What does Maschke's theorem say and why does it require characteristic zero or coprime characteristic?

Maschke's theorem states that if G is a finite group and F is a field whose characteristic does not divide the order of G (in particular, any field of characteristic zero works), then every representation of G on a finite-dimensional F-vector space is completely reducible: it decomposes as a direct sum of irreducible subrepresentations. The key step is that any invariant subspace W has a complementary invariant subspace, constructed by averaging a projection onto W over all group elements using the formula one over the order of G times the sum over g in G of rho of g inverse composed with a projection composed with rho of g. This averaging argument fails when the characteristic divides the group order because dividing by the group order is not possible in that case, and indeed complete reducibility can fail in modular representation theory.

What is the character of a representation and what are its basic properties?

The character of a representation rho of G on V is the function chi from G to F defined by taking the trace of rho of g for each g in G. The trace is basis-independent, so chi is well-defined regardless of how you coordinatize V. Key properties: the character of the identity is the dimension of V; conjugate elements have the same character value, so chi is a class function (constant on conjugacy classes); the character of a direct sum is the sum of the characters; the character of a tensor product is the pointwise product of the characters; and for a unitary representation the character satisfies chi of g inverse equals the complex conjugate of chi of g. Over the complex numbers, the character of an irreducible representation is called an irreducible character, and the set of all irreducible characters forms an orthonormal basis for the space of class functions on G.

What does Schur's lemma say and what are its main consequences?

Schur's lemma has two parts. First: any G-equivariant linear map (intertwining operator) between two irreducible representations is either zero or an isomorphism. Second: any G-equivariant endomorphism of an irreducible complex representation is a scalar multiple of the identity. The consequences are profound. The space of intertwining operators between two non-isomorphic irreducibles is the zero space. The space of intertwining operators from an irrep to itself is one-dimensional (over algebraically closed fields of characteristic zero), spanned by the identity. This implies the character table of a finite group is a square matrix: the number of irreducible representations equals the number of conjugacy classes. It also implies that irreducible representations of abelian groups are one-dimensional, since every group element commutes with everything and must act as a scalar.

What are the orthogonality relations for characters and how are they used?

There are two orthogonality relations for the irreducible characters of a finite group G over the complex numbers. The first (row orthogonality) says that the inner product of two irreducible characters chi and psi, defined as one over the order of G times the sum over g in G of chi of g times the complex conjugate of psi of g, equals 1 if chi and psi are the same irrep and 0 otherwise. The second (column orthogonality) says that for two conjugacy classes with representatives x and y, the sum over all irreps chi of chi of x times the complex conjugate of chi of y equals the order of the centralizer of x if x and y are conjugate and 0 otherwise. These relations let you decompose any representation into irreducibles by computing inner products of its character with each irreducible character, and they let you verify that a character table is correct by checking that the rows and columns are appropriately orthogonal.

Advanced Mathematics

Representation Theory: Groups, Characters, and Lie Algebras

Q: What is a group representation and what does it mean to represent a group on a vector space?

A representation of a group G on a vector space V over a field F is a group homomorphism rho from G to GL(V), the group of invertible linear transformations of V. This means rho assigns to each group element g an invertible linear map rho of g on V such that rho of the identity equals the identity transformation, and rho of the product gh equals rho of g composed with rho of h for all g, h in G. In matrix terms, once you choose a basis, each group element is assigned an invertible matrix, and the assignment is compatible with the group operation: the matrix for gh is the product of the matrix for g and the matrix for h. The dimension of V is called the degree or dimension of the representation.

Q: What is the difference between a reducible and an irreducible representation?

A representation rho of G on V is reducible if there exists a proper nonzero subspace W of V that is invariant under every operator rho of g, meaning rho of g maps W into W for all g in G. Such a W is called a G-invariant subspace or a subrepresentation. If no such proper invariant subspace exists, the representation is called irreducible. Irreducible representations are the atoms of representation theory: every representation of a finite group over a field of characteristic zero decomposes as a direct sum of irreducibles. The irreducible representations are often called irreps, and classifying them is the central problem of representation theory for any given group.

Q: What are the representations of the Lie algebra sl(2) and how are weights and roots used?

The Lie algebra sl(2) over the complex numbers is spanned by three elements traditionally called e (raising), f (lowering), and h (Cartan), satisfying the bracket relations: the bracket of h and e is 2e, the bracket of h and f is negative 2f, and the bracket of e and f is h. A representation of sl(2) on V decomposes into weight spaces: eigenspaces of the action of h. The eigenvalues are called weights, and for finite-dimensional representations they are integers. The finite-dimensional irreducible representations are classified by their highest weight, which is a non-negative integer n, and the corresponding representation has dimension n plus 1. The weights in this representation are n, n minus 2, n minus 4, down to negative n, each appearing with multiplicity one. The operators e and f raise and lower the weight by 2 respectively. The spin-one-half representation (n equals 1) has dimension 2 and corresponds to the two-state spin system in quantum mechanics.

Representation theory asks: how can a group act on a vector space by linear transformations? By answering this question, it turns abstract symmetry into concrete matrices, unlocking deep connections between algebra, geometry, and physics. It is the mathematical foundation of quantum mechanics, particle physics, and much of modern mathematics.

1. Group Representations: Definition and First Examples
2. Reducible and Irreducible Representations
3. Maschke's Theorem and Complete Reducibility
4. Characters and Character Tables
5. Schur's Lemma and Its Consequences
6. Orthogonality Relations for Characters
7. The Regular Representation and Decomposition
8. Induced and Restricted Representations
9. Representations of Symmetric Groups and Young Tableaux
10. Representations of Lie Algebras: sl(2) and Beyond
11. Applications in Physics and Signal Processing
12. Frequently Asked Questions

1. Group Representations: Definition and First Examples

The central idea of representation theory is to study abstract groups through their actions on vector spaces. An abstract group can be difficult to analyze directly, but once you realize that group as a collection of matrices, the full power of linear algebra becomes available.

The Definition

Formal Definition of a Group Representation

Let G be a group and let V be a vector space over a field F. A representation of G on V is a group homomorphism rho from G to GL(V), the group of invertible linear transformations of V. This means: rho of the identity element equals the identity map on V, and for all g and h in G, rho of the product gh equals the composition of rho of g with rho of h.

Equivalently, once a basis for V is chosen, each element g of G is assigned an invertible square matrix rho of g, and the assignment is multiplicative: the matrix of gh is the product of the matrix of g and the matrix of h. The dimension of V is called the degree or dimension of the representation.

Key Terminology

G-Module

A representation of G on V is sometimes called a G-module. This language emphasizes that V is a module over the group algebra F[G], where group elements act on vectors by linear extension. The two frameworks are equivalent, and the module viewpoint connects representation theory to ring and module theory.

Intertwining Operators (G-maps)

Given two representations rho on V and sigma on W, a G-equivariant linear map (also called an intertwining operator or G-map) is a linear map T from V to W such that T composed with rho of g equals sigma of g composed with T for all g in G. These maps are the morphisms in the category of representations of G.

Subrepresentation

A subrepresentation is a G-invariant subspace W of V: a subspace satisfying rho of g sends W into W for every g in G. The zero subspace and V itself are always subrepresentations (the trivial ones). A nonzero subspace that is G-invariant and not equal to all of V is called a proper subrepresentation.

Direct Sum of Representations

Given two representations on V and W, their direct sum is the representation on V plus W where g acts on a pair (v, w) by applying rho of g to v in the first coordinate and sigma of g to w in the second coordinate. In matrix form, the direct sum is a block diagonal matrix with the two individual matrices on the diagonal.

Fundamental Examples

Trivial Representation

Every group G has the trivial representation on a one-dimensional space: every group element acts as the identity (multiplication by 1). It is denoted by the numeral 1 or by the symbol for the trivial character. The trivial representation is always an irreducible representation of any group.

Sign Representation of the Symmetric Group

The symmetric group on n elements has a one-dimensional representation called the sign representation: a permutation acts as multiplication by plus 1 if the permutation is even (expressible as an even number of transpositions) and by minus 1 if the permutation is odd. The sign is a group homomorphism to the multiplicative group with two elements, making it a valid one-dimensional representation.

Standard Representation of the Symmetric Group

The symmetric group on n elements acts naturally on the n-dimensional space with basis vectors labeled by the n positions, by permuting the basis vectors. The subspace of vectors with all coordinates equal is a one-dimensional invariant subspace (the trivial subrepresentation). Its orthogonal complement, the subspace of vectors with coordinate sum equal to zero, is an irreducible representation of dimension n minus 1 called the standard representation.

Rotation Group SO(3) and its Representations

The rotation group SO(3), consisting of all rotations of three-dimensional space, has irreducible representations labeled by non-negative integers l called the angular momentum quantum number. The representation labeled l has dimension 2l plus 1. For l equal to 0, this is the trivial representation (a scalar). For l equal to 1, this is the three-dimensional representation given by the rotations themselves acting on vectors. For l equal to 2, this is the five-dimensional representation corresponding to d-orbital wavefunctions in quantum mechanics.

2. Reducible and Irreducible Representations

Not all representations are created equal. Some break apart into simpler pieces, while others cannot be further decomposed. This distinction, between reducible and irreducible representations, is the organizing principle of the entire subject.

Irreducibility

A nonzero representation V of G is irreducible (also called simple as a G-module) if it has no proper nonzero G-invariant subspaces. In other words, the only invariant subspaces are the zero subspace and V itself. If V does have a proper nonzero invariant subspace, it is reducible.

A representation is completely reducible (or semisimple) if it decomposes as a direct sum of irreducible subrepresentations. Over fields of characteristic zero, finite groups always have completely reducible representations by Maschke's theorem. In the modular setting (characteristic divides group order), complete reducibility can fail, leading to the rich subject of modular representation theory.

Detecting Irreducibility: Practical Criteria

Criterion via Inner Product of Character with Itself

Over the complex numbers, a representation rho with character chi is irreducible if and only if the inner product of chi with chi equals 1. This inner product is computed as one over the order of G times the sum over all g in G of the absolute value squared of chi of g. If this quantity is an integer greater than 1, the representation is reducible and the integer tells you the total number of irreducible constituents counted with multiplicity.

Criterion for Abelian Groups

Every irreducible complex representation of a finite abelian group is one-dimensional. This follows directly from Schur's lemma: in an abelian group, each rho of g commutes with every rho of h, making each rho of g an intertwining operator from the representation to itself. By Schur's lemma, each rho of g must be a scalar multiple of the identity, and since this holds for all g, every subspace is invariant, so irreducibility forces the space to be one-dimensional.

Tensor Products of Representations

Given two representations rho on V and sigma on W, their tensor product is the representation on V tensor W where g acts by the tensor product of rho of g and sigma of g. The character of the tensor product is the pointwise product of the individual characters. The tensor product of two irreducibles may be reducible; decomposing it into irreducibles is the Clebsch-Gordan problem. For example, the tensor product of two spin-one-half representations of SU(2) decomposes into the spin-one and spin-zero representations (the triplet and the singlet), which corresponds to the addition of angular momentum in quantum mechanics.

3. Maschke's Theorem and Complete Reducibility

Maschke's theorem is the cornerstone that makes the representation theory of finite groups tractable. It guarantees that every representation breaks into irreducible pieces, provided the characteristic of the base field does not divide the group order.

Maschke's Theorem (Statement)

Let G be a finite group and F a field whose characteristic does not divide the order of G (for example, any field of characteristic zero, or a field of characteristic p where p does not divide the order of G). Then every representation of G on a finite-dimensional F-vector space is completely reducible: it is isomorphic to a direct sum of irreducible representations.

Proof Strategy: The Averaging Trick

The proof proceeds by showing that every invariant subspace W has an invariant complement. Start with any linear projection P from V onto W (not necessarily G-equivariant). Define a new projection by averaging over the group: replace P with the map that sends v to one over the order of G times the sum over all g in G of rho of g inverse applied to P applied to rho of g applied to v. This averaged projection is G-equivariant by construction, and it is still a projection onto W. Its kernel is therefore a G-invariant complement to W. By induction on the dimension of V, the conclusion follows.

The averaging step requires dividing by the order of G, which is why Maschke's theorem requires the characteristic of F not to divide the group order. When the characteristic does divide the group order, the averaged projection may not be well-defined, and complete reducibility genuinely fails.

Consequences of Maschke's Theorem

Unique Decomposition

Every representation over a field of characteristic zero decomposes as a direct sum of irreducibles, and the multiplicities of each irreducible constituent are uniquely determined by the Krull-Schmidt theorem. The multiplicity of an irreducible chi in a representation rho equals the inner product of the character of rho with chi.

Group Algebra Semisimplicity

Maschke's theorem is equivalent to the statement that the group algebra F[G] is semisimple. By the Artin-Wedderburn theorem, a semisimple algebra over an algebraically closed field decomposes as a product of matrix algebras, one for each irreducible representation, with the matrix size equal to the representation dimension.

Modular Representation Theory

When the characteristic of F divides the order of G, the resulting modular representation theory is far more complex. Representations need not be completely reducible. Instead of simple decompositions, one encounters indecomposable but not irreducible representations, projective covers, injective hulls, and the Auslander-Reiten quiver. Modular representation theory, developed by Richard Brauer, has deep connections to block theory and has been a major area of research for decades.

4. Characters and Character Tables

The character of a representation encodes the most important information about it in a compact form. Characters are functions on the group that are constant on conjugacy classes (class functions), and they determine a representation up to isomorphism.

Definition and Basic Properties

The character of a representation rho of G on V is the function chi mapping each g in G to the trace of the linear map rho of g. The trace of a linear map is the sum of its diagonal entries in any matrix representation, and it is independent of the choice of basis. Therefore chi is a well-defined function from G to F.

Value at identity:chi of the identity element equals the trace of the identity matrix, which is the dimension of V (also called the degree of the character).
Class function:chi of g and chi of hgh-inverse are always equal, so chi is constant on each conjugacy class of G.
Additivity:The character of a direct sum of representations is the sum of their characters.
Multiplicativity:The character of a tensor product of representations is the pointwise product of their characters.
Complex conjugation:For a unitary representation, the character satisfies chi of g inverse equals the complex conjugate of chi of g. For finite groups over the complex numbers, every representation is equivalent to a unitary one, so this always holds.

The Character Table

The character table of a finite group G is a square matrix whose rows are indexed by the irreducible representations (up to isomorphism) and whose columns are indexed by the conjugacy classes. The entry in the row for irrep chi and the column for the conjugacy class containing g is the value chi of g. By Schur's lemma, the number of rows equals the number of columns: the number of irreducible representations equals the number of conjugacy classes of G.

Character Table of the Symmetric Group on 3 Elements

The symmetric group on 3 elements has order 6 and three conjugacy classes: the identity (1 element), transpositions (3 elements), and 3-cycles (2 elements). It has three irreducible representations: the trivial representation (dim 1), the sign representation (dim 1), and the standard representation (dim 2).

Irrep	identity (1 element)	transpositions (3)	3-cycles (2)
trivial	1	1	1
sign	1	-1	1
standard	2	0	-1

Check: 1 squared plus 1 squared plus 2 squared equals 6, which equals the order of the group. This is a general identity: the sum of the squares of the dimensions of all irreducible representations equals the order of the group.

Reading the Character Table

Once the character table is known, one can immediately read off many structural properties of the group and its representations. The dimension of each irrep is the entry in the identity column. Whether two representations are isomorphic is determined by comparing their characters. The multiplicity of an irreducible chi in a representation rho is computed as the inner product of chi with the character of rho. The number of conjugacy classes is visible directly. Many properties of subgroups, normal subgroups, and the group structure are encoded in the character table.

5. Schur's Lemma and Its Consequences

Schur's lemma is a deceptively simple result that has far-reaching consequences. It governs the structure of maps between irreducible representations and is used constantly throughout the subject.

Schur's Lemma (Two Parts)

Let V and W be irreducible representations of a group G, and let T be a G-equivariant linear map from V to W (an intertwining operator).

Part 1: T is either the zero map or an isomorphism. If V and W are non-isomorphic irreps, then T must be zero.
Part 2: If V equals W (so T is an endomorphism of an irreducible representation) and the base field is algebraically closed, then T is a scalar multiple of the identity map on V.

Proof of Part 1

The kernel of T is a G-invariant subspace of V. Since V is irreducible, the kernel is either zero or all of V. If the kernel is all of V, then T is the zero map. If the kernel is zero, then T is injective. Similarly, the image of T is a G-invariant subspace of W, and since W is irreducible, the image is either zero or all of W. If T is injective and the image is nonzero, the image must be all of W, making T an isomorphism.

Proof of Part 2

Over an algebraically closed field, any linear endomorphism T of V has at least one eigenvalue lambda. The map T minus lambda times the identity is then an intertwining operator from V to V that has a nontrivial kernel. By Part 1, this means T minus lambda times the identity is the zero map, so T equals lambda times the identity.

Major Consequences

Number of Irreps Equals Number of Conjugacy Classes

Combined with the theory of the group algebra, Schur's lemma implies that the number of distinct irreducible representations (up to isomorphism) of a finite group equals the number of conjugacy classes. This means the character table is always a square matrix.

Irreps of Abelian Groups are One-Dimensional

In an abelian group, every rho of g commutes with every rho of h, making each rho of g an intertwining operator. By Part 2 of Schur's lemma, each rho of g is a scalar multiple of the identity. Therefore every subspace is invariant, and irreducibility forces the representation to be one-dimensional. Consequently, a finite abelian group of order n has exactly n distinct one-dimensional irreducible representations.

Multiplicity Space Structure

If an irreducible representation chi appears with multiplicity m in a representation rho, then the space of intertwining operators from the chi-isotypic component to itself is isomorphic to the space of m-by-m matrices. This gives the isotypic decomposition a canonical structure that is invisible from the character alone.

6. Orthogonality Relations for Characters

The orthogonality relations are the computational workhorses of character theory. They allow you to decompose any representation into irreducibles, verify that a proposed character table is correct, and extract structural information about the group.

Inner Product on Class Functions

The space of class functions on G (complex-valued functions constant on conjugacy classes) carries a natural inner product. The inner product of two class functions phi and psi is defined as one over the order of G times the sum over all g in G of phi of g times the complex conjugate of psi of g.

Equivalently, since phi and psi are constant on conjugacy classes, this equals the sum over all conjugacy classes C of the size of C over the order of G times phi at a representative of C times the complex conjugate of psi at that representative.

First Orthogonality Relation (Row Orthogonality)

First Orthogonality Theorem

The irreducible characters of a finite group G form an orthonormal set with respect to the inner product defined above. The inner product of two irreducible characters chi and psi equals 1 if they correspond to the same isomorphism class of irreducible representation, and equals 0 if they correspond to different irreducible representations. Furthermore, the irreducible characters form an orthonormal basis for the entire space of class functions on G.

Second Orthogonality Relation (Column Orthogonality)

Second Orthogonality Theorem

For any two elements x and y of G, the sum over all irreducible characters chi of the product of chi of x and the complex conjugate of chi of y equals the order of the centralizer of x in G if x and y are conjugate to each other, and equals zero if x and y are not conjugate. The centralizer of x is the subgroup of elements that commute with x, and its order equals the order of G divided by the size of the conjugacy class containing x.

Applications of Orthogonality

Decomposing a Representation

Given a representation rho with character chi, the multiplicity of an irreducible representation with character psi in rho equals the inner product of chi with psi. This computation requires knowing all the irreducible characters (the character table) and being able to compute the character of rho.

Checking Completeness of a Character Table

The character table is complete (contains all irreducible characters) if and only if the sum of the squares of the dimensions of all listed irreducible representations equals the order of G. Equivalently, the listed rows must be mutually orthogonal and each must have norm 1.

Burnside's Lemma Revisited

The number of orbits of G acting on a set X equals the average number of fixed points: one over the order of G times the sum over g in G of the number of elements fixed by g. This is an application of the character theory of the permutation representation associated with the action of G on X, and the fixed-point count is the character of that representation evaluated at g.

7. The Regular Representation and Decomposition

The regular representation is the most important single representation of a finite group. It is large enough to contain all irreducible representations, and its decomposition encodes the fundamental identity relating dimensions and group order.

Construction of the Regular Representation

Let G be a finite group of order n. The regular representation is the representation of G on the vector space V with basis consisting of one vector labeled by each element of G (so V has dimension n). The action of a group element g on this space is defined by left multiplication on basis vectors: g sends the basis vector labeled by h to the basis vector labeled by gh. This extends to a linear action on all of V, and the assignment g to this linear map is a group homomorphism, making it a valid representation.

Character of the Regular Representation

The action of g on the basis is the permutation matrix that maps each basis vector labeled h to the basis vector labeled gh. The trace (the character value at g) is the number of basis vectors fixed by g, that is, the number of h such that gh equals h. Since G is a group, gh equals h only if g equals the identity element. Therefore:

The character of the regular representation at the identity element equals the order of G (all basis vectors are fixed).
The character of the regular representation at any non-identity element equals 0 (no basis vectors are fixed).

Decomposition Theorem

Fundamental Decomposition of the Regular Representation

The regular representation decomposes as a direct sum of all irreducible representations, each appearing with multiplicity equal to its dimension. If the irreducible representations of G have dimensions d sub 1, d sub 2, through d sub k, then the regular representation is isomorphic to the direct sum with the first irrep appearing d sub 1 times, the second appearing d sub 2 times, and so on. Taking dimensions gives the fundamental identity: the order of G equals the sum of d sub i squared, where the sum runs over all irreducible representations.

The multiplicity formula follows immediately from the orthogonality relations: the multiplicity of an irrep with character chi in the regular representation is the inner product of the regular character with chi, which equals one over the order of G times the regular character value at the identity times the complex conjugate of chi of the identity, which equals the dimension of chi. This confirms that each irrep appears with multiplicity equal to its own dimension.

8. Induced and Restricted Representations

Induction and restriction are the two fundamental operations for relating representations of a group to representations of its subgroups. They are adjoint to each other in a precise categorical sense, captured by Frobenius reciprocity.

Restriction

If H is a subgroup of G and rho is a representation of G on V, the restriction of rho to H, written Res of rho from G to H, is simply the same vector space V with the action restricted to the elements of H. Every G-representation becomes an H-representation by forgetting about group elements outside H. An irreducible G-representation may become reducible upon restriction to H; decomposing Res into H-irreducibles is called the branching problem.

Induction

Induction goes in the other direction. If sigma is a representation of H on W, the induced representation Ind of sigma from H to G is a representation of G constructed from sigma by extending it to all of G. One concrete construction: the induced representation has a basis indexed by coset representatives of H in G, and G acts by permuting the cosets and applying sigma on each coset. The dimension of the induced representation is the index of H in G (the number of cosets) times the dimension of sigma.

Frobenius Reciprocity

Frobenius reciprocity states that induction and restriction are adjoint functors: the inner product (as class functions on G) of the character of Ind of sigma with the character of rho equals the inner product (as class functions on H) of the character of sigma with the character of Res of rho. In other words, the multiplicity of a G-irrep rho in the induced representation Ind of sigma equals the multiplicity of the H-irrep sigma in the restriction Res of rho. This is one of the most useful tools for computing induced representations without explicit matrix calculations.

Mackey's Theorem

Mackey's theorem describes the restriction of an induced representation. If you induce a representation sigma from H to G and then restrict the result to another subgroup K, the result is a direct sum of representations induced from intersections of conjugates of H with K. This theorem is essential in the harmonic analysis of finite groups and provides a systematic way to analyze how induced representations decompose.

9. Representations of Symmetric Groups and Young Tableaux

The representation theory of symmetric groups is one of the most beautiful and combinatorially rich areas of mathematics. The irreducible representations are parametrized by combinatorial objects (Young tableaux), and their dimensions are given by an elegant formula.

Partitions and Young Diagrams

A partition of a positive integer n is a way of writing n as an ordered sum of positive integers with the parts in nonincreasing order. For example, the partitions of 4 are: 4, then 3 plus 1, then 2 plus 2, then 2 plus 1 plus 1, and finally 1 plus 1 plus 1 plus 1. Each partition is visualized as a Young diagram: a left-justified array of boxes where the j-th row from the top contains as many boxes as the j-th part of the partition.

Standard Young Tableaux

A standard Young tableau of shape lambda is a filling of the boxes of the Young diagram for lambda with the numbers 1 through n, each appearing exactly once, such that the entries are strictly increasing across each row from left to right and strictly increasing down each column from top to bottom. The number of standard Young tableaux of a given shape is the dimension of the corresponding irreducible representation of the symmetric group.

Hook Length Formula

The dimension of the irreducible representation of the symmetric group on n elements corresponding to a partition lambda is given by n factorial divided by the product over all boxes in the Young diagram of the hook length of that box. The hook length of a box is the number of boxes directly to the right in the same row plus the number of boxes directly below in the same column plus one (for the box itself). This formula, due to Frame, Robinson, and Thrall, gives an efficient way to compute dimensions without enumerating all standard Young tableaux.

Specht Modules

The irreducible representation corresponding to a partition lambda is called the Specht module for lambda. It can be constructed explicitly from the Young tableau via Young symmetrizers: for each standard Young tableau T of shape lambda, define a Young symmetrizer as the product of two elements of the group algebra of the symmetric group, one that symmetrizes within rows and one that antisymmetrizes within columns. The image of the group algebra under any Young symmetrizer is isomorphic to the Specht module for lambda.

Branching Rules and the RSK Correspondence

The restriction of the Specht module for lambda from the symmetric group on n elements to the symmetric group on n minus 1 elements decomposes as a direct sum of Specht modules for all partitions obtained by removing one box from the Young diagram of lambda in such a way that the result is still a valid Young diagram. This branching rule gives a recursive understanding of all representations. The Robinson-Schensted-Knuth correspondence is a bijection between permutations and pairs of standard Young tableaux of the same shape, with deep connections to the representation theory of symmetric groups and to the theory of symmetric functions.

10. Representations of Lie Algebras: sl(2) and Beyond

Lie algebras arise as the infinitesimal symmetries of Lie groups. Their representation theory parallels the finite group case but uses different techniques: instead of averaging over a finite group, one uses the structure of the Lie bracket and the theory of roots and weights.

What is a Lie Algebra Representation?

A representation of a Lie algebra g on a vector space V is a Lie algebra homomorphism from g to the Lie algebra of endomorphisms of V. This means: each element X of g is assigned a linear map pi of X on V, the map pi is linear in X, and pi preserves the Lie bracket: pi of the bracket of X and Y equals the commutator of pi of X and pi of Y (that is, pi of X composed with pi of Y minus pi of Y composed with pi of X).

The Lie Algebra sl(2) Over the Complex Numbers

The Lie algebra sl(2) consists of all 2-by-2 complex matrices with trace zero. It is three-dimensional with a standard basis: the element e (the raising operator), the element f (the lowering operator), and the element h (the Cartan element).

Bracket of h and e: equals 2 times e. The element e is an eigenvector of the adjoint action of h with eigenvalue 2.
Bracket of h and f: equals negative 2 times f. The element f is an eigenvector of the adjoint action of h with eigenvalue minus 2.
Bracket of e and f: equals h (the Cartan element).

Finite-Dimensional Irreducible Representations of sl(2)

The finite-dimensional irreducible representations of sl(2) over the complex numbers are completely classified. For each non-negative integer n, there is exactly one irreducible representation of dimension n plus 1, up to isomorphism. This representation has a basis of weight vectors labeled v sub 0 through v sub n, where the action of h on the weight vector v sub k has eigenvalue n minus 2k. The weight of v sub k is n minus 2k, so the weights are n, n minus 2, n minus 4, down to negative n, each appearing once. The element e raises the weight by 2 (sending v sub k toward v sub k-1, or to zero if k equals 0), and f lowers the weight by 2 (sending v sub k toward v sub k+1, or to zero if k equals n). The highest weight is n.

Weights, Roots, and Semisimple Lie Algebras

For a general semisimple Lie algebra g, the theory of weights and roots generalizes the sl(2) picture. The Cartan subalgebra is a maximal abelian subalgebra of diagonalizable elements. The roots are the eigenvalues (as linear functionals on the Cartan subalgebra) of the adjoint action, and the root spaces give a decomposition of g called the root space decomposition. In any representation, the weight spaces are the common eigenspaces for the Cartan action, and the raising and lowering operators move between weight spaces by adding or subtracting roots.

Theorem of the Highest Weight (Cartan's Classification)

The finite-dimensional irreducible representations of a complex semisimple Lie algebra are classified by their highest weights, which are dominant integral weights. Two representations are isomorphic if and only if they have the same highest weight. Every dominant integral weight occurs as the highest weight of some irreducible representation. This theorem completely solves the classification problem for finite-dimensional irreducible representations of semisimple Lie algebras.

The Weyl Character Formula

Given the highest weight of a finite-dimensional irreducible representation of a semisimple Lie algebra, the Weyl character formula gives an explicit formula for the character of that representation (encoding all weight multiplicities). The Weyl dimension formula is a corollary that gives the dimension of the representation directly in terms of the highest weight and the roots. For sl(2), the Weyl dimension formula reduces to the statement that the dimension of the representation with highest weight n is n plus 1.

11. Applications in Physics and Signal Processing

Representation theory is not just abstract mathematics: it is the mathematical language of modern physics and has practical applications in engineering and data science.

Quantum Mechanics and Spin

In quantum mechanics, the state space of a quantum system is a Hilbert space, and physical symmetries (rotations, reflections, time reversal) act on this space by unitary operators. Wigner's theorem says that any symmetry of quantum mechanics acts by a unitary or antiunitary operator on the Hilbert space, making the physical symmetry group into a group of unitary operators, that is, a unitary representation of the symmetry group.

The irreducible unitary representations of the rotation group SU(2) are labeled by spin s, which can be a non-negative integer or half-integer. The spin-s representation has dimension 2s plus 1. Elementary particles are classified by their spin: electrons, protons, and quarks are spin one-half (fermions), photons and gluons are spin 1 (bosons), and the Higgs boson is spin 0. The addition of angular momenta in quantum mechanics is precisely the Clebsch-Gordan decomposition of tensor products of representations.

The Standard Model of Particle Physics

The Standard Model is built on three gauge symmetry groups: SU(3) for the strong force, SU(2) for the weak force, and U(1) for electromagnetism. The particles are classified by the representations of these groups under which they transform. Quarks transform in the fundamental three-dimensional representation of SU(3), while gluons transform in the adjoint eight-dimensional representation.

The prediction of new particles has historically been guided by representation theory. Before quarks were observed experimentally, Gell-Mann used the representation theory of SU(3) to organize known hadrons into multiplets (the eightfold way) and to predict the existence of the omega-minus particle, which was subsequently discovered. This is a striking example of mathematics revealing physical reality ahead of experiment.

Harmonic Analysis and Signal Processing

The Fourier transform is a special case of harmonic analysis on groups. For the group of real numbers under addition, the irreducible unitary representations are complex exponentials labeled by frequency, and the Fourier transform decomposes a function into these irreducible components. For the cyclic group of order n, the corresponding analysis gives the discrete Fourier transform, and the fast Fourier transform is an efficient algorithm that exploits the group structure.

For non-abelian groups, the Peter-Weyl theorem generalizes Fourier analysis: functions on a compact group decompose into matrix coefficients of irreducible representations. This non-abelian Fourier theory has applications in image analysis (functions on the sphere decompose using spherical harmonics, which are matrix coefficients of representations of SO(3)), protein structure analysis, and the study of signals on networks with symmetry.

Crystallography and Molecular Symmetry

Crystal structures are classified by their space groups, which are discrete symmetry groups of three-dimensional space combining translations and point symmetries. Representation theory of these groups governs the selection rules for which vibrational modes of a molecule are infrared- or Raman-active, which electronic transitions are allowed by symmetry, and how orbitals split in the presence of a crystal field.

The classification of molecular orbitals in group theory labels each orbital by the irreducible representation of the molecular point group under which it transforms. The labels used (A, B, E, T and their subscripts, from the Mulliken symbol system) directly reflect the representation-theoretic structure of the molecular symmetry group.

12. Frequently Asked Questions

What is a group representation and what does it mean to represent a group on a vector space?▼

A representation of a group G on a vector space V is a group homomorphism rho from G to GL(V), where GL(V) is the group of all invertible linear transformations of V. The homomorphism condition means two things: rho sends the identity of G to the identity transformation on V, and rho of the product of g and h equals rho of g composed with rho of h for all group elements g and h.

In concrete matrix terms: once you choose a basis for V, each group element g gets assigned an invertible square matrix, and the assignment is multiplicative: the matrix of gh equals the matrix of g times the matrix of h. The size of these matrices (the dimension of V) is the dimension or degree of the representation.

The whole point is that abstract group theory is hard, but matrices are computable. A representation lets you study the group by looking at how its elements move vectors around in a concrete linear space. Different representations reveal different facets of the group's structure.

What is the difference between a reducible and an irreducible representation?▼

A representation of G on V is reducible if you can find a proper nonzero subspace W of V that is invariant under the action of every group element. Invariant means: whenever w is in W and g is in G, the element rho of g applied to w is still in W. Such a W is called a subrepresentation.

A representation is irreducible if no such proper invariant subspace exists. The only invariant subspaces are the zero subspace and V itself. Irreducible representations are the building blocks: over fields of characteristic zero, every finite-dimensional representation of a finite group breaks apart as a direct sum of irreducibles by Maschke's theorem, and the multiplicities are uniquely determined.

A helpful analogy: irreducible representations are like prime numbers, and general representations are like positive integers. Just as every positive integer factors uniquely into primes, every representation decomposes uniquely into irreducibles (in the characteristic-zero finite group setting).

What does Maschke's theorem say and why does it need characteristic zero?▼

Maschke's theorem says that every representation of a finite group G over a field of characteristic zero (such as the real or complex numbers) is completely reducible: it decomposes as a direct sum of irreducible subrepresentations. The key idea is that any invariant subspace W has an invariant complement, constructed by averaging a projection over the group.

The averaging construction divides by the order of G: you take any linear projection P onto W and replace it with one over the order of G times the sum over all g in G of rho of g inverse composed with P composed with rho of g. This new projection is G-equivariant, and its kernel is an invariant complement to W.

The problem with characteristic p when p divides the group order: dividing by the order of G means multiplying by the multiplicative inverse of the order of G in the field. If the characteristic of the field divides the order of G, then the order of G is zero in the field, and division by zero is not allowed. In this setting (called modular representation theory), complete reducibility fails. Indecomposable representations need not be irreducible, and the theory becomes much richer and more complicated.

What is the character of a representation and how is it used?▼

The character of a representation rho of G on V is the function chi that assigns to each g in G the trace of the linear map rho of g. The trace is the sum of the diagonal entries in any matrix representation, and it does not depend on the choice of basis, so chi is well-defined.

Key properties: the value of chi at the identity element equals the dimension of V. Conjugate elements always have the same chi value, so chi is a class function. The character of a direct sum of representations is the sum of their characters, and the character of a tensor product is the pointwise product.

The character determines the representation up to isomorphism over the complex numbers: two representations are isomorphic if and only if they have the same character. The orthogonality relations for characters allow you to compute multiplicities: the multiplicity of an irreducible representation with character psi in a representation with character chi is the inner product of chi and psi, computed as one over the order of G times the sum over g of chi of g times the complex conjugate of psi of g.

What does Schur's lemma say and what are its consequences?▼

Schur's lemma has two parts. Part 1: any G-equivariant linear map (intertwining operator) between two irreducible representations is either the zero map or an isomorphism. Part 2: over an algebraically closed field, any G-equivariant endomorphism of an irreducible representation is a scalar multiple of the identity.

The proof of Part 1 is elegant: the kernel of an intertwining operator is an invariant subspace, and by irreducibility of the domain, it must be zero or the whole space. Similarly, the image is an invariant subspace of the codomain, and by irreducibility of the codomain, it is zero or the whole space. Part 2 follows because over an algebraically closed field any endomorphism has an eigenvalue lambda, and T minus lambda times the identity is an intertwining operator with a nontrivial kernel, so by Part 1 it must be zero.

Consequences: the number of irreducible representations of a finite group equals the number of its conjugacy classes. Every irreducible complex representation of an abelian group is one-dimensional. The space of intertwining operators between non-isomorphic irreducibles is zero. These facts are foundational for the entire theory.

How are Young tableaux used to classify representations of symmetric groups?▼

The irreducible complex representations of the symmetric group on n elements are indexed by partitions of n. A partition of n is a list of positive integers that add up to n, written in nonincreasing order. Each partition is visualized as a Young diagram: a staircase-like array of boxes where the j-th row has as many boxes as the j-th part.

A standard Young tableau is a filling of the boxes with numbers 1 through n such that entries increase along each row from left to right and increase down each column. The number of standard Young tableaux of a given shape equals the dimension of the corresponding irreducible representation, computed by the hook length formula: dimension equals n factorial divided by the product of the hook lengths of all boxes.

The actual construction of the irreducible representation, called the Specht module, uses Young symmetrizers built from the tableau combinatorics. The theory is completely explicit: given a partition, you can write down a basis, compute the matrices, and verify the representation axioms. This combinatorial explicitness is one of the great virtues of symmetric group representation theory.

What are the finite-dimensional irreducible representations of sl(2)?▼

The Lie algebra sl(2) is three-dimensional, spanned by elements e (raising), f (lowering), and h (Cartan), with brackets: the bracket of h and e is 2 times e, the bracket of h and f is minus 2 times f, and the bracket of e and f is h.

The finite-dimensional irreducible representations are classified by a non-negative integer n called the highest weight. The representation of highest weight n has dimension n plus 1. It has a basis of vectors labeled v sub 0 through v sub n, where h acts on v sub k with eigenvalue n minus 2k (so the eigenvalues, called weights, are n, n minus 2, down to negative n). The element e raises the weight by 2, and f lowers the weight by 2.

For n equal to 0: the trivial representation. For n equal to 1: the standard two-dimensional representation (also called the spin-one-half representation in physics). For n equal to 2: a three-dimensional representation. These correspond in physics to the spin-0 (singlet), spin-one-half (doublet), and spin-1 (triplet) particles.

What is Frobenius reciprocity and why is it useful?▼

Frobenius reciprocity is the adjunction relation between induction and restriction. If H is a subgroup of G, sigma is an irreducible representation of H, and rho is an irreducible representation of G, then the multiplicity of rho in the induced representation Ind of sigma from H to G equals the multiplicity of sigma in the restriction Res of rho from G to H.

In character-theoretic language: the inner product of the character of Ind of sigma (as a class function on G) with the character of rho equals the inner product of the character of sigma (as a class function on H) with the character of Res of rho restricted to H.

The practical importance: computing the induced representation directly can be difficult. But computing the restriction Res of rho to H and then computing inner products of characters on H is often easy. Frobenius reciprocity lets you bypass the difficult induction calculation and work on the simpler subgroup instead. It is indispensable for computing character tables of groups by leveraging known subgroup information.

How is representation theory used in quantum mechanics?▼

In quantum mechanics, the state space of a quantum system is a Hilbert space, and physical symmetries (rotations, reflections, time reversal) act on it by unitary operators. Wigner's theorem guarantees that any symmetry of quantum mechanics acts by a unitary or antiunitary operator on the Hilbert space. This makes the physical symmetry group into a group of unitary operators, that is, a unitary representation of the symmetry group.

Selection rules (which transitions between quantum states are allowed) are derived from representation theory: a transition is forbidden if the corresponding matrix element is zero by symmetry. Since the matrix element is zero whenever the tensor product of the initial and final state representations does not contain the representation of the interaction, one can determine selection rules purely from representation-theoretic calculations without evaluating the underlying integrals.

What is the regular representation and how does it decompose into irreducibles?▼

The regular representation of a finite group G is the representation on the vector space with one basis vector for each group element (so the dimension equals the order of G), where g acts on the basis by left multiplication: g sends the basis vector labeled h to the basis vector labeled gh.

The character of the regular representation is easy to compute. The action of a non-identity element g sends every basis vector to a different basis vector (since gh is not equal to h when g is not the identity), so the corresponding permutation matrix has no fixed points and its trace is zero. The identity element acts trivially, so its trace equals the dimension, which is the order of G.

The decomposition theorem says the regular representation contains each irreducible representation exactly d times, where d is the dimension of that irreducible. This follows from the multiplicity formula: the multiplicity of an irreducible with character chi in the regular representation equals the inner product of the regular character with chi, which works out to the value of chi at the identity, which is the dimension d. Taking the total dimension gives the fundamental identity: the order of G equals the sum of the squares of the dimensions of all irreducible representations.

Ready to Master Representation Theory?

NailTheTest offers comprehensive practice problems, worked examples, and exam preparation for every level of advanced mathematics. Whether you're preparing for a graduate qualifying exam or deepening your understanding of the mathematical foundations of physics, we have the tools you need.

Start Practicing Browse All Math Topics