Discrete Groups, Symmetric Spaces, and Global Holonomy Author(S): Joseph A

Total Page:16

File Type:pdf, Size:1020Kb

Discrete Groups, Symmetric Spaces, and Global Holonomy Author(S): Joseph A Discrete Groups, Symmetric Spaces, and Global Holonomy Author(s): Joseph A. Wolf Reviewed work(s): Source: American Journal of Mathematics, Vol. 84, No. 4 (Oct., 1962), pp. 527-542 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/2372860 . Accessed: 03/05/2012 14:57 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Mathematics. http://www.jstor.org DISCRETE GROUPS, SYMMETRIC SPACES, AND Gl,OBAL HOLONOMY.* By JOSEPH A. WOLF.' 1. Introduction. Let M be a connectedsimply connected Riemannian manifoldand let r be a properlydiscontinuous group of isometriessuch that M/r is compact. If everysectional curvature of M is negative,in particularif M is a noncompactirreducible symmetric space of rank 1, thena methodof iR. Cartanshows that every abelian subgroupof r is eitherfinite or the productof a finitegroup with an infinitecyclic group. If M is the Euclidean space Rn, then a calculationshows that everyabelian subgroupof r is the productof a finitegroup with a freeabelian groupon ? n generators. These phenomena are unifiedby one of the conclusionsof our Theorem6. 2: If M is Riemannian symmetricand v is the maximumof the dimensionsof thosetotally geodesic submanifoldsof M whichare isometricto Euclidean spaces,then every abelian subgroupof r is the productof a finitegroup with a free abelian group on < v generators,and r has a subgroupwhich is free abelian on v generators. We also provethat an abelian subgroupof r must preservea flat connected totallygeodesic submanifold of M; if M/r is a manifold,it followsthat M/r containsa maximalconnected flat totally geodesic submanifold which is closed, and everyabelian subgroupof -r,(Mr/) can be representedby closed geodesic arcs lying in a connectedflat totallygeodesic submanifold(Corollary 6. 6). In addition,we analyzethe groupof componentsof the homogeneousholonomy group of a locallysymmetric Riemannian manifold N (Theorem 7. 1), prove thatN has compacthomogeneous holonomy group if N is compact2 (Corollary 7.2), and give conditionsfor everymanifold locally isometricto N to have compacthomogeneous holonomy group (Corollary7. 3). Our bounds are obtainedby estimatingthe "size" of abelian subgroups of discreteuniform subgroups of Lie groupsL =E X G whereE is a semi- directproduct of a compactgroup and a vectorgroup, such as the Euclidean group,and G is a reductiveLie group with only finitelymany components. * Received October 26, 1961; revised August 4, 1962. 1 The author thanks the National Science Foundation for fellowship support during the preparation of this paper. 2 If N is flat, this is just the classical Bieberbach Theorem [3]. This does not give a new proof of the Bieberbach Theorem because that result is used in our arguments. 527 528 JOSEPH A. WOLF. The estimatesare firstmade for reductivegroups (Theorem 4. 2), and then extendedby a generalizationof Bieberbach'sTheorem (Theorem 5. 1). Let r be a discreteuniform subgroup of a reductiveLie group G, where G has onlyfinitely many components, and let A be an abelian subgroupof r. Our main idea is that the size of A can be estimatedby findinga Cartan subgroupH of G which is normalizedby A, and observingthat A n H has finiteindex in A. In orderto findH, we firstprove that everyelement of r is a semisimpleelement of G (Theorem 3. 2), and then apply a result of A. Borel and G. D. Mostow (Corollary3. 7). The main tool in our proof of Theorem3. 2 is a geometriccharacterization of the semisimpleelements of G (Lemma 3. 6). 2. Preliminaries. 2.1 Lie groups. Given a Lie group G, Go will denote the identity component,(M will denotethe Lie algebra, and exp: S -> G will denotethe exponentialmap. G and (Mare called reductiveif the adjoint representation of (M (or, equivalently,of Go) is fullyreducible, i. e., if (Mis the directsum of an abelian ideal SCand a semisimpleideal WY;then SCis the centerof (M, A = exp(C) is the identitycomponent of the centerof Go and is called the connectedcenter of G,,,(' is the derivedalgebra of S and is called the semi- simplepart of (M,G' exp(E') is called the semisimplepart of G,, and there is a natural homomorphism(a, g) -> ag of A X G' onto Go. If S is reductive,then the Cartan subalgebrasof (M are the subalgebras of the formSC 0 &', where0 denotesdirect sum of ideals and &' is a Cartan subalgebraof WY;thus the Cartan subalgebrasof (Mare abelian. By Cartan subgroupof a Lie group G, we mean a (necessarilyconnected) group of the formexp (.e) where& is a Cartan subalgebraof (S. Under the adjoint representationof a Lie group G, an elementg C G inducesan automorphismad(g) of (M; we will call g semisimpleif ad(g) is a fully reduciblelinear transformationof (M. If g is a semisimpleelement of a reductiveLie group G, and if Z is the centralizerof g in G, then not onlyis Z reductivebut the adjoint representationof G inducesa fullyreducible representationof 8 on (S. If S and T are subsets of a group G, then the commutator[s, T] denotesthe set of all elements[s, t] sts-'t-1where s C S, t C T. If S and T are groupsand q is a homomorphismof S into the groupof automorphismsof T, then the semidirectproduct S pT is the set S X T with the group structure (s1,t,) (s2,t2) = (s1s2, (cp(s2-) (t,))t2). If S is DISCRETE GROUPS. 529 given as a group of automorphismsof T, then the semidirectproduct is denotedS. T. 2. 2. Discretegroups. A subgroupr of a topologicalgroup G is discrete if G has an open set U such that r n u is just the identityelement 1 E G. A subgroupH of G is uniformif the coset space G/I (I7 is the closureof H in G) is compact. Let r be a discreteuniform subgroup of G. C. L. Siegel [9] has shown that G is locally compact,and, if everycovering of G by open sets has a countablerefinement,3 then G has a compact subsetF such that r- F= G, every g E G has a neighborhoodcontained in a finite union of the yF, rF = {y C r: yF meets F} is finite, and rF generates r if G is connected. It followsthat P finitelygenerated if G/Gois finitelygenerated, but this is betterseen directly[7]. F will be called a fundamentaldomain for the action of r on G by righttranslations. 2. 3. Symmetricspaces. It is well known that a connectedsimply connected Riemannian symmetricspace M is isometric to a product MOXX lXX . X Mt whereMO is a Euclidean space and each M, (i > 0) is an irreducibleRiemannian (non-Euclideanand not isometricto a product of lower dimensionalRiemannian manifolds) symmetricspace; Mo is the Euclidean part of M, ' =-1 X . X Mt is the non-Euclideanpart of M3, and the MT (i > 0) are the irreduciblefactors of 3. We will say that M is strictlynon-Euclidean if M M=M', i. e., if dim.MO = 0, and will say that M is strictlynoncompact if everyirreducible factor of M is noncompact. If M is strictlynoncompact, then everysectional curvature on M is ? 0. Full groups of isometriesare related by I(M) =I((MO) X I(M), and I (M') is generatedby I (M1) X . X I (Mt) togetherwith all permutations on mutuallyisometric sets of Mi. Connectedgroups of isometriesare related by Io (M) = Io (Mo) X Io (Ml) X . X Io (Mt), and I (M)/Io(M) is finite. I (Mo3)is the Euclideangroup E (dim.Mo ), I (MI) is a compactsemisimple Lie groupif M,,is compact,and Io (Mi) is a noncompactcenterless real or complex simpleLie groupif M, is noncompact(i > 0). Let S be a maximalconnected flat (all sectionalcurvatures zero) totally geodesic submanifoldof M. I(M) acts transitivelyon the set of all such submanifolds,and the rank of M (denoted rank.M) is definedto be their commondimension. S is isometricto a productMO X S, X . *X St where Si is a maximalconnected flat totally geodesic submanifold of Mi, whenceit 3 Siegel requires that G have a countable basis for open sets, but uses only this weaker property. 530 JOSEPH A. WOLF. is easily seen that S is a closed submanifoldof M. The symmetryto MI at a point of Si induces a symmetryof Si; it followsthat each Si (and thus S) is a connectedRiemannian symmetric manifold of constantcurvature zero. Thus Si is a flat torus if Mi is compactand Si is a Euclidean space if Mi (i> 0) is noncompact;to see this, we use [10, Theoreme4] or [11, ? 14] togetherwith the factthat Si is the orbitof some Cartansubgroup of maximal vectorrank in I (Mj). If MC is the productof the compactirreducible factors of M and MN is the productof MO withthe noncompactirreducible factors of M, it followsthat S is isometricto the productof a flattorus of dimension rank.Mc with a Euclidean space of dimensionrank.MN. For this reason, we definethe vectorrank of M (denotedv-rank. M) to be rank.MN. Observe that v-rank.M - dim.MO + v-rank.M'. Let r be a subgroupof I (M). The action of r on M is properlydis- continuousif everyelement of M has a neighborhoodwhich meets its trans- formsby only a finitenumber of elementsof r; this is equivalentto r being a discretesubgroup of I (M). The action of r on M is freeif 1 #&y C r and x C M impliesy(x) #x.
Recommended publications
  • Matrix Lie Groups
    Maths Seminar 2007 MATRIX LIE GROUPS Claudiu C Remsing Dept of Mathematics (Pure and Applied) Rhodes University Grahamstown 6140 26 September 2007 RhodesUniv CCR 0 Maths Seminar 2007 TALK OUTLINE 1. What is a matrix Lie group ? 2. Matrices revisited. 3. Examples of matrix Lie groups. 4. Matrix Lie algebras. 5. A glimpse at elementary Lie theory. 6. Life beyond elementary Lie theory. RhodesUniv CCR 1 Maths Seminar 2007 1. What is a matrix Lie group ? Matrix Lie groups are groups of invertible • matrices that have desirable geometric features. So matrix Lie groups are simultaneously algebraic and geometric objects. Matrix Lie groups naturally arise in • – geometry (classical, algebraic, differential) – complex analyis – differential equations – Fourier analysis – algebra (group theory, ring theory) – number theory – combinatorics. RhodesUniv CCR 2 Maths Seminar 2007 Matrix Lie groups are encountered in many • applications in – physics (geometric mechanics, quantum con- trol) – engineering (motion control, robotics) – computational chemistry (molecular mo- tion) – computer science (computer animation, computer vision, quantum computation). “It turns out that matrix [Lie] groups • pop up in virtually any investigation of objects with symmetries, such as molecules in chemistry, particles in physics, and projective spaces in geometry”. (K. Tapp, 2005) RhodesUniv CCR 3 Maths Seminar 2007 EXAMPLE 1 : The Euclidean group E (2). • E (2) = F : R2 R2 F is an isometry . → | n o The vector space R2 is equipped with the standard Euclidean structure (the “dot product”) x y = x y + x y (x, y R2), • 1 1 2 2 ∈ hence with the Euclidean distance d (x, y) = (y x) (y x) (x, y R2).
    [Show full text]
  • The Pure State Space of Quantum Mechanics As Hermitian Symmetric
    The Pure State Space of Quantum Mechanics as Hermitian Symmetric Space. R. Cirelli, M. Gatti, A. Mani`a Dipartimento di Fisica, Universit`adegli Studi di Milano, Italy Istituto Nazionale di Fisica Nucleare, Sezione di Milano, Italy Abstract The pure state space of Quantum Mechanics is investigated as Hermitian Symmetric K¨ahler manifold. The classical principles of Quantum Mechan- ics (Quantum Superposition Principle, Heisenberg Uncertainty Principle, Quantum Probability Principle) and Spectral Theory of observables are dis- cussed in this non linear geometrical context. Subj. Class.: Quantum mechanics 1991 MSC : 58BXX; 58B20; 58FXX; 58HXX; 53C22 Keywords: Superposition principle, quantum states; Riemannian mani- folds; Poisson algebras; Geodesics. 1. Introduction Several models of delinearization of Quantum Mechanics have been pro- arXiv:quant-ph/0202076v1 14 Feb 2002 posed (see f.i. [20] [26], [11], [2] and [5] for a complete list of references). Frequentely these proposals are supported by different motivations, but it appears that a common feature is that, more or less, the delinearization must be paid essentially by the superposition principle. This attitude can be understood if the delinearization program is worked out in the setting of a Hilbert space as a ground mathematical structure. However, as is well known, the groundH mathematical structure of QM is the manifold of (pure) states P( ), the projective space of the Hilbert space . Since, obviously, P( ) is notH a linear object, the popular way of thinking thatH the superposition principleH compels the linearity of the space of states is untenable. The delinearization program, by itself, is not related in our opinion to attempts to construct a non linear extension of QM with operators which 1 2 act non linearly on the Hilbert space .
    [Show full text]
  • Symmetric Spaces with Conformal Symmetry
    Symmetric Spaces with Conformal Symmetry A. Wehner∗ Department of Physics, Utah State University, Logan, Utah 84322 October 30, 2018 Abstract We consider an involutive automorphism of the conformal algebra and the resulting symmetric space. We display a new action of the conformal group which gives rise to this space. The space has an in- trinsic symplectic structure, a group-invariant metric and connection, and serves as the model space for a new conformal gauge theory. ∗Electronic-mail: [email protected] arXiv:hep-th/0107008v1 1 Jul 2001 1 1 Introduction Symmetric spaces are the most widely studied class of homogeneous spaces. They form a subclass of the reductive homogeneous spaces, which can es- sentially be characterized by the fact that they admit a unique torsion-free group-invariant connection. For symmetric spaces the curvature is covari- antly constant with respect to this connection. Many essential results and extensive bibliographies on symmetric spaces can be found, for example, in the standard texts by Kobayashi and Nomizu [1] and Helgason [2]. The physicists’ definition of a (maximally) symmetric space as a metric space which admits a maximum number of Killing vectors is considerably more restrictive. A thorough treatment of symmetric spaces in general rel- ativity, where a pseudo-Riemannian metric and the metric-compatible con- nection are assumed, can be found in Weinberg [3]. The most important symmetric space in gravitational theories is Minkowski space, which is based on the symmetric pair (Poincar´egroup, Lorentz group). Since the Poincar´e group is not semi-simple, it does not admit an intrinsic group-invariant met- ric, which would project in the canonical fashion to the Minkowski metric ηab = diag(−1, 1 ..
    [Show full text]
  • On the Third-Degree Continuous Cohomology of Simple Lie Groups 2
    ON THE THIRD-DEGREE CONTINUOUS COHOMOLOGY OF SIMPLE LIE GROUPS CARLOS DE LA CRUZ MENGUAL ABSTRACT. We show that the class of connected, simple Lie groups that have non-vanishing third-degree continuous cohomology with trivial ℝ-coefficients consists precisely of all simple §ℝ complex Lie groups and of SL2( ). 1. INTRODUCTION ∙ Continuous cohomology Hc is an invariant of topological groups, defined in an analogous fashion to the ordinary group cohomology, but with the additional assumption that cochains— with values in a topological group-module as coefficient space—are continuous. A basic refer- ence in the subject is the book by Borel–Wallach [1], while a concise survey by Stasheff [12] provides an overview of the theory, its relations to other cohomology theories for topological groups, and some interpretations of algebraic nature for low-degree cohomology classes. In the case of connected, (semi)simple Lie groups and trivial real coefficients, continuous cohomology is also known to successfully detect geometric information. For example, the situ- ation for degree two is well understood: Let G be a connected, non-compact, simple Lie group 2 ℝ with finite center. Then Hc(G; ) ≠ 0 if and only G is of Hermitian type, i.e. if its associated symmetric space of non-compact type admits a G-invariant complex structure. If that is the case, 2 ℝ Hc(G; ) is one-dimensional, and explicit continuous 2-cocycles were produced by Guichardet– Wigner in [6] as an obstruction to extending to G a homomorphism K → S1, where K<G is a maximal compact subgroup. The goal of this note is to clarify a similar geometric interpretation of the third-degree con- tinuous cohomology of simple Lie groups.
    [Show full text]
  • Automorphism Groups of Locally Compact Reductive Groups
    PROCEEDINGS OF THE AMERICAN MATHEMATICALSOCIETY Volume 106, Number 2, June 1989 AUTOMORPHISM GROUPS OF LOCALLY COMPACT REDUCTIVE GROUPS T. S. WU (Communicated by David G Ebin) Dedicated to Mr. Chu. Ming-Lun on his seventieth birthday Abstract. A topological group G is reductive if every continuous finite di- mensional G-module is semi-simple. We study the structure of those locally compact reductive groups which are the extension of their identity components by compact groups. We then study the automorphism groups of such groups in connection with the groups of inner automorphisms. Proposition. Let G be a locally compact reductive group such that G/Gq is compact. Then /(Go) is dense in Aq(G) . Let G be a locally compact topological group. Let A(G) be the group of all bi-continuous automorphisms of G. Then A(G) has the natural topology (the so-called Birkhoff topology or ^-topology [1, 3, 4]), so that it becomes a topo- logical group. We shall always adopt such topology in the following discussion. When G is compact, it is well known that the identity component A0(G) of A(G) is the group of all inner automorphisms induced by elements from the identity component G0 of G, i.e., A0(G) = I(G0). This fact is very useful in the study of the structure of locally compact groups. On the other hand, it is also well known that when G is a semi-simple Lie group with finitely many connected components A0(G) - I(G0). The latter fact had been generalized to more general groups ([3]).
    [Show full text]
  • Oldandnewonthe Exceptionalgroupg2 Ilka Agricola
    OldandNewonthe ExceptionalGroupG2 Ilka Agricola n a talk delivered in Leipzig (Germany) on product, the Lie bracket [ , ]; as a purely algebraic June 11, 1900, Friedrich Engel gave the object it is more accessible than the original Lie first public account of his newly discovered group G. If G happens to be a group of matrices, its description of the smallest exceptional Lie Lie algebra g is easily realized by matrices too, and group G2, and he wrote in the corresponding the Lie bracket coincides with the usual commuta- Inote to the Royal Saxonian Academy of Sciences: tor of matrices. In Killing’s and Lie’s time, no clear Moreover, we hereby obtain a direct defi- distinction was made between the Lie group and nition of our 14-dimensional simple group its Lie algebra. For his classification, Killing chose [G2] which is as elegant as one can wish for. a maximal set h of linearly independent, pairwise 1 [En00, p. 73] commuting elements of g and constructed base Indeed, Engel’s definition of G2 as the isotropy vectors Xα of g (indexed over a finite subset R of group of a generic 3-form in 7 dimensions is at elements α ∈ h∗, the roots) on which all elements the basis of a rich geometry that exists only on of h act diagonally through [ , ]: 7-dimensional manifolds, whose full beauty has been unveiled in the last thirty years. [H,Xα] = α(H)Xα for all H ∈ h. This article is devoted to a detailed historical In order to avoid problems when doing so he chose and mathematical account of G ’s first years, in 2 the complex numbers C as the ground field.
    [Show full text]
  • What Is the Group of Conjugate Symplectic Matrices? Howard E
    What is the group of conjugate symplectic matrices? Howard E. Haber Santa Cruz Institute for Particle Physics University of California, Santa Cruz, CA 95064, USA June 26, 2017 Abstract The group of complex symplectic matrices, Sp(n,C) is defined as the set of 2n 2n T 0 In × complex matrices, M GL(2n, C) M JM = J , where J I , and In is the { ∈ | } ≡ − n 0 n n identity matrix. If the transpose is replaced by hermitian conjugate, then one obtains × the group of conjugate symplectic matrices, M GL(2n, C) M †JM = J . In these { ∈ | } notes, we demonstrate that the group of conjugate symplectic matrices is isomorphic to † In 0 U(n,n)= M GL(2n, C) M In,nM = In,n , where In,n I . { ∈ | } ≡ 0 − n The group of complex symplectic matrices, Sp(n,C) is defined as the set of 2n 2n complex matrices,1 × Sp(n, C) M GL(2n, C) M TJM = J , (1) ≡{ ∈ | } where the matrix J is defined in block form by, 0 In J , (2) ≡ In 0 − with 0 denoting the n n zero matrix and In the n n identity matrix. Sp(n,C) is a complex × × 2 simple Lie group. Moreover, the determinant of any complex symplectic matrix is 1. If the transpose is replaced by hermitian conjugate in eq. (1), then one obtains the group of conjugate symplectic matrices, G = M GL(2n, C) M †JM = J , (3) { ∈ | } following the terminology or Refs. [1, 2]. Indeed, the group of conjugate symplectic matrices is a Lie group that is neither semi-simple nor complex.
    [Show full text]
  • Special Unitary Group - Wikipedia
    Special unitary group - Wikipedia https://en.wikipedia.org/wiki/Special_unitary_group Special unitary group In mathematics, the special unitary group of degree n, denoted SU( n), is the Lie group of n×n unitary matrices with determinant 1. (More general unitary matrices may have complex determinants with absolute value 1, rather than real 1 in the special case.) The group operation is matrix multiplication. The special unitary group is a subgroup of the unitary group U( n), consisting of all n×n unitary matrices. As a compact classical group, U( n) is the group that preserves the standard inner product on Cn.[nb 1] It is itself a subgroup of the general linear group, SU( n) ⊂ U( n) ⊂ GL( n, C). The SU( n) groups find wide application in the Standard Model of particle physics, especially SU(2) in the electroweak interaction and SU(3) in quantum chromodynamics.[1] The simplest case, SU(1) , is the trivial group, having only a single element. The group SU(2) is isomorphic to the group of quaternions of norm 1, and is thus diffeomorphic to the 3-sphere. Since unit quaternions can be used to represent rotations in 3-dimensional space (up to sign), there is a surjective homomorphism from SU(2) to the rotation group SO(3) whose kernel is {+ I, − I}. [nb 2] SU(2) is also identical to one of the symmetry groups of spinors, Spin(3), that enables a spinor presentation of rotations. Contents Properties Lie algebra Fundamental representation Adjoint representation The group SU(2) Diffeomorphism with S 3 Isomorphism with unit quaternions Lie Algebra The group SU(3) Topology Representation theory Lie algebra Lie algebra structure Generalized special unitary group Example Important subgroups See also 1 of 10 2/22/2018, 8:54 PM Special unitary group - Wikipedia https://en.wikipedia.org/wiki/Special_unitary_group Remarks Notes References Properties The special unitary group SU( n) is a real Lie group (though not a complex Lie group).
    [Show full text]
  • Holonomy and the Lie Algebra of Infinitesimal Motions of a Riemannian Manifold
    HOLONOMY AND THE LIE ALGEBRA OF INFINITESIMAL MOTIONS OF A RIEMANNIAN MANIFOLD BY BERTRAM KOSTANT Introduction. .1. Let M be a differentiable manifold of class Cx. All tensor fields discussed below are assumed to be of class C°°. Let X be a vector field on M. If X vanishes at a point oCM then X induces, in a natural way, an endomorphism ax of the tangent space F0 at o. In fact if y£ V„ and F is any vector field whose value at o is y, then define axy = [X, Y]„. It is not hard to see that [X, Y]0 does not depend on F so long as the value of F at o is y. Now assume M has an affine connection or as we shall do in this paper, assume M is Riemannian and that it possesses the corresponding affine connection. One may now associate to X an endomorphism ax of F0 for any point oCM which agrees with the above definition and which heuristically indicates how X "winds around o" by defining for vC V0, axv = — VtX, where V. is the symbol of covariant differentiation with respect to v. X is called an infinitesmal motion if the Lie derivative of the metric tensor with respect to X is equal to zero. This is equivalent to the statement that ax is skew-symmetric (as an endomorphism of F0 where the latter is provided with the inner product generated by the metric tensor) for all oCM. §1 contains definitions and formulae which will be used in the remaining sections.
    [Show full text]
  • Symmetric Spaces of Hermitian Type
    Differential Geometry and its Applications 1 (1991) 195-233 195 North-Holland Symmetric spaces of Hermitian type G. ‘Olafsson Mathematisches Znstitut der Uniuersitit Giittingen, Bunsenstraj?e 3-5, D-3400 Giittingen Received 7 February 1991 ‘Olafsson, G., Symmetric spaces of Hermitian type, Diff. Geom. Appl. 1 (1991) 195-233. Abstract: Let M = G/H be a semisimple symmetric space, r the corresponding involution and D = G/K the Riemannian symmetric space. Then we show that the following are equivalent: M is of Hermitian type; r induces a conjugation on D; there exists an open regular H-invariant cone R in q = h’ such that k n 0 # 0. We relate the spaces of Hermitian type to the regular and parahermitian symmetric spaces, analyze the fine structure of D under r and construct an equivariant Cayley transform. We collect also some results on the classification of invariant cones in q. Finally we point out some applications in representations theory. Keywords: Symmetric spaces, semisimple Lie groups, invariant convex cones, causal orientation, ordering, convexity theorem. MS classification: 53635, 57S25, 22E15, 06AlO. Introduction Bounded symmetric domains and their unbounded counterparts, the Siegel domains, have long been an important part of different fields of mathematics, e.g., number theory, algebraic geometry, harmonic analysis and representations theory. So the holomorphic discrete series and other interesting representations of a group live in spaces of holo- morphic functions on such domains. In the last years some interplays with harmonic analysis on affine symmetric spaces have also become apparent, e.g., a construction of non-zero harmonic forms related to the discrete series of such spaces (see [44] and the literature there).
    [Show full text]
  • Normalizers and Centralizers of Reductive Subgroups of Almost Connected Lie Groups
    Journal of Lie Theory Volume 8 (1998) 211{217 C 1998 Heldermann Verlag Normalizers and centralizers of reductive subgroups of almost connected Lie groups Detlev Poguntke Communicated by K. H. Hofmann Abstract. If M is an almost connected Lie subgroup of an almost connected Lie group G such that the adjoint group M is reductive then the product of M and its centralizer is of finite index in the normalizer. The method of proof also gives a different approach to the Weyl groups in the sense of [6]. An almost connected Lie group is a Lie group with finitely many connected compo- nents. For a subgroup U of any group H denote by Z(U; H) and by N(U; H) the centralizer and the normalizer of U in H , respectively. The product UZ(U; H) is always contained in N(U; H). By Theorem A of [3], if H is an almost connected Lie group and U is compact, the difference is not too "big," i.e., UZ(U; H) is of finite index in N(U; H). While from the point of view of applications, for instance to probability theory of Lie groups, the hypothesis of the compactness of U may be the most interesting one, this short note presents a theorem, which shows that, from the point of view of the proof, it is not the topological property of compact- ness which is crucial, but an algebraic consequence of it, namely, the reductivity of the adjoint group. Moreover, we provide additional information on the almost connectedness of certain normalizers.
    [Show full text]
  • Symmetric Spaces
    CHAPTER 2 Symmetric Spaces Introduction We remark that a prerequisite for this course is a basic understanding of Rie- mannian geometry, see for instance [Boo86]or[Hel79], which at the same time is a standard reference for the theory of symmetric spaces. See also [Bor98]fora much more condensed version of the latter. In 1926, É. Cartan began to study those Riemannian manifolds,forwhich central symmetries are distance preserving. The term central symmetry is going to be made precise later but you can think of it as a reflection about a point, as in Rn for instance. These spaces are now known as symmetric spaces;henoticedthatthe classification of these is essentially equivalent to the classification of real semisimple Lie algebras. The local definition of symmetric spaces is a generalisation of constant sectional curvature. Just as a differentiable function from R to R is constant if its derivative vanishes identically, one may ask for the Riemanncurvaturetensorto have covariant derivative zero. As incomprehensible as thisstatementmaybeatthe moment, it yields an important equivalent characterizationofsymmetricspaces. Example. The following list of examples of symmetric spaces contains the constant curvature cases. n n (i) Euclidean n-space E =(R ,geucl) has constant sectional curvature zero and isometry group Iso(En) = O(n) !Rn. n n ∼ (ii) The n-sphere S =(S ,geucl) has constant sectional curvature one. Its Riemannian metric arises through restriction of the Riemannian metric of the ambient space Rn+1 Sn to the tangent bundle of Sn.Theisometry group of Sn is O(n). ⊃ (iii) Hyperbolic n-space Hn has constant sectional curvature minus one.
    [Show full text]