{
  "id": "lorentz-group",
  "title": "Lorentz Group",
  "lang": "en",
  "source_lang": "en",
  "summary": "The Lorentz group is the group of linear transformations preserving the spacetime interval, governing relativistic geometry, fields, and particle spin.",
  "category": "science",
  "subcategory": "physics",
  "type": "concept",
  "aliases": [
    "Homogeneous Lorentz group",
    "Indefinite orthogonal group O(1,3)"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/lorentz-group/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/lorentz-group/",
      "json": "https://aiwiki.page/zh/lorentz-group.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The Lorentz group is the mathematical group of linear transformations that preserve the spacetime interval of special relativity. It includes spatial rotations, transformations between observers moving at constant relative velocity, and discrete operations such as spatial inversion and reversal of the time coordinate. Named after Hendrik Antoon Lorentz, it expresses the equivalence of inertial descriptions without selecting a preferred direction in space or a preferred state of uniform motion.\n\nIts importance extends beyond changes of coordinates. The Lorentz group determines how relativistic vectors, tensors, and spinors transform; constrains the form of physical equations; and supplies part of the mathematical framework for classifying elementary particles. Its six continuous parameters describe three rotations and three boosts. The subgroup continuously connected to the identity is particularly important in physics, while its double cover explains why spinor states behave differently from ordinary vectors under a full rotation.\n\nThe Lorentz group is not the entire symmetry group of flat spacetime: translations must also be included to obtain the Poincaré group. In curved spacetime, Lorentz symmetry survives locally as the freedom to choose orthonormal frames, even when the spacetime has no global symmetries."
    },
    {
      "heading": "Definition, invariant interval, and connected components",
      "text": "Consider four-dimensional spacetime with coordinates \\(x^\\mu=(ct,x,y,z)\\), where \\(c\\) is the speed of light. Using the metric convention\n\n\\[\\eta=\\operatorname{diag}(1,-1,-1,-1),\\]\n\nthe squared interval between neighboring events is\n\n\\[ds^2=\\eta_{\\mu\\nu}\\,dx^\\mu dx^\\nu =c^2dt^2-dx^2-dy^2-dz^2.\\]\n\nRepeated Greek indices are summed from \\(0\\) to \\(3\\). A linear transformation \\(x'^\\mu=\\Lambda^\\mu{}_\\nu x^\\nu\\) is Lorentz precisely when its real matrix satisfies\n\n\\[\\Lambda^{\\mathsf T}\\eta\\Lambda=\\eta.\\]\n\nThese matrices form \\(O(1,3)\\). The alternative notation \\(O(3,1)\\) commonly accompanies the opposite metric convention and describes an isomorphic group.\n\nThe defining equation ensures closure under multiplication and gives the inverse explicitly:\n\n\\[\\Lambda^{-1}=\\eta^{-1}\\Lambda^{\\mathsf T}\\eta.\\]\n\nTaking determinants shows that \\((\\det\\Lambda)^2=1\\), so every Lorentz transformation has determinant \\(+1\\) or \\(-1\\). The preservation of an indefinite metric distinguishes this group from the ordinary orthogonal group of Euclidean space. It preserves the division of nonzero vectors into timelike, null, and spacelike classes, rather than preserving a positive spatial length.\n\nThere are four connected components. Two independent distinctions label them: whether \\(\\det\\Lambda\\) is positive, and whether future-directed timelike vectors remain future-directed. In the stated convention, the latter condition is \\(\\Lambda^0{}_0\\geq1\\); transformations reversing time orientation have \\(\\Lambda^0{}_0\\leq-1\\). A transformation with determinant \\(+1\\) is called proper, and one preserving time orientation is called orthochronous. The identity component is the proper orthochronous Lorentz group, written \\(SO^+(1,3)\\), \\(SO_0(1,3)\\), or \\(L_+^\\uparrow\\).\n\nSpatial inversion and time-coordinate reversal have representative matrices\n\n\\[P=\\operatorname{diag}(1,-1,-1,-1),\\qquad T=\\operatorname{diag}(-1,1,1,1).\\]\n\nMultiplying the identity component by \\(P\\), \\(T\\), or \\(PT\\) produces the other three components. Notably, \\(PT=-I\\) has determinant \\(+1\\) but reverses time orientation. Thus “proper” alone does not mean continuously connected to the identity.\n\nA Lorentz transformation fixes the coordinate origin. A more general interval-preserving change of inertial coordinates has the form \\(x'=\\Lambda x+a\\), with a translation \\(a\\). Keeping these linear and affine transformations distinct is essential: the Lorentz group has six continuous parameters, whereas the full group of flat-spacetime isometries has ten."
    },
    {
      "heading": "Rotations, boosts, and rapidity",
      "text": "Ordinary spatial rotations appear as Lorentz transformations of the form\n\n\\[\\Lambda_R= \\begin{pmatrix} 1&0\\\\ 0&R \\end{pmatrix}, \\qquad R\\in SO(3).\\]\n\nThey leave the time coordinate unchanged. The other continuous transformations are boosts, which relate inertial reference frames in relative motion. For a passive change to a frame moving at velocity \\(v\\) along the \\(x\\)-axis,\n\n\\[ct'=\\gamma(ct-\\beta x),\\qquad x'=\\gamma(x-\\beta ct),\\]\n\nwith \\(y'=y\\), \\(z'=z\\), \\(\\beta=v/c\\), and \\(\\gamma=(1-\\beta^2)^{-1/2}\\). These equations preserve \\(c^2t^2-x^2\\) directly. Active transformations of physical vectors can use the inverse matrix, so boost signs must always be interpreted with their convention.\n\nA useful parameter is rapidity, denoted \\(\\chi\\), defined by\n\n\\[\\beta=\\tanh\\chi,\\qquad \\gamma=\\cosh\\chi,\\qquad \\gamma\\beta=\\sinh\\chi.\\]\n\nThe boost matrix in the time–\\(x\\) plane becomes\n\n\\[B_x(\\chi)= \\begin{pmatrix} \\cosh\\chi&-\\sinh\\chi\\\\ -\\sinh\\chi&\\cosh\\chi \\end{pmatrix}.\\]\n\nUnlike velocity, rapidity ranges over the whole real line. Collinear boosts add their rapidities:\n\n\\[B_x(\\chi_1)B_x(\\chi_2)=B_x(\\chi_1+\\chi_2).\\]\n\nThe corresponding velocity-composition law is\n\n\\[v_{\\mathrm{combined}} =\\frac{v_1+v_2}{1+v_1v_2/c^2}.\\]\n\nFor example, two successive boosts of speed \\(0.6c\\) in the same direction combine to a speed of approximately \\(0.882c\\), not \\(1.2c\\).\n\nEvery proper orthochronous Lorentz transformation can be factored into a rotation and a rotation-free boost. Once the order of the factors is fixed, the factorization is unique. Nevertheless, boosts do not form a subgroup: composing boosts along different directions generally produces a transformation containing a rotation. Reversing the order changes the result. This noncommutativity produces the Wigner rotation, and its infinitesimal accumulation along an accelerated trajectory underlies Thomas precession.\n\nThe geometry of boosts is hyperbolic rather than circular. A rotation mixes two spatial coordinates through sine and cosine while preserving a sum of squares; a boost mixes time and space through hyperbolic functions while preserving a difference of squares. The limit \\(|v|\\to c\\) corresponds to \\(|\\chi|\\to\\infty\\). Consequently, no finite boost transforms the rest frame of a massive object into a frame moving at light speed. For \\(|v|\\ll c\\), the boost equations approach Galilean kinematics, with relativity-of-simultaneity and higher-order velocity corrections becoming negligible in the appropriate limit."
    },
    {
      "heading": "Lie algebra and infinitesimal structure",
      "text": "The Lorentz group is a six-dimensional Lie group: its continuous transformations form a smooth manifold, and multiplication and inversion are smooth operations. Its local structure is encoded by its Lie algebra, \\(\\mathfrak{so}(1,3)\\). Writing a transformation close to the identity as \\(\\Lambda=I+\\omega+\\cdots\\), the metric-preservation condition gives\n\n\\[\\omega^{\\mathsf T}\\eta+\\eta\\omega=0.\\]\n\nEquivalently, the lowered object \\(\\omega_{\\mu\\nu}=\\eta_{\\mu\\rho}\\omega^\\rho{}_\\nu\\) is antisymmetric. An antisymmetric \\(4\\times4\\) matrix has six independent components, accounting for three rotations and three boosts.\n\nIn the conventional physics normalization, rotation generators \\(J_i\\) and boost generators \\(K_i\\) obey\n\n\\[[J_i,J_j]=i\\epsilon_{ijk}J_k,\\]\n\n\\[[J_i,K_j]=i\\epsilon_{ijk}K_k,\\]\n\n\\[[K_i,K_j]=-i\\epsilon_{ijk}J_k,\\]\n\nwhere \\(\\hbar=1\\), and \\(\\epsilon_{ijk}\\) is the three-dimensional antisymmetric symbol. For a representation \\(D\\), a transformation near the identity may be written\n\n\\[D(\\Lambda)= \\exp\\!\\left[-i\\left(\\boldsymbol{\\theta}\\cdot\\mathbf J+ \\boldsymbol{\\xi}\\cdot\\mathbf K\\right)\\right],\\]\n\nwith signs of the parameters chosen consistently with the active or passive convention. In real matrix conventions the factors of \\(i\\) are absorbed into the generators.\n\nThe first commutator is the familiar algebra of angular momentum. The second states that boosts transform as spatial vectors under rotations. The third is characteristic of relativistic geometry: two infinitesimal boosts in different directions generate a rotation, with the opposite sign from the rotation–rotation relation. For small boost parameters \\(\\epsilon\\) and \\(\\delta\\), the group commutator\n\n\\[e^{\\epsilon X}e^{\\delta Y}e^{-\\epsilon X}e^{-\\delta Y} =I+\\epsilon\\delta[X,Y]+\\cdots\\]\n\nexhibits this rotational effect directly. The associated correction is second order in small velocities.\n\nThe real Lorentz algebra is noncompact and simple, but its complexification splits into two commuting factors. Define\n\n\\[A_i=\\frac{J_i+iK_i}{2},\\qquad B_i=\\frac{J_i-iK_i}{2}.\\]\n\nThen\n\n\\[[A_i,A_j]=i\\epsilon_{ijk}A_k,\\qquad [B_i,B_j]=i\\epsilon_{ijk}B_k,\\qquad [A_i,B_j]=0.\\]\n\nThis decomposition is the basis of the two-spin labeling of finite-dimensional complex representations. It is a statement about the complexified algebra, not an assertion that the real Lorentz group is a product of two ordinary rotation groups. In particular, it does not remove the noncompact boost directions."
    },
    {
      "heading": "Topology, double covering, and spinors",
      "text": "The local commutation relations do not determine all global properties of a group. The connected Lorentz group is not simply connected: its fundamental group is\n\n\\[\\pi_1\\!\\left(SO^+(1,3)\\right)\\cong\\mathbb Z_2.\\]\n\nIts universal covering group is the special linear group \\(SL(2,\\mathbb C)\\), consisting of complex \\(2\\times2\\) matrices with determinant one. The map from this group onto \\(SO^+(1,3)\\) is two-to-one, with kernel \\(\\{I,-I\\}\\). In spin-group notation,\n\n\\[\\operatorname{Spin}^+(1,3)\\cong SL(2,\\mathbb C).\\]\n\nThe covering map has an explicit construction. Associate a real four-vector with the Hermitian matrix\n\n\\[X=x^\\mu\\sigma_\\mu= \\begin{pmatrix} x^0+x^3&x^1-ix^2\\\\ x^1+ix^2&x^0-x^3 \\end{pmatrix},\\]\n\nwhere \\(\\sigma_0=I\\) and the remaining matrices are the Pauli matrices. Its determinant is\n\n\\[\\det X=(x^0)^2-(x^1)^2-(x^2)^2-(x^3)^2.\\]\n\nFor \\(A\\in SL(2,\\mathbb C)\\), the transformation\n\n\\[X\\longmapsto AXA^\\dagger\\]\n\npreserves Hermiticity and determinant, and therefore induces a Lorentz transformation. Both \\(A\\) and \\(-A\\) induce exactly the same transformation of \\(X\\).\n\nThe subgroup \\(SU(2)\\subset SL(2,\\mathbb C)\\) covers spatial rotations. A continuous \\(2\\pi\\) rotation lifts to a path ending at \\(-I\\), whereas a \\(4\\pi\\) rotation lifts to a closed path ending at \\(I\\). This accounts for the sign reversal of a spin-\\(\\tfrac12\\) spinor under a full spatial rotation. The sign reversal is not a reversal of a classical vector. Although an overall sign of an isolated state does not change its observable probabilities, relative phases can matter in interference.\n\nThe topology can also be understood through polar decomposition. Every \\(A\\in SL(2,\\mathbb C)\\) can be written uniquely as \\(A=HU\\), with \\(H\\) positive Hermitian of determinant one and \\(U\\in SU(2)\\). The three real parameters in \\(H\\) describe boosts and form a space topologically equivalent to \\(\\mathbb R^3\\). Correspondingly, \\(SO^+(1,3)\\) has the topology of \\(SO(3)\\times\\mathbb R^3\\), although this is not a direct-product decomposition of groups.\n\nThis distinction between the Lorentz group and its cover is indispensable in physics. Ordinary tensors carry representations of \\(SO^+(1,3)\\), while spinors generally require \\(SL(2,\\mathbb C)\\). Extending spinors to include reflections requires additional structure, described by appropriate Pin groups rather than by the connected spin group alone."
    },
    {
      "heading": "Finite-dimensional representations and relativistic fields",
      "text": "A representation assigns a linear transformation of a vector space to each group element while preserving multiplication. Finite-dimensional complex irreducible representations of the Lorentz Lie algebra, and of its simply connected group \\(SL(2,\\mathbb C)\\), are labeled by\n\n\\[(j_L,j_R),\\]\n\nwhere each label is a nonnegative integer or half-integer. Their complex dimension is\n\n\\[(2j_L+1)(2j_R+1).\\]\n\nThe labels arise from the two commuting angular-momentum algebras in the complexified Lorentz algebra. They describe transformation properties, rather than particle masses or interactions.\n\nThe simplest examples organize familiar relativistic quantities. A scalar transforms as \\((0,0)\\). The two inequivalent two-component Weyl spinors transform as \\((\\tfrac12,0)\\) and \\((0,\\tfrac12)\\). A four-vector transforms as \\((\\tfrac12,\\tfrac12)\\). A four-component Dirac spinor carries the reducible representation\n\n\\[(\\tfrac12,0)\\oplus(0,\\tfrac12).\\]\n\nSpatial parity interchanges the two Weyl representations, explaining why a single Weyl representation does not by itself furnish a parity-invariant spinor multiplet.\n\nUnder the rotation subgroup, \\((j_L,j_R)\\) contains angular momenta\n\n\\[j=|j_L-j_R|,\\ |j_L-j_R|+1,\\ldots,j_L+j_R.\\]\n\nThus a four-vector contains a rotational scalar, its time component, and a rotational three-vector, its spatial components. The label \\((\\tfrac12,\\tfrac12)\\) therefore does not mean that all four components describe physical spin-one polarizations. Equations of motion and constraints determine the actual propagating content of a field.\n\nAn antisymmetric rank-two tensor has six components and, after complexification, decomposes as\n\n\\[(1,0)\\oplus(0,1).\\]\n\nFor the electromagnetic field, these sectors can be expressed through complex combinations of the electric and magnetic fields. With consistent units, \\(\\mathbf E+ic\\mathbf B\\) and \\(\\mathbf E-ic\\mathbf B\\) belong to the two complementary sectors, with the assignment depending on conventions. A symmetric rank-two tensor decomposes into a traceless nine-dimensional representation \\((1,1)\\) and a scalar trace \\((0,0)\\).\n\nNontrivial finite-dimensional representations of the connected Lorentz group are not unitary with respect to a positive-definite inner product. The group’s noncompact simple structure prevents nontrivial continuous finite-dimensional unitary representations. This does not conflict with quantum mechanics: field components transform in finite-dimensional, generally nonunitary representations, while physical states transform in unitary representations on a typically infinite-dimensional Hilbert space. Lorentz-invariant spinor bilinears and tensor contractions use appropriate invariant structures, not necessarily positive-definite norms on component spaces."
    },
    {
      "heading": "The Poincaré group and the classification of particles",
      "text": "The spacetime symmetry relevant to isolated relativistic particles includes both Lorentz transformations and translations. It is the Poincaré group,\n\n\\[ISO(1,3)=\\mathbb R^{1,3}\\rtimes O(1,3),\\]\n\nor an appropriate connected component or covering group. The semidirect-product symbol records that Lorentz transformations act on translation vectors. For transformations written as pairs \\((a,\\Lambda)\\), composition is\n\n\\[(a,\\Lambda)(b,M)=(a+\\Lambda b,\\Lambda M).\\]\n\nTranslations therefore form a normal subgroup, but they do not commute with general Lorentz transformations.\n\nIn quantum theory, translations are generated by four-momentum \\(P^\\mu\\), whose components encode energy and momentum. Two important Casimir operators, which commute with the connected Poincaré action, are\n\n\\[P_\\mu P^\\mu,\\qquad W_\\mu W^\\mu,\\]\n\nwhere\n\n\\[W^\\mu=\\frac12\\epsilon^{\\mu\\nu\\rho\\sigma}P_\\nu M_{\\rho\\sigma}\\]\n\nis the Pauli–Lubanski vector. For a massive irreducible representation,\n\n\\[P_\\mu P^\\mu=m^2c^2,\\qquad W_\\mu W^\\mu=-m^2c^2\\hbar^2s(s+1).\\]\n\nThe invariant labels are therefore mass \\(m\\) and spin \\(s\\), together with a choice of energy sector.\n\nEugene Wigner’s classification proceeds by choosing a standard four-momentum and studying the subgroup leaving it fixed, called its little group. For a massive particle, the standard momentum can be chosen as \\((mc,0,0,0)\\). Its connected Lorentz little group is \\(SO(3)\\), whose double cover \\(SU(2)\\) supplies integer and half-integer spin. A massive spin-\\(s\\) particle has \\(2s+1\\) spin states at fixed momentum.\n\nFor a massless particle with nonzero momentum, the connected little group is isomorphic to the Euclidean motion group \\(ISO(2)\\). In the familiar finite-helicity representations, its translation-like part acts trivially, and rotations around the momentum direction characterize helicity. The photon has helicities \\(+1\\) and \\(-1\\). Other mathematical possibilities include continuous-spin representations, in which the translation-like part acts nontrivially; no elementary particle has been experimentally established in such a representation.\n\nThis classification distinguishes particle states from covariant field components. An electromagnetic four-potential has four Lorentz-vector components, but a free photon has only two physical polarization states. Gauge redundancy and field equations remove the surplus description. Likewise, a Dirac field contains components needed to describe both particle and antiparticle sectors rather than merely listing the spin states of one fixed-momentum particle."
    },
    {
      "heading": "Lorentz covariance, interactions, and discrete symmetries",
      "text": "An equation is Lorentz covariant when its transformed fields satisfy the same tensorial equation in transformed coordinates. This is more precise than requiring every component to remain numerically unchanged. In electromagnetism, the electric and magnetic fields are different components of an antisymmetric field-strength tensor \\(F^{\\mu\\nu}\\). A boost generally mixes them. The covariant form of Maxwell’s equations makes clear that the electric–magnetic split depends on the observer, while the electromagnetic field itself is a spacetime object.\n\nLorentz-invariant contractions provide systematic building blocks for physical equations. For a scalar field, a kinetic term is proportional to\n\n\\[\\partial_\\mu\\phi\\,\\partial^\\mu\\phi.\\]\n\nFor electromagnetism, it is proportional to\n\n\\[F_{\\mu\\nu}F^{\\mu\\nu}.\\]\n\nThe wave operator\n\n\\[\\Box=\\partial_\\mu\\partial^\\mu =\\frac{1}{c^2}\\frac{\\partial^2}{\\partial t^2}-\\nabla^2\\]\n\nis also Lorentz invariant. Such expressions help construct a Lagrangian whose action respects spacetime symmetry, although Lorentz symmetry alone does not determine the interactions or their numerical strengths.\n\nIn a flat-spacetime field theory, continuous Lorentz symmetry is related by Noether’s theorem to conserved angular-momentum currents, including boost charges. In a formulation using a symmetric conserved stress–energy tensor, the current takes the form\n\n\\[\\mathcal J^{\\lambda\\mu\\nu} =x^\\mu T^{\\lambda\\nu}-x^\\nu T^{\\lambda\\mu}.\\]\n\nWith a canonical stress–energy tensor, an explicit spin-current term may also appear. The conserved quantities combine orbital and intrinsic contributions. Appropriate boundary behavior is needed for integrated charges to exist.\n\nInvariance under the identity component does not require invariance under spatial parity or time reversal. The weak interaction violates parity, yet remains Lorentz covariant under proper orthochronous transformations. This is one reason that “Lorentz invariant” in physical discussions often refers specifically to the connected subgroup or its spin cover rather than to all four components of \\(O(1,3)\\).\n\nTime reversal also requires a distinction between spacetime geometry and quantum-state transformations. The coordinate operation \\(t\\mapsto-t\\) is represented by a real Lorentz matrix. A time-reversal symmetry on quantum states is instead implemented antiunitarily in the usual positive-energy setting. The CPT theorem concerns the combined operation of charge conjugation, parity, and time reversal in local relativistic quantum field theory under its standard assumptions. It is not a statement that charge conjugation belongs to the Lorentz group, nor does connected Lorentz symmetry by itself imply separate parity or time-reversal invariance."
    },
    {
      "heading": "Hyperbolic geometry and local Lorentz symmetry in gravitation",
      "text": "The Lorentz group has a geometric action beyond its role in changing inertial coordinates. Consider the future unit hyperboloid\n\n\\[\\mathcal H^3= \\{u\\in\\mathbb R^{1,3}:\\eta_{\\mu\\nu}u^\\mu u^\\nu=1,\\ u^0>0\\}.\\]\n\nThe connected Lorentz group acts transitively on this surface. The subgroup fixing the point \\((1,0,0,0)\\) is \\(SO(3)\\), so\n\n\\[\\mathcal H^3\\cong SO^+(1,3)/SO(3).\\]\n\nWith the positive metric induced by restricting \\(-\\eta\\) to tangent vectors, this is three-dimensional hyperbolic space of curvature \\(-1\\). It also represents the space of future-directed unit four-velocities.\n\nFor two such vectors \\(u\\) and \\(v\\), their relative rapidity satisfies\n\n\\[\\cosh\\chi=\\eta_{\\mu\\nu}u^\\mu v^\\nu.\\]\n\nRapidity is therefore a hyperbolic distance in velocity space. A massive observer’s normalized four-velocity lies on this hyperboloid, whereas null directions lie on its ideal boundary. This interpretation connects the non-Euclidean geometry of velocities with relativistic velocity composition and boost-induced rotations.\n\nThe action on future null directions gives another useful picture. After identifying proportional future null vectors, the light cone becomes a two-sphere, often called the celestial sphere. Through the covering group \\(SL(2,\\mathbb C)\\), connected Lorentz transformations act on a complex stereographic coordinate \\(z\\) by Möbius transformations,\n\n\\[z\\longmapsto\\frac{az+b}{cz+d}, \\qquad ad-bc=1.\\]\n\nThis connects Lorentz symmetry to conformal transformations of the sphere and provides a geometric description of how observers in relative motion assign different directions to incoming light.\n\nIn general relativity, a curved spacetime need not admit any nontrivial global Lorentz transformations as isometries. Nevertheless, each tangent space has a metric of Lorentzian signature. A local orthonormal frame, or tetrad, expresses the spacetime metric as\n\n\\[g_{\\mu\\nu}(x)=\\eta_{ab}e^a{}_\\mu(x)e^b{}_\\nu(x).\\]\n\nReplacing the tetrad by\n\n\\[e'^a{}_\\mu(x)=\\Lambda^a{}_b(x)e^b{}_\\mu(x)\\]\n\nleaves \\(g_{\\mu\\nu}\\) unchanged whenever \\(\\Lambda(x)\\) is Lorentz. Here the transformation may vary from point to point: it is a local freedom in choosing frames, not necessarily a physical symmetry mapping one spacetime point to another.\n\nA spin connection relates neighboring frames and makes covariant differentiation compatible with local Lorentz transformations. Spinor fields require the frame bundle to admit an appropriate spin lift, a global topological condition not supplied merely by writing the metric locally in orthonormal form. Local Lorentz freedom, coordinate invariance, and global spacetime isometries are thus related but distinct mathematical structures."
    },
    {
      "heading": "Historical development and mathematical consolidation",
      "text": "The Lorentz group emerged from nineteenth-century attempts to reconcile electromagnetism with the motion of bodies. Maxwell’s theory supplied a characteristic propagation speed that did not fit naturally into Galilean transformations. In 1887 Woldemar Voigt introduced a transformation related to the later Lorentz transformation in studying the wave equation, but his transformation included an overall scale factor and was not identical to the standard Lorentz transformation. Lorentz developed transformations in his electron theory during the 1890s and presented a further formulation in 1904.\n\nHenri Poincaré made the group structure explicit in 1905. He recognized that the relevant transformations form a group, used the name “Lorentz transformations,” and clarified the restrictions on an additional scale factor. His longer paper, published in 1906, developed the mathematical treatment further. These contributions distinguished the closure and composition properties of the transformations from their earlier use as auxiliary changes of variables in electromagnetic theory.\n\nIn 1905 Albert Einstein derived the transformations from the relativity principle and the constancy of light speed, without requiring a stationary luminiferous ether. This changed their physical interpretation: they expressed relations between equally valid inertial descriptions rather than mathematical corrections associated with motion through a preferred medium. Hermann Minkowski’s spacetime formulation, developed in 1907–1908 and presented prominently in his 1908 lecture “Space and Time,” made preservation of the four-dimensional interval the central geometric idea.\n\nQuantum theory then exposed the importance of covering groups and representations. Wolfgang Pauli’s two-component spin formalism in 1927 and Paul Dirac’s relativistic electron equation in 1928 made spinorial transformation laws central to physics. Hermann Weyl’s 1929 work clarified the role of two-component relativistic spinors. These developments showed why ordinary vector representations of the Lorentz group were insufficient for describing all matter fields.\n\nWigner’s 1939 paper on unitary representations of the inhomogeneous Lorentz group established the systematic relation between relativistic spacetime symmetry, mass, and spin. Work by mathematicians including Valentin Bargmann, Harish-Chandra, and Israel Gelfand and Mark Naimark developed the representation theory of noncompact groups, including Lorentz-related groups. This mathematical program distinguished finite-dimensional covariant representations from the infinite-dimensional unitary representations relevant to quantum states.\n\nModern usage retains several related meanings. In a geometric discussion, “Lorentz group” usually denotes \\(O(1,3)\\); in much particle physics it may mean \\(SO^+(1,3)\\); and in treatments of spinors the operative group is often \\(SL(2,\\mathbb C)\\). Specifying the component, covering group, and type of representation resolves differences that are invisible at the level of infinitesimal coordinate transformations but essential for reflections, spin, and global topology."
    }
  ],
  "links": [
    "special-relativity",
    "spacetime",
    "speed-of-light",
    "matrix",
    "lorentz-transformation",
    "inertial-reference-frame",
    "lie-group",
    "lie-algebra",
    "angular-momentum",
    "spinor",
    "representation-theory",
    "quantum-mechanics",
    "energy",
    "momentum",
    "photon",
    "electromagnetism",
    "maxwells-equations",
    "lagrangian",
    "noethers-theorem",
    "weak-interaction",
    "parity-physics",
    "quantum-field-theory",
    "hyperbolic-geometry",
    "general-relativity",
    "albert-einstein"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}