{
  "id": "lie-algebra",
  "title": "Lie Algebra",
  "lang": "en",
  "source_lang": "en",
  "summary": "A Lie algebra is a vector space with a bracket operation that encodes infinitesimal symmetries and satisfies the Jacobi identity.",
  "category": "mathematics",
  "subcategory": "algebra",
  "type": "concept",
  "aliases": [
    "Infinitesimal Lie algebra"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/lie-algebra/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/lie-algebra/",
      "json": "https://aiwiki.page/zh/lie-algebra.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A Lie algebra is a vector space equipped with a bilinear operation, called the Lie bracket, that is alternating and satisfies the Jacobi identity. These axioms abstract the algebraic behavior of commutators of matrices and differential operators. Lie algebras provide a language for studying continuous symmetry: whereas a Lie group describes finite transformations, its Lie algebra describes the infinitesimal transformations near the identity.\n\nNamed after Sophus Lie, Lie algebras connect abstract algebra, differential geometry, and representation theory. They also organize many of the symmetry principles used in classical mechanics and quantum mechanics. Their theory includes a complete classification of finite-dimensional complex semisimple Lie algebras, extensive methods for constructing representations, and much less uniform classification problems for solvable, infinite-dimensional, and positive-characteristic examples."
    },
    {
      "heading": "Definition and algebraic language",
      "text": "Let \\(k\\) be a field. A Lie algebra over \\(k\\) is a \\(k\\)-vector space \\(\\mathfrak g\\) together with a bilinear map\n\n\\[[\\ ,\\ ]:\\mathfrak g\\times\\mathfrak g\\longrightarrow\\mathfrak g\\]\n\nsatisfying\n\n\\[[x,x]=0\\]\n\nand\n\n\\[[x,[y,z]]+[y,[z,x]]+[z,[x,y]]=0\\]\n\nfor all \\(x,y,z\\in\\mathfrak g\\). The second equation is the Jacobi identity. Alternation implies \\([x,y]=-[y,x]\\) by expanding \\([x+y,x+y]\\). Over a field of characteristic other than \\(2\\), this skew-symmetry is equivalent to alternation. In characteristic \\(2\\), alternation must be imposed separately; skew-symmetry alone is not sufficient.\n\nThe bracket need not be associative. Instead, the Jacobi identity controls its failure to behave like an associative product. One particularly useful rearrangement is\n\n\\[[x,[y,z]]=[[x,y],z]+[y,[x,z]].\\]\n\nThus, for a fixed \\(x\\), the map \\(\\operatorname{ad}_x(y)=[x,y]\\) acts as a derivation of the bracket: it differentiates a bracket by acting on each argument. This identity is central to the theory because it turns the algebra's own multiplication into a family of symmetry operators.\n\nA linear subspace \\(\\mathfrak h\\subseteq\\mathfrak g\\) is a Lie subalgebra if \\([\\mathfrak h,\\mathfrak h]\\subseteq\\mathfrak h\\). It is an ideal if the stronger condition \\([\\mathfrak g,\\mathfrak h]\\subseteq\\mathfrak h\\) holds. Ideals allow the construction of quotient Lie algebras: the quotient vector space \\(\\mathfrak g/\\mathfrak h\\) has the bracket\n\n\\[[x+\\mathfrak h,y+\\mathfrak h]=[x,y]+\\mathfrak h.\\]\n\nThe ideal condition makes this expression independent of the chosen representatives. A Lie algebra homomorphism is a linear map \\(f\\) preserving brackets, so that \\(f([x,y])=[f(x),f(y)]\\). Its kernel is an ideal, and the usual first isomorphism theorem applies.\n\nIn finite dimension, a choice of basis \\(e_1,\\ldots,e_n\\) describes the entire bracket through structure constants:\n\n\\[[e_i,e_j]=\\sum_{r=1}^{n}c_{ij}^{r}e_r.\\]\n\nAlternation imposes linear conditions on these constants, while Jacobi imposes quadratic conditions. Changing basis generally changes the constants without changing the Lie algebra up to isomorphism. Consequently, classification is not simply a matter of listing solutions of equations: it also requires identifying solutions related by invertible changes of coordinates. This perspective connects Lie algebra classification with algebraic varieties, group orbits, and deformation problems."
    },
    {
      "heading": "Matrix algebras and fundamental examples",
      "text": "Every associative algebra becomes a Lie algebra when equipped with the commutator\n\n\\[[x,y]=xy-yx.\\]\n\nThe Jacobi identity follows by expanding the three double commutators and canceling terms. The most familiar example is \\(\\mathfrak{gl}_n(k)\\), the space of all \\(n\\times n\\) matrices over \\(k\\). Its underlying vector space has dimension \\(n^2\\). Matrix multiplication itself is associative, but the induced Lie bracket is generally not. An arbitrary Lie algebra need not come with an associative multiplication, even though many important examples admit faithful realizations by matrices.\n\nSeveral classical families arise by imposing matrix conditions. The special linear algebra\n\n\\[\\mathfrak{sl}_n(k)=\\{X:\\operatorname{tr}X=0\\}\\]\n\nis closed under commutators because \\(\\operatorname{tr}(XY)=\\operatorname{tr}(YX)\\), where \\(\\operatorname{tr}\\) denotes the matrix trace. Over fields of characteristic other than \\(2\\), the orthogonal algebra associated with a nondegenerate symmetric matrix \\(Q\\) is\n\n\\[\\mathfrak{so}(Q)=\\{X:X^{T}Q+QX=0\\}.\\]\n\nFor a nondegenerate alternating matrix \\(J\\) of size \\(2n\\), the analogous condition defines the symplectic algebra\n\n\\[\\mathfrak{sp}_{2n}(k)=\\{X:X^{T}J+JX=0\\}.\\]\n\nThese equations express infinitesimal preservation of bilinear forms.\n\nAn abelian Lie algebra has zero bracket. Every vector space can be made into one, and two finite-dimensional abelian Lie algebras over the same field are isomorphic precisely when their dimensions agree. The simplest nonabelian example has basis \\(h,e\\) with \\([h,e]=e\\). It can be realized by\n\n\\[h=\\begin{pmatrix}1&0\\\\0&0\\end{pmatrix}, \\qquad e=\\begin{pmatrix}0&1\\\\0&0\\end{pmatrix}.\\]\n\nA different basic example is the three-dimensional Heisenberg algebra, with basis \\(p,q,z\\) and only one independent nonzero bracket, \\([p,q]=z\\). The element \\(z\\) commutes with everything, making this algebra a prototype for central extensions and canonical commutation relations.\n\nOver the real numbers, \\(\\mathbb R^3\\) with the cross product is a Lie algebra isomorphic to \\(\\mathfrak{so}_3(\\mathbb R)\\). Another indispensable example is \\(\\mathfrak{sl}_2(k)\\), with basis\n\n\\[e=\\begin{pmatrix}0&1\\\\0&0\\end{pmatrix},\\quad f=\\begin{pmatrix}0&0\\\\1&0\\end{pmatrix},\\quad h=\\begin{pmatrix}1&0\\\\0&-1\\end{pmatrix}\\]\n\nand relations\n\n\\[[h,e]=2e,\\qquad [h,f]=-2f,\\qquad [e,f]=h.\\]\n\nIn characteristic zero, copies of this three-dimensional algebra govern the root structure of every complex semisimple Lie algebra. Its explicit representations also provide the mathematical model for angular momentum in quantum theory."
    },
    {
      "heading": "Infinitesimal symmetry and Lie groups",
      "text": "For a real or complex Lie group \\(G\\), its Lie algebra is the tangent space \\(T_eG\\) at the identity element. To define the bracket intrinsically, each tangent vector is extended uniquely to a left-invariant vector field on \\(G\\). The commutator of two such vector fields is again left-invariant, so evaluating it at the identity defines a bracket on \\(T_eG\\). This construction explains why the Jacobi identity is unavoidable: commutators of derivations already satisfy it.\n\nFor a matrix group, the tangent-space description produces the matrix commutator directly. If \\(X\\) belongs to its Lie algebra, the curve\n\n\\[t\\longmapsto \\exp(tX)\\]\n\nis a one-parameter subgroup. The exponential map is defined for matrices by the convergent series\n\n\\[\\exp X=\\sum_{m=0}^{\\infty}\\frac{X^m}{m!}.\\]\n\nFor an arbitrary finite-dimensional Lie group, it is defined through one-parameter subgroups and their differential equations. Its derivative at zero is the identity, so it identifies neighborhoods of zero and the group identity, though it need not be globally injective or surjective.\n\nThe Baker–Campbell–Hausdorff formula makes the passage between brackets and local group multiplication explicit:\n\n\\[\\log(\\exp X\\exp Y) =X+Y+\\frac12[X,Y] +\\frac1{12}[X,[X,Y]] +\\frac1{12}[Y,[Y,X]]+\\cdots.\\]\n\nFor sufficiently small \\(X,Y\\), the series expresses multiplication entirely through the Lie algebra structure. In a nilpotent Lie algebra, sufficiently long nested brackets vanish, so the series terminates. A related calculation shows that the group commutator begins with the Lie bracket:\n\n\\[\\exp(tX)\\exp(tY)\\exp(-tX)\\exp(-tY) =I+t^2[X,Y]+O(t^3).\\]\n\nThe bracket therefore measures the leading failure of infinitesimal transformations to commute.\n\nLie's third theorem states that every finite-dimensional real Lie algebra is the Lie algebra of a connected, simply connected Lie group, unique up to Lie group isomorphism. Here simple connectedness removes a global ambiguity: distinct groups can have the same infinitesimal structure. For example, \\(SU(2)\\) and \\(SO(3)\\) have isomorphic real Lie algebras, although the first is a double cover of the second. More generally, a connected Lie group is obtained from its simply connected covering group by quotienting by an appropriate discrete central subgroup. Thus the Lie algebra determines local multiplication but does not record all global topology."
    },
    {
      "heading": "Ideals, solvability, and nilpotence",
      "text": "Ideals describe the parts of a Lie algebra that can be separated by quotients. The center\n\n\\[Z(\\mathfrak g)=\\{z:[z,x]=0\\text{ for every }x\\in\\mathfrak g\\}\\]\n\nis an ideal, as is the derived algebra \\([\\mathfrak g,\\mathfrak g]\\), the linear span of all brackets. The quotient \\(\\mathfrak g/[\\mathfrak g,\\mathfrak g]\\) is abelian and is called the abelianization. It captures everything that can be detected by homomorphisms from \\(\\mathfrak g\\) into abelian Lie algebras. The center, by contrast, identifies elements whose adjoint action is zero.\n\nThe derived series is defined by\n\n\\[\\mathfrak g^{(0)}=\\mathfrak g,\\qquad \\mathfrak g^{(r+1)}=[\\mathfrak g^{(r)},\\mathfrak g^{(r)}].\\]\n\nA Lie algebra is solvable if this series eventually becomes zero. The lower central series is\n\n\\[\\gamma_1(\\mathfrak g)=\\mathfrak g,\\qquad \\gamma_{r+1}(\\mathfrak g)=[\\mathfrak g,\\gamma_r(\\mathfrak g)].\\]\n\nIt defines nilpotence when it eventually becomes zero. Nilpotence implies solvability, but not conversely. The two-dimensional algebra with \\([h,e]=e\\) is solvable, while its lower central series stabilizes at \\(ke\\). The Heisenberg algebra is nilpotent of class two because its commutator algebra is nonzero and central.\n\nTwo triangularization theorems reveal why these definitions are useful. Lie's theorem states that a finite-dimensional representation of a solvable Lie algebra over an algebraically closed field of characteristic zero admits a basis in which all representing matrices are upper triangular. Engel's theorem states that if every operator \\(\\operatorname{ad}_x\\) is nilpotent, then \\(\\mathfrak g\\) is nilpotent. Its representation-theoretic form gives simultaneous strict triangularization for Lie algebras consisting of nilpotent endomorphisms. The distinction between triangular and strictly triangular matrices mirrors the distinction between solvable and nilpotent structure.\n\nEvery finite-dimensional Lie algebra has a largest solvable ideal, its radical \\(\\mathfrak r\\), and a largest nilpotent ideal, its nilradical. Over characteristic zero, the Levi decomposition gives\n\n\\[\\mathfrak g=\\mathfrak s\\ltimes\\mathfrak r,\\]\n\nwhere \\(\\mathfrak s\\) is semisimple. The symbol \\(\\ltimes\\) indicates that \\(\\mathfrak s\\) acts on the radical by derivations; it is not generally a direct sum of ideals. Levi subalgebras need not be unique as subspaces, although a conjugacy theorem relates them. This decomposition reduces much structural analysis to semisimple algebras, solvable algebras, and the possible actions connecting them. It does not supply a comparably short classification of all solvable algebras: their isomorphism types already occur in continuous families in relatively modest dimensions."
    },
    {
      "heading": "Semisimple algebras, roots, and classification",
      "text": "A Lie algebra is simple if it is nonabelian and has no nonzero proper ideals. It is semisimple if its solvable radical is zero. In finite dimension over characteristic zero, semisimplicity is equivalent to being a direct sum of simple ideals. A fundamental diagnostic is the Killing form,\n\n\\[B(x,y)=\\operatorname{tr}(\\operatorname{ad}_x\\operatorname{ad}_y).\\]\n\nIt is symmetric and invariant:\n\n\\[B([x,y],z)=B(x,[y,z]).\\]\n\nCartan's criterion states that, in this setting, a Lie algebra is semisimple precisely when its Killing form is nondegenerate. This criterion turns a structural question about ideals into a calculation with a bilinear form.\n\nFor a complex semisimple algebra, choose a Cartan subalgebra \\(\\mathfrak h\\), which in this setting is a maximal abelian subalgebra whose elements act diagonally in the adjoint representation. Simultaneous eigenspaces give the decomposition\n\n\\[\\mathfrak g=\\mathfrak h\\oplus \\bigoplus_{\\alpha\\in\\Phi}\\mathfrak g_\\alpha, \\qquad \\mathfrak g_\\alpha =\\{x:[h,x]=\\alpha(h)x\\text{ for all }h\\in\\mathfrak h\\}.\\]\n\nThe nonzero linear functionals \\(\\alpha\\) are called roots. Each root space is one-dimensional, and brackets obey\n\n\\[[\\mathfrak g_\\alpha,\\mathfrak g_\\beta] \\subseteq\\mathfrak g_{\\alpha+\\beta},\\]\n\nwith the zero-weight space interpreted as \\(\\mathfrak h\\). This converts many bracket calculations into addition of roots.\n\nThe roots form a finite root system in a real Euclidean space. Reflections in root hyperplanes generate the Weyl group, a finite symmetry group that controls conjugacy and representation theory. A choice of positive roots determines simple roots, from which all positive roots are nonnegative integer combinations. Their lengths and relative angles are encoded by a Cartan matrix and, more visually, a Dynkin diagram. The rank of the algebra is the dimension of \\(\\mathfrak h\\), equivalently the number of simple roots.\n\nThe finite-dimensional complex simple Lie algebras belong to four classical families,\n\n\\[A_n=\\mathfrak{sl}_{n+1}(\\mathbb C),\\quad B_n=\\mathfrak{so}_{2n+1}(\\mathbb C),\\quad C_n=\\mathfrak{sp}_{2n}(\\mathbb C),\\quad D_n=\\mathfrak{so}_{2n}(\\mathbb C),\\]\n\nand five exceptional types,\n\n\\[G_2,\\quad F_4,\\quad E_6,\\quad E_7,\\quad E_8.\\]\n\nA nonredundant standard list uses \\(n\\geq1\\) for \\(A_n\\), \\(n\\geq2\\) for \\(B_n\\), \\(n\\geq3\\) for \\(C_n\\), and \\(n\\geq4\\) for \\(D_n\\). Low-rank coincidences include \\(B_2\\cong C_2\\) and \\(D_3\\cong A_3\\). The exceptional dimensions are respectively \\(14,52,78,133,\\) and \\(248\\). For example, \\(G_2\\) can be realized as the derivation algebra of the octonions.\n\nThis classification is specific to the complex semisimple setting. It does not classify all finite-dimensional Lie algebras, nor does it classify real simple algebras without additional information. A semisimple complex algebra is specified by a finite collection of simple factors, but the solvable radical and its interactions introduce substantially more varied behavior in the general case."
    },
    {
      "heading": "Representations and highest weights",
      "text": "A representation of \\(\\mathfrak g\\) on a vector space \\(V\\) is a Lie algebra homomorphism\n\n\\[\\rho:\\mathfrak g\\longrightarrow\\mathfrak{gl}(V).\\]\n\nEquivalently, \\(V\\) is a \\(\\mathfrak g\\)-module satisfying\n\n\\[[x,y]\\cdot v=x\\cdot(y\\cdot v)-y\\cdot(x\\cdot v).\\]\n\nRepresentations translate abstract brackets into linear operators. A representation is irreducible if it has no nonzero proper invariant subspace, and faithful if \\(\\rho\\) is injective. The adjoint representation is \\(\\rho(x)=\\operatorname{ad}_x\\); its kernel is the center.\n\nAdo's theorem states that every finite-dimensional Lie algebra over a field of characteristic zero has a faithful finite-dimensional representation. Thus every such algebra can be realized as a matrix Lie algebra, although its most natural or smallest faithful representation may not be apparent from its definition. For semisimple algebras in characteristic zero, Weyl's complete reducibility theorem gives a stronger simplification: every finite-dimensional module is a direct sum of irreducible modules. This fails for general solvable algebras and in positive characteristic.\n\nThe representation theory of \\(\\mathfrak{sl}_2(\\mathbb C)\\) provides the model. For each integer \\(m\\geq0\\), there is one irreducible representation of dimension \\(m+1\\). Its \\(h\\)-eigenvalues are\n\n\\[m,m-2,\\ldots,-m.\\]\n\nThe operators \\(e\\) and \\(f\\) raise and lower these weights by two. A concrete realization is the space of homogeneous polynomials of degree \\(m\\) in two variables. Tensor products decompose according to\n\n\\[V_m\\otimes V_n \\cong \\bigoplus_{r=0}^{\\min(m,n)}V_{m+n-2r}.\\]\n\nIn angular-momentum notation, these representations correspond to spin \\(j=m/2\\).\n\nFor a general complex semisimple algebra, irreducible finite-dimensional representations are classified by dominant integral highest weights. After choosing positive roots, a highest-weight vector is annihilated by all positive root spaces. Its weight \\(\\lambda\\) satisfies\n\n\\[\\langle\\lambda,\\alpha_i^\\vee\\rangle \\in\\mathbb Z_{\\geq0}\\]\n\nfor every simple coroot \\(\\alpha_i^\\vee\\). Applying negative root operators generates the module. The Weyl dimension formula gives\n\n\\[\\dim V_\\lambda =\\prod_{\\alpha>0} \\frac{\\langle\\lambda+\\rho,\\alpha^\\vee\\rangle} {\\langle\\rho,\\alpha^\\vee\\rangle},\\]\n\nwhere \\(\\rho\\) here denotes half the sum of the positive roots, rather than a representation map.\n\nLie algebra representations and Lie group representations are closely related but are not interchangeable without global conditions. Every finite-dimensional representation of a real Lie algebra integrates to its simply connected Lie group. It descends to a quotient group only when the quotient's discrete kernel acts trivially. Consequently, half-integer-spin representations integrate to \\(SU(2)\\), while only integer-spin representations descend to \\(SO(3)\\). The allowed weight lattice therefore depends on the global group as well as on its Lie algebra."
    },
    {
      "heading": "Enveloping algebras, invariants, and cohomology",
      "text": "The universal enveloping algebra \\(U(\\mathfrak g)\\) is an associative algebra that packages all products of Lie algebra operators. Starting with the tensor algebra \\(T(\\mathfrak g)\\), it is defined by\n\n\\[U(\\mathfrak g)= T(\\mathfrak g)\\big/ \\langle x\\otimes y-y\\otimes x-[x,y]\\rangle.\\]\n\nIts defining universal property says that a Lie algebra homomorphism from \\(\\mathfrak g\\) to the commutator algebra of an associative algebra extends uniquely to an associative algebra homomorphism from \\(U(\\mathfrak g)\\). A representation of \\(\\mathfrak g\\) is therefore the same thing as a unital module over \\(U(\\mathfrak g)\\).\n\nThe Poincaré–Birkhoff–Witt theorem gives a precise description of this construction. If \\(x_1,\\ldots,x_n\\) is an ordered basis of \\(\\mathfrak g\\), then the ordered monomials\n\n\\[x_1^{a_1}\\cdots x_n^{a_n}, \\qquad a_i\\geq0,\\]\n\nform a basis of \\(U(\\mathfrak g)\\). In particular, the natural map from \\(\\mathfrak g\\) into its enveloping algebra is injective. Filtering by degree gives\n\n\\[\\operatorname{gr}U(\\mathfrak g)\\cong S(\\mathfrak g),\\]\n\nthe symmetric algebra. This identifies the enveloping algebra as a noncommutative counterpart of a polynomial algebra, with the bracket specifying the leading correction to commutativity.\n\nElements in the center of \\(U(\\mathfrak g)\\) supply invariant operators. For a semisimple algebra, a basis \\(x_i\\) and its Killing-dual basis \\(x^i\\) define the quadratic Casimir element\n\n\\[\\Omega=\\sum_i x_i x^i.\\]\n\nIt commutes with every element of \\(\\mathfrak g\\), and acts by a scalar on each finite-dimensional irreducible complex representation. Such scalars help distinguish representations and appear as eigenvalues of invariant differential operators. The familiar total angular-momentum operator is an instance of this construction.\n\nLie algebra cohomology studies representations through a cochain complex. For a module \\(V\\), its degree-\\(p\\) cochains are\n\n\\[C^p(\\mathfrak g,V)=\\operatorname{Hom}(\\Lambda^p\\mathfrak g,V).\\]\n\nThe differential combines the bracket with the module action, and its square vanishes because of Jacobi and the representation identity. Degree zero records invariant vectors. Degree one measures derivations modulo inner derivations in the adjoint case, while degree two classifies abelian extensions with a specified module action; for trivial coefficients it classifies central extensions. For finite-dimensional semisimple \\(\\mathfrak g\\) and finite-dimensional \\(V\\) in characteristic zero, Whitehead's lemmas state that \\(H^1(\\mathfrak g,V)\\) and \\(H^2(\\mathfrak g,V)\\) vanish. These results explain important rigidity and splitting phenomena rather than merely furnishing numerical invariants."
    },
    {
      "heading": "Real forms and positive characteristic",
      "text": "A complex classification does not by itself determine the corresponding real algebras. A real form of a complex Lie algebra \\(\\mathfrak g\\) is a real Lie algebra \\(\\mathfrak g_0\\) such that\n\n\\[\\mathfrak g_0\\otimes_{\\mathbb R}\\mathbb C\\cong\\mathfrak g.\\]\n\nFor example, \\(\\mathfrak{sl}_2(\\mathbb R)\\) and \\(\\mathfrak{su}_2\\) are nonisomorphic real forms of \\(\\mathfrak{sl}_2(\\mathbb C)\\). The former is split and noncompact; the latter is compact and can be identified with \\(\\mathfrak{so}_3(\\mathbb R)\\). Their complex representation theory shares a common algebraic framework, but their associated groups have quite different geometry and harmonic analysis.\n\nThe Killing form detects part of this distinction. For a compact real semisimple algebra it is negative definite, while noncompact real forms have indefinite Killing forms. A Cartan involution \\(\\theta\\) of a real semisimple algebra gives a decomposition\n\n\\[\\mathfrak g_0=\\mathfrak k\\oplus\\mathfrak p\\]\n\ninto its \\(+1\\) and \\(-1\\) eigenspaces, with\n\n\\[[\\mathfrak k,\\mathfrak k]\\subseteq\\mathfrak k,\\qquad [\\mathfrak k,\\mathfrak p]\\subseteq\\mathfrak p,\\qquad [\\mathfrak p,\\mathfrak p]\\subseteq\\mathfrak k.\\]\n\nThe form \\(-B(x,\\theta y)\\) is positive definite. This structure links real Lie algebras to Riemannian symmetric spaces and supplies a useful replacement for the positive-definite geometry available in compact cases.\n\nOver a field of positive characteristic \\(p\\), several characteristic-zero theorems require modification or fail. A simple illustration occurs in \\(\\mathfrak{sl}_n(k)\\): if \\(p\\) divides \\(n\\), the identity matrix has trace zero and therefore lies in this algebra's center. Familiar assertions of simplicity must consequently be checked rather than transferred unchanged. The Killing form may also fail to distinguish semisimple behavior in the same way, and finite-dimensional representations of semisimple algebras need not be completely reducible.\n\nA major additional structure is a restricted Lie algebra, equipped with an operation \\(x\\mapsto x^{[p]}\\) compatible with the \\(p\\)-th powers of adjoint operators:\n\n\\[\\operatorname{ad}(x^{[p]})=(\\operatorname{ad}x)^p.\\]\n\nFor Lie algebras of suitable matrix groups, this operation is induced by matrix \\(p\\)-th powers. Restricted enveloping algebras incorporate these extra relations and connect the theory to infinitesimal group schemes. Finite-dimensional simple algebras in positive characteristic include analogues of the classical families as well as Cartan-type families built from derivations and differential forms. Characteristic \\(5\\) also admits the Melikian family. The modular theory is therefore not simply the characteristic-zero classification with coefficients reduced modulo a prime."
    },
    {
      "heading": "Infinite-dimensional algebras and applications",
      "text": "Infinite-dimensional Lie algebras arise naturally from spaces of transformations and differential operators. The smooth vector fields on a manifold form a Lie algebra under their commutator as derivations of smooth functions. If\n\n\\[X=\\sum_i X^i\\partial_i,\\qquad Y=\\sum_i Y^i\\partial_i,\\]\n\nthen\n\n\\[[X,Y] =\\sum_j\\left(\\sum_i X^i\\partial_iY^j-Y^i\\partial_iX^j\\right)\\partial_j.\\]\n\nThis bracket measures the failure of the local flows of \\(X\\) and \\(Y\\) to commute. Symmetry analysis of differential equations uses such algebras to organize transformations preserving their solution sets.\n\nLoop algebras provide another construction:\n\n\\[L\\mathfrak g=\\mathfrak g\\otimes k[t,t^{-1}], \\qquad [x\\otimes t^m,y\\otimes t^n]=[x,y]\\otimes t^{m+n}.\\]\n\nCentral extensions of these algebras, together with degree derivations, yield affine examples of Kac–Moody algebras. More general Kac–Moody algebras are defined from generalized Cartan matrices by generators and relations. They extend aspects of root systems and highest-weight theory beyond finite dimension, though their root sets and representation categories are substantially more complicated.\n\nThe Virasoro algebra is a central extension of the algebra of Laurent-polynomial vector fields on the circle. With generators \\(L_n\\), indexed by integers, and central element \\(c\\), its standard relations are\n\n\\[[L_m,L_n]=(m-n)L_{m+n} +\\frac{m^3-m}{12}\\delta_{m+n,0}c, \\qquad [c,L_n]=0.\\]\n\nThe central term cannot be removed by changing generators. In a representation, \\(c\\) often acts as a scalar called the central charge. Virasoro representations are fundamental in two-dimensional conformal field theory and connect algebraic structure to scaling dimensions and the organization of quantum states.\n\nFinite-dimensional Lie algebras also have direct physical roles. Rotations are governed by \\(\\mathfrak{so}_3\\); relativistic spacetime symmetries involve the Lorentz and Poincaré algebras. Gauge theories use Lie-algebra-valued fields, with the Standard Model's local gauge algebra\n\n\\[\\mathfrak{su}_3\\oplus\\mathfrak{su}_2\\oplus\\mathfrak u_1.\\]\n\nThe algebra specifies local commutation relations, while the global gauge group contains additional information. In Hamiltonian mechanics, observables form a Lie algebra under the Poisson bracket, and quantization relates this bracket to operator commutators without furnishing a universal exact correspondence for all observables. Lie brackets also enter geometric control theory: repeated brackets of available control vector fields reveal directions of motion accessible through combinations of flows, even when those directions are not directly actuated."
    },
    {
      "heading": "Historical development",
      "text": "Sophus Lie developed the theory of continuous transformation groups in the 1870s, partly in pursuit of methods for differential equations analogous to the role of permutation groups in algebraic equations. Infinitesimal transformations made it possible to replace nonlinear questions about transformations by linear calculations with vector fields and their commutators. Lie's work with Friedrich Engel was presented systematically in the three-volume Theorie der Transformationsgruppen, published from 1888 to 1893. The modern axiomatic separation of Lie algebras from transformation groups emerged from this program rather than being its original starting point.\n\nWilhelm Killing investigated the structure and classification of complex simple Lie algebras during the late 1880s. He identified the root configurations underlying the classical and exceptional families, though his arguments contained gaps. Élie Cartan's 1894 doctoral thesis supplied a rigorous classification and developed the structural methods associated with Cartan subalgebras and the Killing form. These achievements revealed that the apparently open-ended problem of complex simple infinitesimal symmetry has a remarkably short answer encoded by a finite list of root-system types.\n\nDuring the early twentieth century, representation theory became a central organizing principle. Hermann Weyl connected semisimple Lie algebras, compact groups, invariant integration, and the representation-theoretic structures used in quantum mechanics. The enveloping-algebra basis theorem bears the names of Henri Poincaré, Garrett Birkhoff, and Ernst Witt, reflecting contributions made at different stages to the relationship between Lie brackets and associative algebras. Igor Ado's theorem in the 1930s established that finite-dimensional characteristic-zero Lie algebras always admit faithful finite-dimensional matrix representations.\n\nLater developments widened both the available fields and the scale of the objects studied. Claude Chevalley's integral constructions made it possible to build related Lie algebras and groups over fields other than the real and complex numbers. Samuel Eilenberg and Chevalley developed Lie algebra cohomology in the 1940s, incorporating extensions and representations into homological algebra. In the 1960s, Victor Kac and Robert Moody independently introduced the generalized algebras now carrying their names; these subsequently became important in representation theory and mathematical physics. The classification of modular simple algebras, the geometry of representation spaces, and the representation theory of infinite-dimensional algebras developed as distinct research programs, each requiring tools beyond the finite-dimensional complex classification."
    }
  ],
  "links": [
    "vector-space",
    "lie-group",
    "abstract-algebra",
    "differential-geometry",
    "representation-theory",
    "classical-mechanics",
    "quantum-mechanics",
    "field-mathematics",
    "linear-subspace",
    "quotient-vector-space",
    "homomorphism",
    "basis-linear-algebra",
    "isomorphism",
    "matrix",
    "matrix-trace",
    "cross-product",
    "tangent-space",
    "vector-field",
    "exponential-function",
    "baker-campbell-hausdorff-formula",
    "universal-property",
    "differential-equation"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}