quantum groups, introduced by Vladimir Drinfeld and Michio Jimbo in the 1980s, are deformations of enveloping algebras that arise in mathematical physics; they have become a major industry in their own right, with deep connections to knot theory, statistical mechanics, and the theory of quantum integrable systems.
Despite its maturity, noncommutative algebra retains a number of central open problems. The classification of simple Noetherian rings is far from complete; even the classification of simple finitely generated algebras over the complex numbers is unknown in general. The Goldie rank of a simple Noetherian ring—the number of summands in a decomposition of the ring as a module over its center—is known to be a subtle invariant, and its behavior under extension of scalars is not fully understood. The Zelmanov problem on the structure of prime Jordan algebras, and the related questions about the structure of division rings, remain open in important cases.
The most famous open problem in the subject is the Gelfand–Kirillov conjecture, which asks whether the skew field of fractions of an enveloping algebra of a nilpotent Lie algebra is isomorphic to a Weyl field. The conjecture is known to hold in many cases but fails in general; the precise boundary between the cases where it holds and where it fails is not understood.
A more recent open direction concerns the classification of finite-dimensional algebras of wild representation type. The trichotomy of finite/tame/wild is known, but the internal structure of the wild case is largely unexplored: there is no good notion of "how wild" a wild algebra is, and no satisfactory way to organize the uncountably many indecomposable modules that such algebras possess.
Noncommutative algebra is often described as a collection of techniques rather than a single theory, and there is truth to this description. The subject draws on ring theory, module theory, homological algebra, category theory, and increasingly on geometry and analysis. Yet there is a unifying thread: the attempt to understand what happens when the commutative law fails, and to find the right language for describing the structures that arise. The Wedderburn–Artin theorem, the Jacobson radical, the Auslander–Reiten theory, and the derived category are all answers to the same question—what is a ring?—asked with increasing sophistication and from increasingly external perspectives.
The subject's history shows a pattern of successive broadening: from the internal structure theory of the 1920s–1960s, to the module-theoretic and homological approaches of the 1970s–1980s, to the geometric and categorical perspectives of the modern era. Each broadening has not replaced its predecessors but has absorbed them, providing new tools and new questions while retaining the old results as special cases. The result is a field that is both deeply technical and broadly connected, with a rich internal structure and a wide range of applications to other parts of mathematics and to physics.