1 Foundations
Homological algebra studies algebraic objects through sequences of maps, kernels, cokernels, and invariants built from them. Its central idea is that many constructions become clearer when organized into exact sequences and chain complexes, where the failure of exactness is measured by homology or cohomology.
The subject developed from topological methods but now serves as a general language for comparing modules, rings, sheaves, and other structures. It is especially effective when a problem has a natural notion of “relations among relations,” which can be encoded in higher-order algebraic data.
1.1 Algebraic motivation
A basic motivation is that many algebraic constructions are not exact: they do not preserve kernels, images, or quotient behavior perfectly. Homological algebra provides systematic tools for measuring these defects and organizing them into computable invariants.
This perspective is useful in problems where one wishes to replace a complicated object by simpler ones that still retain essential information. Resolutions, exact sequences, and derived functors all arise from this desire to study objects indirectly but effectively.
1.2 Historical development
The field emerged in the first half of the 20th century from algebraic topology, where invariants were needed to distinguish spaces up to continuous deformation. Ideas of cycles, boundaries, and exactness were abstracted and gradually extended to algebraic settings.
Later developments in category theory and module theory gave the subject a broader conceptual framework. By the mid-20th century, homological algebra had become a standard toolkit in topology, algebraic geometry, and representation theory.
1.3 Basic categorical language
Homological algebra is most naturally expressed using category theory. Categories describe collections of objects and morphisms, while functorial constructions explain how algebraic information is transported between settings.
1.3.1 Categories and functors
A category consists of objects, morphisms between objects, and a composition law satisfying associativity and identity axioms. In homological algebra, typical examples include the category of abelian groups, modules over a ring, or sheaves on a space.
A functor maps objects and morphisms from one category to another while preserving composition and identities. Many central constructions, such as kernels, tensor products, and homology groups, are best understood as functorial assignments.
1.3.2 Natural transformations
A natural transformation compares two functors in a way that is compatible with morphisms in the source category. This compatibility expresses that the comparison is not accidental but coherent across all objects.
Natural transformations are important because they formalize when two constructions are essentially the same from a categorical point of view. In homological algebra, they appear in the study of derived functors, comparison maps, and long exact sequences.
1.3.3 Abelian categories
An abelian category is a setting in which kernels and cokernels exist and behave much like they do for modules or abelian groups. It provides enough structure to define exactness, homology, and derived functors abstractly.
Many familiar algebraic categories are abelian, and this abstraction allows results to be stated once and applied broadly. It is one of the main reasons homological algebra can move fluidly between different mathematical contexts.
1.4 Exactness
Exactness describes how images and kernels fit together in a sequence of morphisms. It is the condition that makes a sequence behave like a precise algebraic record of relations and constraints.
1.4.1 Exact sequences
A sequence is exact at a term when the image of the preceding map equals the kernel of the following map. Exact sequences encode the idea that every relation arises from a previous one and that no information is lost or duplicated at the exact spot.
They are among the most basic tools in the subject. Many structural theorems are proved by embedding an object into an exact sequence and analyzing the surrounding terms.
1.4.2 Split exact sequences
A split exact sequence is one that decomposes into a direct sum in a particularly simple way. Such sequences are the algebraic analogue of situations in which an object can be cleanly separated into independent parts.
Split exactness is stronger than exactness alone. It often signals that the sequence is especially tractable and that the middle object can be reconstructed from the ends without hidden extension data.
1.4.3 Short and long exact sequences
A short exact sequence has three terms and captures a single extension problem. It is a compact way to express that one object sits inside another with a well-defined quotient.
Long exact sequences arise when a short exact sequence of complexes, pairs, or other structures is passed through a homological construction. They are indispensable because they connect multiple homology or cohomology groups in a coherent chain of relations.
2 Chain complexes
Chain complexes are the basic objects from which homology is built. They consist of a graded sequence of algebraic objects connected by maps whose successive compositions vanish.
2.1 Definitions
A chain complex is a sequence of objects and differentials, typically written so that each differential maps one degree to the next and the composition of two consecutive differentials is zero. This zero-composition property ensures that boundaries are always cycles.
Chain complexes may be finite or infinite, bounded or unbounded. Their graded structure allows one to track information degree by degree, which is essential for defining homological invariants.
2.2 Chain maps
A chain map is a morphism between chain complexes that commutes with the differentials. It preserves the complex structure and therefore induces maps on the associated homology groups.
Chain maps provide the correct notion of morphism in this context. They let one compare resolutions, build functorial constructions, and transfer information between different algebraic models.
2.3 Homotopy of chain maps
Two chain maps are chain homotopic if their difference can be expressed through a controlled intermediate family of maps. Chain homotopy is an equivalence relation that identifies maps inducing the same behavior at the level of homology.
This notion is central because it allows the study of complexes up to a flexible notion of equivalence. Homotopy invariance is one of the reasons homology is such a stable invariant.
2.4 Chain complexes in an abelian category
Chain complexes can be defined not only for modules but in any abelian category. This broad setting keeps the formalism intact while allowing applications to sheaves, representations, and other structured objects.
Once complexes are available in an abelian category, one can define cycles, boundaries, and homology objects in a purely categorical way. This abstraction is a major step toward derived categories and modern homological methods.
2.5 Mapping cones
The mapping cone of a chain map is a new complex that measures how far the map is from being a quasi-isomorphism. It packages the source, target, and connecting data into a single construction.
Mapping cones are especially useful in triangulated settings, where they help define distinguished triangles. They also serve as a concrete model for homotopical and derived behavior.
3 Homology and cohomology
Homology and cohomology are the primary invariants produced from chain complexes. They detect algebraic or topological features that survive after boundaries are identified as trivial.
3.1 Homology groups
Homology groups are obtained by taking cycles modulo boundaries in each degree. They measure the extent to which a complex fails to be exact.
These groups often capture hidden structure, such as generators, relations, and obstructions. In topology they record holes and voids; in algebra they reflect nontrivial syzygies and extension phenomena.
3.2 Cohomology groups
Cohomology is dual in spirit to homology and is typically built from cochain complexes. It is often better suited for encoding operations, products, and duality principles.
Cohomology groups frequently carry richer algebraic structure than homology groups alone. They play a central role in deformation theory, classification problems, and geometric applications.
3.3 Reduced homology
Reduced homology is a variant designed to simplify behavior in low degrees, especially for connected spaces. It adjusts ordinary homology so that a point has trivial reduced homology.
This modification makes many statements cleaner and more uniform. It is particularly helpful in formulations where the presence of a basepoint or connectedness should not introduce exceptional cases.
3.4 Relative homology
Relative homology compares a space or object with a subobject. It measures what is present in the larger object beyond what is already accounted for by the smaller one.
Relative groups are natural whenever one studies inclusions, pairs, or extension problems. They also lead to long exact sequences that relate absolute and relative information.
3.5 Connecting homomorphisms
Connecting homomorphisms arise in long exact sequences and link one degree of homology or cohomology to the next. They are the maps that carry boundary information across a short exact sequence of complexes or a pair of objects.
These homomorphisms are crucial because they encode how local data propagates into global invariants. They often provide the mechanism by which exact sequences become computational tools.
4 Derived functors
Derived functors extend ordinary functors to a homological setting so that failures of exactness can be measured systematically. They are one of the main achievements of the subject.
4.1 Motivation and basic idea
Many functors are only partially exact, preserving some exact sequences but not others. Derived functors correct this by replacing an object with a resolution and then applying the functor in a controlled way.
The resulting invariants capture higher-order obstruction data. They explain how a functor behaves beyond the simplest exact cases and often organize previously isolated computations into a coherent theory.
4.2 Right derived functors
Right derived functors are associated with left exact functors and are typically constructed using injective resolutions. They measure the extent to which a functor fails to preserve surjections or more generally right exact behavior.
4.2.1 Injective resolutions
An injective resolution replaces an object by a complex of injective objects that is exact except at the starting term. This makes it possible to compute right derived functors in a stable and often practical way.
Injective resolutions are especially valuable because injective objects have strong lifting properties. Those properties make calculations and comparison arguments significantly easier.
4.2.2 Ext functors
Ext functors are among the most important right derived functors. They classify extensions and measure higher extension data between objects such as modules or sheaves.
The lowest Ext group often describes equivalence classes of short exact sequences, while higher Ext groups encode more subtle compatibility conditions. These functors appear throughout algebra, geometry, and representation theory.
4.3 Left derived functors
Left derived functors are associated with right exact functors and are usually built from projective resolutions. They record the failure of exactness on the left side.
4.3.1 Projective resolutions
A projective resolution expresses an object as the homology of a complex built from projective objects. Such resolutions are designed so that applying a right exact functor produces meaningful higher information.
Projective resolutions are often easier to construct in module categories than in more general settings. They provide a concrete route to computations of Tor and related invariants.
4.3.2 Tor functors
Tor functors are the standard left derived functors of tensor product. They measure how tensoring interacts with exactness and detect hidden torsion phenomena.
These groups are fundamental in commutative algebra and algebraic topology. They often reveal whether two objects interact cleanly or whether their interaction produces higher obstruction terms.
4.4 Universality and effaceable functors
Derived functors are characterized by universal properties that distinguish them among possible extensions of a functor. These properties ensure that the construction is canonical rather than dependent on arbitrary choices.
The notion of effaceability helps explain why certain resolutions yield the correct derived objects. It formalizes the idea that higher obstructions can be made invisible after passing to suitable replacements.
5 Resolutions and constructions
Resolutions are the practical machinery behind derived methods. They replace an object by a more tractable complex that retains the information needed for homological analysis.
5.1 Projective resolutions
Projective resolutions are built from projective objects and are tailored for left derived functors. They are often used to compute Tor and to analyze modules through free or projective approximations.
A good projective resolution simplifies computations while remaining faithful to the original object. Its existence depends on the ambient category, and when available it becomes a powerful computational device.
5.2 Injective resolutions
Injective resolutions are dual to projective ones and support right derived functors. They are particularly effective for cohomological calculations.
Because injective objects admit extension of morphisms from subobjects, they provide a flexible environment for encoding global information. This makes them central to the theory of Ext and sheaf cohomology.
5.3 Flat resolutions
Flat resolutions use flat objects, which preserve exactness after tensoring with arbitrary modules. They are especially useful when projective resolutions are inconvenient or unavailable.
Flatness gives a weaker but often sufficient condition for homological computations. In many algebraic settings, flat resolutions provide a practical bridge between abstract theory and explicit calculation.
5.4 Free resolutions
Free resolutions are resolutions built from free objects, usually in module categories. They are among the most concrete and accessible tools in the subject.
Because free modules are easy to manipulate, free resolutions are widely used in computations of homology, syzygies, and algebraic invariants. They often serve as the starting point for explicit examples.
5.5 Bar resolutions
The bar resolution is a standard construction in homological algebra and algebraic topology. It provides an explicit free resolution with a combinatorial form that is useful for calculations.
Bar resolutions appear in group homology, algebra extensions, and other contexts where iterated multiplication or composition is present. Their explicit nature makes them a bridge between abstract definitions and concrete formulas.
6 Spectral sequences
Spectral sequences are computational devices that organize complicated homological data into successive approximations. They are especially useful when a problem has a filtration or multi-stage structure.
6.1 Definition and filtration
A spectral sequence is a sequence of pages, each carrying graded objects and differentials, that approximates a target homology or cohomology theory. It often arises from a filtered complex or another layered construction.
Filtration is the mechanism that makes the approximation possible. By analyzing successive quotients, one can break a hard computation into manageable stages.
6.2 Convergence
Convergence describes whether and how the successive pages of a spectral sequence recover the intended invariant. It is a crucial issue because the early pages may only represent partial information.
When convergence holds well, the spectral sequence becomes a reliable computational tool. Understanding convergence conditions is often as important as constructing the sequence itself.
6.3 Spectral sequences from filtered complexes
A filtered complex gives rise naturally to a spectral sequence by examining the associated graded pieces. This method is one of the most common sources of spectral sequences.
Such sequences help track how local or layerwise data assemble into global homological information. They are frequently used in topology, geometry, and algebraic calculations.
6.4 Spectral sequences from exact couples
Exact couples provide another standard mechanism for producing spectral sequences. They encode a recursive exactness pattern from which successive approximations emerge.
This approach is particularly well suited to long exact sequences and recursive structures. It organizes boundary maps and connecting homomorphisms into a systematic iterative process.
6.5 Applications
Spectral sequences are used to compare homology theories, compute invariants of fiber bundles, and analyze filtered algebraic objects. They are also valuable in settings where direct calculation would be unwieldy.
Although technically demanding, spectral sequences often turn impossible-looking problems into finite algebraic steps. Their strength lies in revealing structure that is otherwise hidden by complexity.
7 Homological dimensions
Homological dimensions quantify how complicated objects are from a homological standpoint. They measure the length of the shortest possible resolutions of a given type.
7.1 Projective dimension
Projective dimension is the minimal length of a projective resolution of an object. A small projective dimension indicates that the object is close to being projective.
This invariant is important because it controls many algebraic properties and often governs the vanishing of derived functors. It is a standard measure of complexity in module theory.
7.2 Injective dimension
Injective dimension is the dual notion, defined using injective resolutions. It records how far an object is from being injective.
It plays a major role in cohomological algebra and duality theory. Objects of finite injective dimension often have particularly well-behaved cohomological properties.
7.3 Flat dimension
Flat dimension measures the minimal length of a flat resolution. It is useful in contexts where tensor products and exactness after tensoring are central.
Flat dimension is especially relevant in commutative algebra and algebraic geometry. It helps identify objects whose tensor behavior is controlled and predictable.
7.4 Global dimension
Global dimension is a ring-theoretic invariant defined as the supremum of projective dimensions of all modules over the ring. It gives a broad measure of the ring’s homological complexity.
Rings of finite global dimension often admit particularly orderly homological behavior. This invariant connects local module properties with the overall structure of the ring.
7.5 Regularity conditions
Regularity conditions describe situations in which homological dimensions are small or well controlled. They often signal that an algebraic object behaves as if it were locally simple.
Such conditions are important because they link geometric intuition with algebraic exactness properties. In commutative algebra, they frequently characterize especially well-behaved rings and modules.
8 Derived categories
Derived categories provide a framework in which complexes are studied up to homological equivalence. They refine the language of homological algebra by making quasi-isomorphisms invertible.
8.1 Localization of homotopy categories
The derived category is obtained by localizing a homotopy category at quasi-isomorphisms. This construction identifies complexes that have the same homology, even if they are not isomorphic as complexes.
Localization is a powerful idea because it focuses attention on the information that matters homologically. It eliminates irrelevant distinctions while preserving the essential structure of the theory.
8.2 Triangulated categories
Triangulated categories are the natural ambient setting for derived categories. They come equipped with a shift functor and a class of triangles that abstract the behavior of exact sequences of complexes.
This structure captures how objects are related by mapping cones and extensions. It provides a flexible language for formulating homological arguments at a higher level of abstraction.
8.3 Distinguished triangles
A distinguished triangle is the triangulated analogue of a short exact sequence. It usually arises from a morphism and its mapping cone.
Distinguished triangles encode the basic relations among complexes in the derived category. They are central to many formal arguments because they retain enough exactness to support homological reasoning.
8.4 Derived functors in derived categories
In derived categories, derived functors can often be defined and computed more cleanly than at the level of individual complexes. The categorical framework makes their functorial properties transparent.
This setting unifies left and right derived constructions and clarifies how resolutions work conceptually. It is one of the main reasons derived categories became standard in modern algebraic geometry and representation theory.
8.5 t-structures
A t-structure is additional structure on a triangulated category that recovers an abelian category of “heart” objects. It organizes a derived category into nonnegative and nonpositive parts.
t-Structures are valuable because they bridge derived and classical viewpoints. They make it possible to extract ordinary algebraic objects from a derived environment in a controlled way.
9 Applications
Homological algebra has wide-ranging applications because many mathematical problems reduce to understanding exactness, extensions, and derived invariants. Its methods are especially effective in settings with layered or relational structure.
9.1 Algebraic topology
In algebraic topology, homological algebra provides the language for homology and cohomology theories. It is used to study spaces through algebraic invariants that are stable under deformation.
Exact sequences, spectral sequences, and derived constructions are indispensable in this area. They allow topological data to be computed and compared systematically.
9.2 Algebraic geometry
Algebraic geometry uses homological methods to study sheaves, cohomology, and geometric properties of varieties and schemes. Many important geometric questions are translated into statements about complexes and derived functors.
Derived techniques help organize local-to-global principles and clarify how geometric objects glue together. They are now a standard part of modern algebraic geometry.
9.3 Commutative algebra
Commutative algebra employs homological algebra to analyze modules over rings, ideals, and resolution-based invariants. Tor, Ext, and homological dimensions are central tools in this field.
These methods reveal subtle information about rings and modules that is not visible from generators and relations alone. They are especially useful in studying depth, regularity, and singular behavior.
9.4 Representation theory
Representation theory uses homological algebra to study modules over algebras and the relations among representations. Extensions and derived categories help organize complex families of representations.
Homological invariants often detect how representations can be built from simpler ones. They also support classification problems and the comparison of different representation-theoretic settings.
9.5 Sheaf theory
Sheaf theory relies heavily on cohomological methods to assemble local data into global conclusions. Homological algebra supplies the exactness and derived machinery needed for this process.
Sheaf cohomology is one of the most important examples of a derived functor in practice. It is central to modern geometry and many related fields.
10 Advanced topics
Advanced homological algebra extends the core theory into more flexible, abstract, or computationally refined directions. These topics often interact with topology, geometry, and higher category theory.
10.1 Cohomological operations
Cohomological operations are maps between cohomology groups that preserve algebraic structure and reveal additional symmetries. They enrich cohomology beyond its role as a graded invariant.
Such operations can encode secondary information not visible from groups alone. They are especially important in algebraic topology and related computational frameworks.
10.2 Homological perturbation theory
Homological perturbation theory studies how homological data changes under controlled modifications of differentials or chain-level structures. It is useful for transferring algebraic structures between equivalent complexes.
This theory provides explicit formulas and methods for simplifying complicated homotopical situations. It is widely used in deformation theory, algebraic topology, and higher algebra.
10.3 Model categories and homotopical algebra
Model categories provide an abstract framework for homotopy theory in which weak equivalences, fibrations, and cofibrations are specified axiomatically. They connect homological algebra with broader homotopical methods.
Homotopical algebra uses these structures to study derived and homotopy-invariant phenomena in a systematic way. It offers a setting where resolutions and localization can be treated conceptually.
10.4 Derived categories of sheaves
Derived categories of sheaves combine sheaf theory with derived methods to study geometric objects at a deeper level. They are crucial in many modern geometric constructions.
This framework allows one to handle local-to-global questions, duality statements, and cohomological phenomena uniformly. It is one of the most influential applications of derived ideas.
10.5 Noncommutative homological algebra
Noncommutative homological algebra extends the methods of the field beyond commutative rings and classical geometric settings. It studies modules, complexes, and derived invariants in noncommutative contexts.
The resulting theory retains the central themes of exactness, resolutions, and derived functors, while often requiring new techniques and subtler interpretations. It has become important in modern algebra and noncommutative geometry.