Lie algebra

Lie algebras were introduced to study the concept of infinitesimal transformations by Sophus Lie in the 1870s,^[1] and independently discovered by Wilhelm Killing^[2] in the 1880s. The name Lie algebra was given by Hermann Weyl in the 1930s; in older texts, the term infinitesimal group was used.

A Lie algebra is a vector space ${\displaystyle \,{\mathfrak {g}}}$ over a field ${\displaystyle F}$ together with a binary operation ${\displaystyle [\,\cdot \,,\cdot \,]:{\mathfrak {g}}\times {\mathfrak {g}}\to {\mathfrak {g}}}$ called the Lie bracket, satisfying the following axioms:^[a]

• Bilinearity,

${\displaystyle [ax+by,z]=a[x,z]+b[y,z],}$

${\displaystyle [z,ax+by]=a[z,x]+b[z,y]}$

for all scalars ${\displaystyle a,b}$ in ${\displaystyle F}$ and all elements ${\displaystyle x,y,z}$ in ${\displaystyle {\mathfrak {g}}}$.

• The Alternating property,

${\displaystyle [x,x]=0\ }$

for all ${\displaystyle x}$ in ${\displaystyle {\mathfrak {g}}}$.

• The Jacobi identity,

${\displaystyle [x,[y,z]]+[z,[x,y]]+[y,[z,x]]=0\ }$

for all ${\displaystyle x,y,z}$ in ${\displaystyle {\mathfrak {g}}}$.

Given a Lie group, the Jacobi identity for its Lie algebra follows from the associativity of the group operation.

Using bilinearity to expand the Lie bracket ${\displaystyle [x+y,x+y]}$ and using the alternating property shows that ${\displaystyle [x,y]+[y,x]=0}$ for all ${\displaystyle x,y}$ in ${\displaystyle {\mathfrak {g}}}$. Thus bilinearity and the alternating property together imply

• Anticommutativity,

${\displaystyle [x,y]=-[y,x],\ }$

for all ${\displaystyle x,y}$ in ${\displaystyle {\mathfrak {g}}}$. If the field does not have characteristic 2, then anticommutativity implies the alternating property, since it implies ${\displaystyle [x,x]=-[x,x].}$^[3]

• Derivation property, the anti commutativity of the Lie bracket allows to rewrite the Jacobi identity as a "Leibnitz rule" for ${\displaystyle \mathrm {ad} _{x}=[x,-]}$:

${\displaystyle [x,[y,z]]=[[x,y],z]+[y,[x,z]],\ }$

for all ${\displaystyle x,y,z}$ in ${\displaystyle {\mathfrak {g}}}$.

It is customary to denote a Lie algebra by a lower-case fraktur letter such as ${\displaystyle {\mathfrak {g,h,b,n}}}$. If a Lie algebra is associated with a Lie group, then the algebra is denoted by the fraktur version of the group's name: for example, the Lie algebra of SU(n) is ${\displaystyle {\mathfrak {su}}(n)}$.

The dimension of a Lie algebra over a field means its dimension as a vector space. In physics, a vector space basis of the Lie algebra of a Lie group G may be called a set of generators for G. (They are "infinitesimal generators" for G, so to speak.) In mathematics, a set S of generators for a Lie algebra ${\displaystyle {\mathfrak {g}}}$ means a subset of ${\displaystyle {\mathfrak {g}}}$ such that any Lie subalgebra (as defined below) that contains S must be all of ${\displaystyle {\mathfrak {g}}}$. Equivalently, ${\displaystyle {\mathfrak {g}}}$ is spanned (as a vector space) by all iterated brackets of elements of S.

A Lie algebra is called abelian if its Lie bracket is identically zero. Any vector space ${\displaystyle V}$ endowed with the identically zero Lie bracket becomes a Lie algebra. Every one-dimensional Lie algebra is abelian, by the alternating property of the Lie bracket.

• On an associative algebra ${\displaystyle A}$ over a field ${\displaystyle F}$ with multiplication written as ${\displaystyle xy}$, a Lie bracket may be defined by the commutator ${\displaystyle [x,y]=xy-yx}$. With this bracket, ${\displaystyle A}$ is a Lie algebra. (The Jacobi identity follows from the associativity of the multiplication on ${\displaystyle A}$.) ^[4]
• The endomorphism ring of an ${\displaystyle F}$-vector space ${\displaystyle V}$ with the above Lie bracket is denoted ${\displaystyle {\mathfrak {gl}}(V)}$.
• For a field F and a positive integer n, the space of n × n matrices over F, denoted ${\displaystyle {\mathfrak {gl}}(n,F)}$ or ${\displaystyle {\mathfrak {gl}}_{n}(F)}$, is a Lie algebra with bracket given by the commutator of matrices: ${\displaystyle [X,Y]=XY-YX}$.^[5] This is a special case of the previous example; it is a key example of a Lie algebra. It is called the general linear Lie algebra.

When F is the real numbers, ${\displaystyle {\mathfrak {gl}}(n,\mathbb {R} )}$ is the Lie algebra of the general linear group ${\displaystyle \mathrm {GL} (n,\mathbb {R} )}$, the group of invertible n x n real matrices (or equivalently, matrices with nonzero determinant), where the group operation is matrix multiplication. Likewise, ${\displaystyle {\mathfrak {gl}}(n,\mathbb {C} )}$ is the Lie algebra of the complex Lie group ${\displaystyle \mathrm {GL} (n,\mathbb {C} )}$. The Lie bracket on ${\displaystyle {\mathfrak {gl}}(n,\mathbb {R} )}$ describes the failure of commutativity for matrix multiplication, or equivalently for the composition of linear maps. For any field F, ${\displaystyle {\mathfrak {gl}}(n,F)}$ can be viewed as the Lie algebra of the algebraic group ${\displaystyle \mathrm {GL} (n)}$ over F.

The Lie bracket is not required to be associative, meaning that ${\displaystyle [[x,y],z]}$ need not be equal to ${\displaystyle [x,[y,z]]}$. Nonetheless, much of the terminology for associative rings and algebras (and also for groups) has analogs for Lie algebras. A Lie subalgebra is a linear subspace ${\displaystyle {\mathfrak {h}}\subseteq {\mathfrak {g}}}$ which is closed under the Lie bracket. An ideal ${\displaystyle {\mathfrak {i}}\subseteq {\mathfrak {g}}}$ is a linear subspace that satisfies the stronger condition:^[6]

${\displaystyle [{\mathfrak {g}},{\mathfrak {i}}]\subseteq {\mathfrak {i}}.}$

In the correspondence between Lie groups and Lie algebras, subgroups correspond to Lie subalgebras, and normal subgroups correspond to ideals.

A Lie algebra homomorphism is a linear map compatible with the respective Lie brackets:

${\displaystyle \phi \colon {\mathfrak {g}}\to {\mathfrak {h}},\quad \phi ([x,y])=[\phi (x),\phi (y)]\ {\text{for all}}\ x,y\in {\mathfrak {g}}.}$

An isomorphism of Lie algebras is a bijective homomorphism.

As with normal subgroups in groups, ideals in Lie algebras are precisely the kernels of homomorphisms. Given a Lie algebra ${\displaystyle {\mathfrak {g}}}$ and an ideal ${\displaystyle {\mathfrak {i}}}$ in it, the quotient Lie algebra ${\displaystyle {\mathfrak {g}}/{\mathfrak {i}}}$ is defined, with a surjective homomorphism ${\displaystyle {\mathfrak {g}}\to {\mathfrak {g}}/{\mathfrak {i}}}$ of Lie algebras. The first isomorphism theorem holds for Lie algebras: for any homomorphism ${\displaystyle \phi \colon {\mathfrak {g}}\to {\mathfrak {h}}}$ of Lie algebras, the image of ${\displaystyle \phi }$ is a Lie subalgebra of ${\displaystyle {\mathfrak {h}}}$ that is isomorphic to ${\displaystyle {\mathfrak {g}}/{\text{ker}}(\phi )}$.

For the Lie algebra of a Lie group, the Lie bracket is a kind of infinitesimal commutator. As a result, for any Lie algebra, two elements ${\displaystyle x,y\in {\mathfrak {g}}}$ are said to commute if their bracket vanishes: ${\displaystyle [x,y]=0}$.

The centralizer subalgebra of a subset ${\displaystyle S\subset {\mathfrak {g}}}$ is the set of elements commuting with ${\displaystyle S}$: that is, ${\displaystyle {\mathfrak {z}}_{\mathfrak {g}}(S)=\{x\in {\mathfrak {g}}:[x,s]=0\ {\text{ for all }}s\in S\}}$. The centralizer of ${\displaystyle {\mathfrak {g}}}$ itself is the center ${\displaystyle {\mathfrak {z}}({\mathfrak {g}})}$. Similarly, for a subspace S, the normalizer subalgebra of ${\displaystyle S}$ is ${\displaystyle {\mathfrak {n}}_{\mathfrak {g}}(S)=\{x\in {\mathfrak {g}}:[x,s]\in S\ {\text{ for all}}\ s\in S\}}$.^[7] If ${\displaystyle S}$ is a Lie subalgebra, ${\displaystyle {\mathfrak {n}}_{\mathfrak {g}}(S)}$ is the largest subalgebra such that ${\displaystyle S}$ is an ideal of ${\displaystyle {\mathfrak {n}}_{\mathfrak {g}}(S)}$.

The subspace ${\displaystyle {\mathfrak {t}}_{n}}$ of diagonal matrices in ${\displaystyle {\mathfrak {gl}}(n,F)}$ is an abelian Lie subalgebra. (It is a Cartan subalgebra of ${\displaystyle {\mathfrak {gl}}(n)}$, analogous to a maximal torus in the theory of compact Lie groups.) Here ${\displaystyle {\mathfrak {t}}_{n}}$ is not an ideal in ${\displaystyle {\mathfrak {gl}}(n)}$ for ${\displaystyle n\geq 2}$. For example, when ${\displaystyle n=2}$, this follows from the calculation:

${\displaystyle {\begin{aligned}\left[{\begin{bmatrix}a&b\\c&d\end{bmatrix}},{\begin{bmatrix}x&0\\0&y\end{bmatrix}}\right]&={\begin{bmatrix}ax&by\\cx&dy\\\end{bmatrix}}-{\begin{bmatrix}ax&bx\\cy&dy\\\end{bmatrix}}\\&={\begin{bmatrix}0&b(y-x)\\c(x-y)&0\end{bmatrix}}\end{aligned}}}$

(which is not always in ${\displaystyle {\mathfrak {t}}_{2}}$).

Every one-dimensional linear subspace of a Lie algebra ${\displaystyle {\mathfrak {g}}}$ is an abelian Lie subalgebra, but it need not be an ideal.

For two Lie algebras ${\displaystyle {\mathfrak {g}}}$ and ${\displaystyle {\mathfrak {g'}}}$, the product Lie algebra is the vector space ${\displaystyle {\mathfrak {g}}\times {\mathfrak {g'}}}$ consisting of all ordered pairs ${\displaystyle (x,x'),\,x\in {\mathfrak {g}},\ x'\in {\mathfrak {g'}}}$, with Lie bracket^[8]

${\displaystyle [(x,x'),(y,y')]=([x,y],[x',y']).}$

This is the product in the category of Lie algebras. Note that the copies of ${\displaystyle {\mathfrak {g}}}$ and ${\displaystyle {\mathfrak {g}}'}$ in ${\displaystyle {\mathfrak {g}}\times {\mathfrak {g'}}}$ commute with each other: ${\displaystyle [(x,0),(0,x')]=0.}$

Let ${\displaystyle {\mathfrak {g}}}$ be a Lie algebra and ${\displaystyle {\mathfrak {i}}}$ an ideal of ${\displaystyle {\mathfrak {g}}}$. If the canonical map ${\displaystyle {\mathfrak {g}}\to {\mathfrak {g}}/{\mathfrak {i}}}$ splits (i.e., admits a section ${\displaystyle {\mathfrak {g}}/{\mathfrak {i}}\to {\mathfrak {g}}}$, as a homomorphism of Lie algebras), then ${\displaystyle {\mathfrak {g}}}$ is said to be a semidirect product of ${\displaystyle {\mathfrak {i}}}$ and ${\displaystyle {\mathfrak {g}}/{\mathfrak {i}}}$, ${\displaystyle {\mathfrak {g}}={\mathfrak {g}}/{\mathfrak {i}}\ltimes {\mathfrak {i}}}$. See also semidirect sum of Lie algebras.

For an algebra A over a field F, a derivation of A over F is a linear map ${\displaystyle D\colon A\to A}$ that satisfies the Leibniz rule

${\displaystyle D(xy)=D(x)y+xD(y)}$

for all ${\displaystyle x,y\in A}$. (The definition makes sense for a possibly non-associative algebra.) Given two derivations ${\displaystyle D_{1}}$ and ${\displaystyle D_{2}}$, their commutator ${\displaystyle [D_{1},D_{2}]:=D_{1}D_{2}-D_{2}D_{1}}$ is again a derivation. This operation makes the space ${\displaystyle {\text{Der}}_{k}(A)}$ of all derivations of A over F into a Lie algebra.^[9]

Informally speaking, the space of derivations of A is the Lie algebra of the automorphism group of A. (This is literally true when the automorphism group is a Lie group, for example when F is the real numbers and A has finite dimension as a vector space.) For this reason, spaces of derivations are a natural way to construct Lie algebras: they are the "infinitesimal automorphisms" of A. Indeed, writing out the condition that

${\displaystyle (1+\epsilon D)(xy)\equiv (1+\epsilon D)(x)\cdot (1+\epsilon D)(y){\pmod {\epsilon ^{2}}}}$

(where 1 denotes the identity map on A) gives exactly the definition of D being a derivation.

Example: the Lie algebra of vector fields. Let A be the ring ${\displaystyle C^{\infty }(X)}$ of smooth functions on a smooth manifold X. Then a derivation of A over ${\displaystyle \mathbb {R} }$ is equivalent to a vector field on X. (A vector field v gives a derivation of the space of smooth functions by differentiating functions in the direction of v.) This makes the space ${\displaystyle {\text{Vect}}(X)}$ of vector fields into a Lie algebra (see Lie bracket of vector fields).^[10] Informally speaking, ${\displaystyle {\text{Vect}}(X)}$ is the Lie algebra of the diffeomorphism group of X. So the Lie bracket of vector fields describes the non-commutativity of the diffeomorphism group. An action of a Lie group G on a manifold X determines a homomorphism of Lie algebras ${\displaystyle {\mathfrak {g}}\to {\text{Vect}}(X)}$. (An example is illustrated below.)

A Lie algebra can be viewed as a non-associative algebra, and so each Lie algebra ${\displaystyle {\mathfrak {g}}}$ over a field F determines its Lie algebra of derivations, ${\displaystyle {\text{Der}}_{F}({\mathfrak {g}})}$. That is, a derivation of ${\displaystyle {\mathfrak {g}}}$ is a linear map ${\displaystyle D\colon {\mathfrak {g}}\to {\mathfrak {g}}}$ such that

${\displaystyle D([x,y])=[D(x),y]+[x,D(y)]}$.

The inner derivation associated to any ${\displaystyle x\in {\mathfrak {g}}}$ is the adjoint mapping ${\displaystyle \mathrm {ad} _{x}}$ defined by ${\displaystyle \mathrm {ad} _{x}(y):=[x,y]}$. (This is a derivation as a consequence of the Jacobi identity.) That gives a homomorphism of Lie algebras, ${\displaystyle \operatorname {ad} \colon {\mathfrak {g}}\to {\text{Der}}_{F}({\mathfrak {g}})}$. The image ${\displaystyle {\text{Inn}}_{F}({\mathfrak {g}})}$ is an ideal in ${\displaystyle {\text{Der}}_{F}({\mathfrak {g}})}$, and the Lie algebra of outer derivations is defined as the quotient Lie algebra, ${\displaystyle {\text{Out}}_{F}({\mathfrak {g}})={\text{Der}}_{F}({\mathfrak {g}})/{\text{Inn}}_{F}({\mathfrak {g}})}$. (This is exactly analogous to the outer automorphism group of a group.) For a semisimple Lie algebra (defined below) over a field of characteristic zero, every derivation is inner.^[11] This is related to the theorem that the outer automorphism group of a semisimple Lie group is finite.^[12]

In contrast, an abelian Lie algebra has many outer derivations. Namely, for a vector space ${\displaystyle V}$ with Lie bracket zero, the Lie algebra ${\displaystyle {\text{Out}}_{F}(V)}$ can be identified with ${\displaystyle {\mathfrak {gl}}(V)}$.

A matrix group is a Lie group consisting of invertible matrices, ${\displaystyle G\subset \mathrm {GL} (n,\mathbb {R} )}$, where the group operation of G is matrix multiplication. The corresponding Lie algebra ${\displaystyle {\mathfrak {g}}}$ is the space of matrices which are tangent vectors to G inside the linear space ${\displaystyle M_{n}(\mathbb {R} )}$: this consists of derivatives of smooth curves in G at the identity matrix ${\displaystyle I}$:

${\displaystyle {\mathfrak {g}}=\{X=c'(0)\in M_{n}(\mathbb {R} ):{\text{ smooth }}c:\mathbb {R} \to G,\ c(0)=I\}.}$

The Lie bracket of ${\displaystyle {\mathfrak {g}}}$ is given by the commutator of matrices, ${\displaystyle [X,Y]=XY-YX}$. Given a Lie algebra ${\displaystyle {\mathfrak {g}}\subset {\mathfrak {gl}}(n,\mathbb {R} )}$, one can recover the Lie group as the subgroup generated by the matrix exponential of elements of ${\displaystyle {\mathfrak {g}}}$.^[13] (To be precise, this gives the identity component of G, if G is not connected.) Here the exponential mapping ${\displaystyle \exp :M_{n}(\mathbb {R} )\to M_{n}(\mathbb {R} )}$ is defined by ${\displaystyle \exp(X)=I+X+{\tfrac {1}{2!}}X^{2}+{\tfrac {1}{3!}}X^{3}+\cdots }$, which converges for every matrix ${\displaystyle X}$.

The same comments apply to complex Lie subgroups of ${\displaystyle GL(n,\mathbb {C} )}$ and the complex matrix exponential, ${\displaystyle \exp :M_{n}(\mathbb {C} )\to M_{n}(\mathbb {C} )}$ (defined by the same formula).

Here are some matrix Lie groups and their Lie algebras.^[14]

• For a positive integer n, the special linear group ${\displaystyle \mathrm {SL} (n,\mathbb {R} )}$ consists of all real n × n matrices with determinant 1. This is the group of linear maps from ${\displaystyle \mathbb {R} ^{n}}$ to itself that preserve volume and orientation. More abstractly, ${\displaystyle \mathrm {SL} (n,\mathbb {R} )}$ is the commutator subgroup of the general linear group ${\displaystyle \mathrm {GL} (n,\mathbb {R} )}$. Its Lie algebra ${\displaystyle {\mathfrak {sl}}(n,\mathbb {R} )}$ consists of all real n × n matrices with trace 0. Similarly, one can define the analogous complex Lie group ${\displaystyle {\rm {SL}}(n,\mathbb {C} )}$ and its Lie algebra ${\displaystyle {\mathfrak {sl}}(n,\mathbb {C} )}$.
• The orthogonal group ${\displaystyle \mathrm {O} (n)}$ plays a basic role in geometry: it is the group of linear maps from ${\displaystyle \mathbb {R} ^{n}}$ to itself that preserve the length of vectors. For example, rotations and reflections belong to ${\displaystyle \mathrm {O} (n)}$. Equivalently, this is the group of n x n orthogonal matrices, meaning that ${\displaystyle A^{\mathrm {T} }=A^{-1}}$, where ${\displaystyle A^{\mathrm {T} }}$ denotes the transpose of a matrix. The orthogonal group has two connected components; the identity component is called the special orthogonal group ${\displaystyle \mathrm {SO} (n)}$, consisting of the orthogonal matrices with determinant 1. Both groups have the same Lie algebra ${\displaystyle {\mathfrak {so}}(n)}$, the subspace of skew-symmetric matrices in ${\displaystyle {\mathfrak {gl}}(n,\mathbb {R} )}$ (${\displaystyle X^{\rm {T}}=-X}$). See also infinitesimal rotations with skew-symmetric matrices.

The complex orthogonal group ${\displaystyle \mathrm {O} (n,\mathbb {C} )}$, its identity component ${\displaystyle \mathrm {SO} (n,\mathbb {C} )}$, and the Lie algebra ${\displaystyle {\mathfrak {so}}(n,\mathbb {C} )}$ are given by the same formulas applied to n x n complex matrices. Equivalently, ${\displaystyle \mathrm {O} (n,\mathbb {C} )}$ is the subgroup of ${\displaystyle \mathrm {GL} (n,\mathbb {C} )}$ that preserves the standard symmetric bilinear form on ${\displaystyle \mathbb {C} ^{n}}$.

• The unitary group ${\displaystyle \mathrm {U} (n)}$ is the subgroup of ${\displaystyle \mathrm {GL} (n,\mathbb {C} )}$ that preserves the length of vectors in ${\displaystyle \mathbb {C} ^{n}}$ (with respect to the standard Hermitian inner product). Equivalently, this is the group of n × n unitary matrices (satisfying ${\displaystyle A^{*}=A^{-1}}$, where ${\displaystyle A^{*}}$ denotes the conjugate transpose of a matrix). Its Lie algebra ${\displaystyle {\mathfrak {u}}(n)}$ consists of the skew-hermitian matrices in ${\displaystyle {\mathfrak {gl}}(n,\mathbb {C} )}$ (${\displaystyle X^{*}=-X}$). This is a Lie algebra over ${\displaystyle \mathbb {R} }$, not over ${\displaystyle \mathbb {C} }$. (Indeed, i times a skew-hermitian matrix is hermitian, rather than skew-hermitian.) Likewise, the unitary group ${\displaystyle \mathrm {U} (n)}$ is a real Lie subgroup of the complex Lie group ${\displaystyle \mathrm {GL} (n,\mathbb {C} )}$. For example, ${\displaystyle \mathrm {U} (1)}$ is the circle group, and its Lie algebra (from this point of view) is ${\displaystyle i\mathbb {R} \subset \mathbb {C} ={\mathfrak {gl}}(1,\mathbb {C} )}$.
• The special unitary group ${\displaystyle \mathrm {SU} (n)}$ is the subgroup of matrices with determinant 1 in ${\displaystyle \mathrm {U} (n)}$. Its Lie algebra ${\displaystyle {\mathfrak {su}}(n)}$ consists of the skew-hermitian matrices with trace zero.
• The symplectic group ${\displaystyle \mathrm {Sp} (2n,\mathbb {R} )}$ is the subgroup of ${\displaystyle \mathrm {GL} (2n,\mathbb {R} )}$ that preserves the standard alternating bilinear form on ${\displaystyle \mathbb {R} ^{2n}}$. Its Lie algebra is the symplectic Lie algebra ${\displaystyle {\mathfrak {sp}}(2n,\mathbb {R} )}$.
• The classical Lie algebras are those listed above, along with variants over any field.

Some Lie algebras of low dimension are described here. See the classification of low-dimensional real Lie algebras for further examples.

• There is a unique nonabelian Lie algebra ${\displaystyle {\mathfrak {g}}}$ of dimension 2 over any field F, up to isomorphism.^[15] Here ${\displaystyle {\mathfrak {g}}}$ has a basis ${\displaystyle X,Y}$ for which the bracket is given by ${\displaystyle \left[X,Y\right]=Y}$. (This determines the Lie bracket completely, because the axioms imply that ${\displaystyle [X,X]=0}$ and ${\displaystyle [Y,Y]=0}$.) Over the real numbers, ${\displaystyle {\mathfrak {g}}}$ can be viewed as the Lie algebra of the Lie group ${\displaystyle G=\mathrm {Aff} (1,\mathbb {R} )}$ of affine transformations of the real line, ${\displaystyle x\mapsto ax+b}$.

The affine group G can be identified with the group of matrices

${\displaystyle \left({\begin{array}{cc}a&b\\0&1\end{array}}\right)}$

under matrix multiplication, with ${\displaystyle a,b\in \mathbb {R} }$, ${\displaystyle a\neq 0}$. Its Lie algebra is the Lie subalgebra ${\displaystyle {\mathfrak {g}}}$ of ${\displaystyle {\mathfrak {gl}}(2,\mathbb {R} )}$ consisting of all matrices

${\displaystyle \left({\begin{array}{cc}c&d\\0&0\end{array}}\right).}$

In these terms, the basis above for ${\displaystyle {\mathfrak {g}}}$ is given by the matrices

${\displaystyle X=\left({\begin{array}{cc}1&0\\0&0\end{array}}\right),\qquad Y=\left({\begin{array}{cc}0&1\\0&0\end{array}}\right).}$

For any field ${\displaystyle F}$, the 1-dimensional subspace ${\displaystyle F\cdot Y}$ is an ideal in the 2-dimensional Lie algebra ${\displaystyle {\mathfrak {g}}}$, by the formula ${\displaystyle [X,Y]=Y\in F\cdot Y}$. Both of the Lie algebras ${\displaystyle F\cdot Y}$ and ${\displaystyle {\mathfrak {g}}/(F\cdot Y)}$ are abelian (because 1-dimensional). In this sense, ${\displaystyle {\mathfrak {g}}}$ can be broken into abelian "pieces", meaning that it is solvable (though not nilpotent), in the terminology below.

• The Heisenberg algebra ${\displaystyle {\mathfrak {h}}_{3}(F)}$ over a field F is the three-dimensional Lie algebra with a basis ${\displaystyle X,Y,Z}$ such that^[16]

${\displaystyle [X,Y]=Z,\quad [X,Z]=0,\quad [Y,Z]=0}$.

It can be viewed as the Lie algebra of 3×3 strictly upper-triangular matrices, with the commutator Lie bracket and the basis

${\displaystyle X=\left({\begin{array}{ccc}0&1&0\\0&0&0\\0&0&0\end{array}}\right),\quad Y=\left({\begin{array}{ccc}0&0&0\\0&0&1\\0&0&0\end{array}}\right),\quad Z=\left({\begin{array}{ccc}0&0&1\\0&0&0\\0&0&0\end{array}}\right)~.\quad }$

Over the real numbers, ${\displaystyle {\mathfrak {h}}_{3}(\mathbb {R} )}$ is the Lie algebra of the Heisenberg group ${\displaystyle \mathrm {H} _{3}(\mathbb {R} )}$, that is, the group of matrices

${\displaystyle \left({\begin{array}{ccc}1&a&c\\0&1&b\\0&0&1\end{array}}\right)}$

under matrix multiplication.

For any field F, the center of ${\displaystyle {\mathfrak {h}}_{3}(F)}$ is the 1-dimensional ideal ${\displaystyle F\cdot Z}$, and the quotient ${\displaystyle {\mathfrak {h}}_{3}(F)/(F\cdot Z)}$ is abelian, isomorphic to ${\displaystyle F^{2}}$. In the terminology below, it follows that ${\displaystyle {\mathfrak {h}}_{3}(F)}$ is nilpotent (though not abelian).

• The Lie algebra ${\displaystyle {\mathfrak {so}}(3)}$ of the rotation group SO(3) is the space of skew-symmetric 3 x 3 matrices over ${\displaystyle \mathbb {R} }$. A basis is given by the three matrices^[17]

${\displaystyle F_{1}=\left({\begin{array}{ccc}0&0&0\\0&0&-1\\0&1&0\end{array}}\right),\quad F_{2}=\left({\begin{array}{ccc}0&0&1\\0&0&0\\-1&0&0\end{array}}\right),\quad F_{3}=\left({\begin{array}{ccc}0&-1&0\\1&0&0\\0&0&0\end{array}}\right)~.\quad }$

The commutation relations among these generators are

${\displaystyle [F_{1},F_{2}]=F_{3},}$

${\displaystyle [F_{2},F_{3}]=F_{1},}$

${\displaystyle [F_{3},F_{1}]=F_{2}.}$

The cross product of vectors in ${\displaystyle \mathbb {R} ^{3}}$ is given by the same formula in terms of the standard basis; so that Lie algebra is isomorphic to ${\displaystyle {\mathfrak {so}}(3)}$. Also, ${\displaystyle {\mathfrak {so}}(3)}$ is equivalent to the Spin (physics) angular-momentum component operators for spin-1 particles in quantum mechanics.^[18]

The Lie algebra ${\displaystyle {\mathfrak {so}}(3)}$ cannot be broken into pieces in the way that the previous examples can: it is simple, meaning that it is not abelian and its only ideals are 0 and all of ${\displaystyle {\mathfrak {so}}(3)}$.

• Another simple Lie algebra of dimension 3, in this case over ${\displaystyle \mathbb {C} }$, is the space ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$ of 2 x 2 matrices of trace zero. A basis is given by the three matrices

${\displaystyle H=\left({\begin{array}{cc}1&0\\0&-1\end{array}}\right),\ E=\left({\begin{array}{cc}0&1\\0&0\end{array}}\right),\ F=\left({\begin{array}{cc}0&0\\1&0\end{array}}\right).}$

H

E

F

The action of ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$ on the Riemann sphere ${\displaystyle \mathbb {CP} ^{1}}$. In particular, the Lie brackets of the vector fields shown are: ${\displaystyle [H,E]=2E}$, ${\displaystyle [H,F]=-2F}$, ${\displaystyle [E,F]=H}$.

The Lie bracket is given by:

${\displaystyle [H,E]=2E,}$

${\displaystyle [H,F]=-2F,}$

${\displaystyle [E,F]=H.}$

Using these formulas, one can show that the Lie algebra ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$ is simple, and classify its finite-dimensional representations (defined below).^[19] In the terminology of quantum mechanics, one can think of E and F as raising and lowering operators. Indeed, for any representation of ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$, the relations above imply that E maps the c-eigenspace of H (for a complex number c) into the ${\displaystyle (c+2)}$-eigenspace, while F maps the c-eigenspace into the ${\displaystyle (c-2)}$-eigenspace.

The Lie algebra ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$ is isomorphic to the complexification of ${\displaystyle {\mathfrak {so}}(3)}$, meaning the tensor product ${\displaystyle {\mathfrak {so}}(3)\otimes _{\mathbb {R} }\mathbb {C} }$. The formulas for the Lie bracket are easier to analyze in the case of ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$. As a result, it is common to analyze complex representations of the group ${\displaystyle \mathrm {SO} (3)}$ by relating them to representations of the Lie algebra ${\displaystyle {\mathfrak {sl}}(2,\mathbb {C} )}$.

• The Lie algebra of vector fields on a smooth manifold of positive dimension is an infinite-dimensional Lie algebra over ${\displaystyle \mathbb {R} }$.
• The Kac–Moody algebras are a large class of infinite-dimensional Lie algebras, say over ${\displaystyle \mathbb {C} }$, with structure much like that of the finite-dimensional simple Lie algebras (such as ${\displaystyle {\mathfrak {sl}}(n,\mathbb {C} )}$).
• The Moyal algebra is an infinite-dimensional Lie algebra that contains all the classical Lie algebras as subalgebras.
• The Virasoro algebra is important in string theory.
• The functor that takes a Lie algebra over a field F to the underlying vector space has a left adjoint ${\displaystyle V\mapsto L(V)}$, called the free Lie algebra on a vector space V. It is spanned by all iterated Lie brackets of elements of V, modulo only the relations coming from the definition of a Lie algebra. The free Lie algebra ${\displaystyle L(V)}$ is infinite-dimensional for V of dimension at least 2.^[20]

Given a vector space V, let ${\displaystyle {\mathfrak {gl}}(V)}$ denote the Lie algebra consisting of all linear maps from V to itself, with bracket given by ${\displaystyle [X,Y]=XY-YX}$. A representation of a Lie algebra ${\displaystyle {\mathfrak {g}}}$ on V is a Lie algebra homomorphism

${\displaystyle \pi \colon {\mathfrak {g}}\to {\mathfrak {gl}}(V).}$

That is, ${\displaystyle \pi }$ sends each element of ${\displaystyle {\mathfrak {g}}}$ to a linear map from V to itself, in such a way that the Lie bracket on ${\displaystyle {\mathfrak {g}}}$ corresponds to the commutator of linear maps.

A representation is said to be faithful if its kernel is zero. Ado's theorem states that every finite-dimensional Lie algebra over a field of characteristic zero has a faithful representation on a finite-dimensional vector space. Kenkichi Iwasawa extended this result to finite-dimensional Lie algebras over a field of any characteristic.^[21] Equivalently, every finite-dimensional Lie algebra over a field F is isomorphic to a Lie subalgebra of ${\displaystyle {\mathfrak {gl}}(n,F)}$ for some positive integer n.

For any Lie algebra ${\displaystyle {\mathfrak {g}}}$, the adjoint representation is the representation

${\displaystyle \operatorname {ad} \colon {\mathfrak {g}}\to {\mathfrak {gl}}({\mathfrak {g}})}$

given by ${\displaystyle \operatorname {ad} (x)(y)=[x,y]}$. (This is a representation of ${\displaystyle {\mathfrak {g}}}$ by the Jacobi identity.)

One important aspect of the study of Lie algebras (especially semisimple Lie algebras, as defined below) is the study of their representations. Although Ado's theorem is an important result, the primary goal of representation theory is not to find a faithful representation of a given Lie algebra ${\displaystyle {\mathfrak {g}}}$. Indeed, in the semisimple case, the adjoint representation is already faithful. Rather, the goal is to understand all possible representations of ${\displaystyle {\mathfrak {g}}}$. For a semisimple Lie algebra over a field of characteristic zero, Weyl's theorem^[22] says that every finite-dimensional representation is a direct sum of irreducible representations (those with no nontrivial invariant subspaces). The finite-dimensional irreducible representations are well understood from several points of view; see the representation theory of semisimple Lie algebras and the Weyl character formula.

The functor that takes an associative algebra A over a field F to A as a Lie algebra (by ${\displaystyle [X,Y]:=XY-YX}$) has a left adjoint ${\displaystyle {\mathfrak {g}}\mapsto U({\mathfrak {g}})}$, called the universal enveloping algebra. To construct this: given a Lie algebra ${\displaystyle {\mathfrak {g}}}$ over F, let

${\displaystyle T({\mathfrak {g}})=F\oplus {\mathfrak {g}}\oplus ({\mathfrak {g}}\otimes {\mathfrak {g}})\oplus ({\mathfrak {g}}\otimes {\mathfrak {g}}\otimes {\mathfrak {g}})\oplus \cdots }$

be the tensor algebra on ${\displaystyle {\mathfrak {g}}}$, also called the free associative algebra on the vector space ${\displaystyle {\mathfrak {g}}}$. Here ${\displaystyle \otimes }$ denotes the tensor product of F-vector spaces. Let I be the two-sided ideal in ${\displaystyle T({\mathfrak {g}})}$ generated by the elements ${\displaystyle XY-YX-[X,Y]}$ for ${\displaystyle X,Y\in {\mathfrak {g}}}$; then the universal enveloping algebra is the quotient ring ${\displaystyle U({\mathfrak {g}})=T({\mathfrak {g}})/I}$. It satisfies the Poincaré–Birkhoff–Witt theorem: if ${\displaystyle e_{1},\ldots ,e_{n}}$ is a basis for ${\displaystyle {\mathfrak {g}}}$ as an F-vector space, then a basis for ${\displaystyle U({\mathfrak {g}})}$ is given by all ordered products ${\displaystyle e_{1}^{i_{1}}\cdots e_{n}^{i_{n}}}$ with ${\displaystyle i_{1},\ldots ,i_{n}}$ natural numbers. In particular, the map ${\displaystyle {\mathfrak {g}}\to U({\mathfrak {g}})}$ is injective.^[23]

Representations of ${\displaystyle {\mathfrak {g}}}$ are equivalent to modules over the universal enveloping algebra. The fact that ${\displaystyle {\mathfrak {g}}\to U({\mathfrak {g}})}$ is injective implies that every Lie algebra (possibly of infinite dimension) has a faithful representation (of infinite dimension), namely its representation on ${\displaystyle U({\mathfrak {g}})}$. This also shows that every Lie algebra is contained in the Lie algebra associated to some associative algebra.

The representation theory of Lie algebras plays an important role in various parts of theoretical physics. There, one considers operators on the space of states that satisfy certain natural commutation relations. These commutation relations typically come from a symmetry of the problem—specifically, they are the relations of the Lie algebra of the relevant symmetry group. An example is the angular momentum operators, whose commutation relations are those of the Lie algebra ${\displaystyle {\mathfrak {so}}(3)}$ of the rotation group ${\displaystyle \mathrm {SO} (3)}$. Typically, the space of states is far from being irreducible under the pertinent operators, but one can attempt to decompose it into irreducible pieces. In doing so, one needs to know the irreducible representations of the given Lie algebra. In the study of the hydrogen atom, for example, quantum mechanics textbooks classify (more or less explicitly) the finite-dimensional irreducible representations of the Lie algebra ${\displaystyle {\mathfrak {so}}(3)}$.^[18]

Lie algebras can be classified to some extent. This is a powerful approach to the classification of Lie groups.

Analogously to abelian, nilpotent, and solvable groups, one can define abelian, nilpotent, and solvable Lie algebras.

A Lie algebra ${\displaystyle {\mathfrak {g}}}$ is abelian if the Lie bracket vanishes; that is, [x,y] = 0 for all x and y in ${\displaystyle {\mathfrak {g}}}$. In particular, the Lie algebra of an abelian Lie group (such as the group ${\displaystyle \mathbb {R} ^{n}}$ under addition or the torus group ${\displaystyle \mathbb {T} ^{n}}$) is abelian. Every finite-dimensional abelian Lie algebra over a field ${\displaystyle F}$ is isomorphic to ${\displaystyle F^{n}}$ for some ${\displaystyle n\geq 0}$, meaning an n-dimensional vector space with Lie bracket zero.

A more general class of Lie algebras is defined by the vanishing of all commutators of given length. First, the commutator subalgebra (or derived subalgebra) of a Lie algebra ${\displaystyle {\mathfrak {g}}}$ is ${\displaystyle [{\mathfrak {g}},{\mathfrak {g}}]}$, meaning the linear subspace spanned by all brackets ${\displaystyle [x,y]}$ with ${\displaystyle x,y\in {\mathfrak {g}}}$. The commutator subalgebra is an ideal in ${\displaystyle {\mathfrak {g}}}$, in fact the smallest ideal such that the quotient Lie algebra is abelian. It is analogous to the commutator subgroup of a group.

A Lie algebra ${\displaystyle {\mathfrak {g}}}$ is nilpotent if the lower central series

${\displaystyle {\mathfrak {g}}\supseteq [{\mathfrak {g}},{\mathfrak {g}}]\supseteq [[{\mathfrak {g}},{\mathfrak {g}}],{\mathfrak {g}}]\supseteq [[[{\mathfrak {g}},{\mathfrak {g}}],{\mathfrak {g}}],{\mathfrak {g}}]\supseteq \cdots }$

becomes zero after finitely many steps. Equivalently, ${\displaystyle {\mathfrak {g}}}$ is nilpotent if there is a finite sequence of ideals in ${\displaystyle {\mathfrak {g}}}$,

${\displaystyle 0={\mathfrak {a}}_{0}\subseteq {\mathfrak {a}}_{1}\subseteq \cdots \subseteq {\mathfrak {a}}_{r}={\mathfrak {g}},}$

such that ${\displaystyle {\mathfrak {a}}_{j}/{\mathfrak {a}}_{j-1}}$ is central in ${\displaystyle {\mathfrak {g}}/{\mathfrak {a}}_{j-1}}$ for each j. By Engel's theorem, a Lie algebra over any field is nilpotent if and only if for every u in ${\displaystyle {\mathfrak {g}}}$ the adjoint endomorphism

${\displaystyle \operatorname {ad} (u):{\mathfrak {g}}\to {\mathfrak {g}},\quad \operatorname {ad} (u)v=[u,v]}$

is nilpotent.^[24]

More generally, a Lie algebra ${\displaystyle {\mathfrak {g}}}$ is said to be solvable if the derived series:

${\displaystyle {\mathfrak {g}}\supseteq [{\mathfrak {g}},{\mathfrak {g}}]\supseteq [[{\mathfrak {g}},{\mathfrak {g}}],[{\mathfrak {g}},{\mathfrak {g}}]]\supseteq [[[{\mathfrak {g}},{\mathfrak {g}}],[{\mathfrak {g}},{\mathfrak {g}}]],[[{\mathfrak {g}},{\mathfrak {g}}],[{\mathfrak {g}},{\mathfrak {g}}]]]\supseteq \cdots }$