Abstract
This article is part of an ongoing project aiming at the connections between causal structures on homogeneous spaces, Algebraic Quantum Field Theory, modular theory of operator algebras and unitary representations of Lie groups. In this article we concentrate on non-compactly causal symmetric spaces G/H. This class contains de Sitter space but also other spaces with invariant partial ordering. The central ingredient is an Euler element h in the Lie algebra of \({{\mathfrak {g}}}\). We define three different kinds of wedge domains depending on h and the causal structure on G/H. Our main result is that the connected component containing the base point eH of these seemingly different domains all agree. Furthermore we discuss the connectedness of those wedge domains. We show that each of these spaces has a natural extension to a non-compactly causal symmetric space of the form \(G_{{\mathbb {C}}}/G^c\) where \(G^c\) is a certain real form of the complexification \(G_{{\mathbb {C}}}\) of G. As \(G_{{\mathbb {C}}}/G^c\) is non-compactly causal, it also contains three types of wedge domains. Our results says that the intersection of these domains with G/H agree with the wedge domains in G/H.
Similar content being viewed by others
Data availibility
Data sharing is not applicable to this article as no dataset were used or generated in the article.
Notes
We write \(C^\circ \) for the relative interior of the cone C in its span.
As in the proof of Lemma 4.8 below, this follows from the \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}({{\mathfrak {h}}}_1)\)-invariance of the characteristic function of \(C_1\) by taking \(D = \varphi ^{-1}([1,\infty ))\).
A subspace is called hyperbolic if it consists of hyperbolic elements.
We want to keep some flexibility in choosing \(K_{{\mathbb {C}}}\) and hence the complexification \(M^r_{{\mathbb {C}}}\) because the crown domain, which is simply connected, can be realized in many “complexifications”.
References
Akhiezer, D.N., Gindikin, S.G.: On Stein extensions of real symmetric spaces. Math. Ann. 286, 1–12 (1990)
Araki, H., Woods, E.J.: Representations of the canonical commutation relations describing a nonrelativistic infinite free Bose gas. J. Math. Phys. 4, 637–662 (1963)
Baumgärtel, H., Jurke, M., Lledo, F.: Twisted duality of the CAR-algebra. J. Math. Phys. 43(8), 4158–4179 (2002)
Borchers, H.-J.: On the net of von Neumann algebras assiciated with a wedge and wedge-causal manifold (2009). Preprint, available at http://www.theorie.physik.uni-goettingen.de/forschung/qft/publications/2009
Borchers, H.-J., Buchholz, D.: Global properties of vacuum states in de Sitter space. Annales Poincare Phys. Theor. A 70, 23–40 (1999)
Bratteli, O., Robinson, D.W.: Operator Algebras and Quantum Statistical Mechanics I, 2nd edn., Texts and Monographs in Physics. Springer, Berlin (1987)
Bros, J., Moschella, U.: Two-point functions and quantum fields in de Sitter universe. Rev. Math. Phys. 8, 327–391 (1996)
Buchholz, D., Mund, J., Summers, S.J.: Transplantation of local nets and geometric modular action on Robertson-Walker Space-Times. Fields Inst. Commun. 30, 65–81 (2001)
Buchholz, D., Summers, S.J.: Stable quantum systems in anti-de Sitter space: causality, independence and spectral properties. J. Math. Phys. 45, 4810–4831 (2004)
Ciolli, F., Longo, R., Ranallo, A., Ruzzi, G.: Relative entropy and curved spacetimes. J. Geom. Phys. 172, Paper No. 104416, 16 pp (2022)
Dappiaggi, C., Lechner, G., Morfa-Morales, E.: Deformations of quantum field theories on spacetimes with Killing vector fields. Commun. Math. Phys. 305(1), 99–130 (2011)
Dörr, N., Neeb, K.-H.: On wedges in Lie triple systems and ordered symmetric spaces. Geometriae Dedicata 46, 1–34 (1993)
Faraut, J., Koranyi, A.: Analysis on Symmetric Cones. Oxford Math Monographs, Oxford University Press, Oxford (1994)
Gindikin, S., Krötz, B.: Complex crowns of Riemannian symmetric spaces and non-compactly causal symmetric spaces. Trans. Am. Math. Soc. 354(8), 3299–3327 (2002)
Haag, R.: Local Quantum Physics. Fields, Particles, Algebras, 2nd edn, Texts and Monographs in Physics. Springer, Berlin (1996)
Hilgert, J., Neeb, K.-H.: Lie Semigroups and Their Applications, Lecture Notes in Math, vol. 1552. Springer, Berlin (1993)
Hilgert, J., Neeb, K.-H.: Structure and Geometry of Lie Groups. Springer, Berlin (2012)
Hilgert, J., Ólafsson, G.: Causal Symmetric Spaces, Geometry and Harmonic Analysis, Perspectives in Mathematics, vol. 18. Academic Press, New York (1997)
Krötz, B.: Corner view on the crown domain. Jpn. J. Math. 2, 303–311 (2007)
Krötz, B., Neeb, K.-H.: On hyperbolic cones and mixed symmetric spaces. J. Lie Theory 6(1), 69–146 (1996)
Krötz, B., Stanton, R.J.: Holomorphic extensions of representations. II. Geometry and harmonic analysis. Geom. Funct. Anal. 15(1), 190–245 (2005)
Lauridsen-Ribeiro, P.: Structural and Dynamical Aspects of the AdS/CFT Correspondence: a Rigorous Approach. PhD thesis, Sao Paulo (2007) ar**v:0712.0401v3 [math-ph] (2008)
Lechner, G.: Algebraic Constructive Quantum Field Theory: Integrable Models and Deformation Techniques. In: Brunetti, R., et al. (eds.) Advances in Algebraic Quantum Field Theory, Math. Phys. Stud, pp. 397–449. Springer, Berlin (2015). ar**v:1503.03822
Loos, O.: Symmetric Spaces I: General Theory. W. A. Benjamin Inc, New York (1969)
Morinelli, V., Neeb, K.-H.: Covariant homogeneous nets of standard subspaces. Commun. Math. Phys. 386, 305–358 (2021). ar**v:2010.07128
Morinelli, V., Neeb, K.-H., Ólafsson, G.: From Euler elements and \(3\)-gradings to non-compactly causal symmetric spaces. J. Lie Theory 33(1) (2023) (to appear). ar**v:2207.14034
Morinelli, V., Neeb, K.-H., Ólafsson, G.: Modular geodesics and wedge domains in general non-compactly causal symmetric spaces (in preparation)
Neeb, K.-H.: On the complex geometry of invariant domains in complexified symmetric spaces. Annales de l’Inst. Fourier 49(1), 177–225 (1999)
Neeb, K.-H.: Holomorphy and Convexity in Lie Theory, Expositions in Mathematics, vol. 28. de Gruyter Verlag, Berlin (2000)
Neeb, K.-H.: On differentiable vectors for representations of infinite dimensional Lie groups. J. Funct. Anal. 259, 2814–2855 (2010)
Neeb, K.-H., Ólafsson, G.: Antiunitary representations and modular theory. In: Grabowska, K., et al. (eds.) 50th Sophus Lie Seminar, Banach Center Publications, vol. 113, pp. 291–362. ar**v:1704.01336
Neeb, K.-H., Ólafsson, G.: Wedge domains in compactly causal symmetric spaces. Int. Math. Res. Notices (to appear). ar**v:2107.13288
Neeb, K.-H., Ólafsson, G., Ørsted, B.: Standard subspaces of Hilbert spaces of holomorphic functions on tube domains. Commun. Math. Phys. 386, 1437–1487 (2021). ar**v:2007.14797
Neumann, A., Ólafsson, G.: Minimal and maximal semigroups related to causal symmetric spaces. Semigroup Forum 61, 57–85 (2000)
Oeh, D.: Classification of 3-graded causal subalgebras of real simple Lie algebras appeared. Trans. Groups 27, 1393–1430 (2022)
Ólafsson, G.: Symmetric spaces of Hermitian type. Differ. Geom. Appl. 1, 195–233 (1991)
Ólafsson, G., Ørsted, B.: The holomorphic discrete series of an affine symmetric space and representations with reproducing kernels. Trans. Am. Math. Soc. 326, 385–405 (1991)
Schroer, B.: Wigner representation theory of the Poincaré group, localization, statistics and the S-matrix. Nuclear Phys. B 499–3, 519–546 (1997)
Simon, B.: The \(P(\Phi )_2\) Euclidean (Quantum) Field Theory. Princeton University Press, Princeton (1974)
Thomas, L.J., Wichmann, E.H.: On the causal structure of Minkowski spacetime. J. Math. Phys. 38, 5044–5086 (1997)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The research of K.-H. Neeb was partially supported by DFG-Grant NE 413/10-1. The research of G. Ólafsson was partially supported by Simons Grant 586106.
Appendices
Appendix
1.1 A Irreducible modular ncc symmetric Lie algebras
See Table 1.
B Some calculations in \(\mathop {{\mathfrak {sl}}}\nolimits _2({{\mathbb {R}}})\)
Arguments are often reduced to relatively simple \(\mathop {{\mathfrak {sl}}}\nolimits _2({{\mathbb {R}}})\) calculations. We therefore collect the basic notations and calculations here in one place for reference. For \({{\mathfrak {g}}}= \mathop {{\mathfrak {sl}}}\nolimits _2({{\mathbb {R}}})\), we fix the Cartan involution \(\theta (x) = - x^\top \), so that
The basis elements
and
satisfy
For the involution
we have
is a hyperbolic \(\mathop {\mathrm{Inn}}\nolimits ({{\mathfrak {h}}})\)-invariant cone in \({{\mathfrak {q}}}\), containing \(h^1\) as a causal Euler element. Further
lead to
so that
The subspace \({{\mathfrak {t}}}_{{\mathfrak {q}}}:= {{\mathbb {R}}}(e^0 - f^0) = \mathop {{\mathfrak {so}}}\nolimits _2({{\mathbb {R}}})\) of \({{\mathfrak {q}}}\) is maximal elliptic. For
and
we then have
More generally, we have for \(t \in {{\mathbb {R}}}\)
because \( [e^0 - f^0, h^1] = 2 h^0, \quad [e^0 - f^0, h^0] = -2 h^1. \)
Now fix the Euler element
and similarly drop the \({}^0\) in other places. Then we have
Recall that
We have
For \(y = \lambda e + \mu f \in {{\mathfrak {g}}}^{-\tau _h}\), we have
We define the function \(S : {{\mathbb {C}}}\rightarrow {{\mathbb {C}}}\) by \(S(z) = \sum _{k = 0}^\infty \frac{(-1)^k}{(2k+1)!} z^k\), so that \(\sin (z) = z S(z^2).\) Then
This element is contained in \(-C_{{\mathfrak {g}}}^\circ \) if \(\lambda , \mu > 0\) and \(\sin (2\sqrt{\lambda \mu }) > 0\). This is satisfied for \(0< 4\mu \lambda < \pi ^2\). Then y is hyperbolic with \(\mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits y) = \{ \pm 2 \sqrt{\lambda \mu }\} \subseteq (-\pi , \pi )\).
Example B.1
(cf. Example 6.2) Let G be a connected Lie group with Lie algebra \({{\mathfrak {g}}}\) and \(h \in {{\mathfrak {g}}}\) an Euler element. In general \(G^h\) may be much larger than \(G^{\tau _h}\). This can be seen for \(G={\widetilde{\mathop {\mathrm{SL}}\nolimits }}_2({{\mathbb {R}}})\) (the simply connected covering group of \(\mathop {\mathrm{SL}}\nolimits _2({{\mathbb {R}}})\)) and the Euler element \(h = \mathop {\mathrm{diag}}\nolimits (1/2,-1/2)\). The group \(G^h\) contains \(Z(G) \cong {{\mathbb {Z}}}\) and \(G^h_e = \exp ({{\mathbb {R}}}h)\), which shows that \(\pi _0(G^h) \cong {{\mathbb {Z}}}\). The involution \(\tau _h\) acts on \(Z(G)\subseteq \exp (\mathop {{\mathfrak {so}}}\nolimits _2({{\mathbb {R}}}))\) by inversion (cf. [25, Ex. 2.10(d)]).
For \(G = \mathop {\mathrm{SL}}\nolimits _2({{\mathbb {R}}})\) and the same Euler element h, the centralizer \(G^h\) is the non-connected subgroup of diagonal matrices, which coincides with the fixed point group \(G^{\tau _h}\) of the corresponding involution
On the other hand, if we consider the centerfree group \(G=\mathop {\mathrm{SO}}\nolimits _{1,2}({{\mathbb {R}}})_e \cong \mathop {\mathrm{PSL}}\nolimits _2({{\mathbb {R}}})\), then \(G^h=G^h_e\), but the subgroup \(G^{\tau _h}\) has 2 connected components: \(G^{\tau _h}= G^h\cup \theta G^h\), where \(\theta \) is a Cartan involution with \(\theta (h)=-h\).
C Polar maps
In this section we discuss polar maps associated to an involution on a symmetric space, resp., to a pair of commuting involutions on a Lie group. Key properties are collected in Lemma C.4. These results are used in particular in Sect. 4.2 to obtain a polar decomposition of the crown domain of the Riemannian symmetric space \(M^r = G/K\) and in the characterization of this domain as a subset of the tube domain of the cone C (Theorem 4.10).
1.1 C.1 Some spectral theory
Let V be a finite dimensional real vector space and \(A \in \mathop {\mathrm{End}}\nolimits (V)\).
Lemma C.1
\(\ker \big (\frac{\sinh (A)}{A}\big ) = \bigoplus _{n \not =0} \ker (A^2 + n^2\pi ^2 \mathbf{1 })\).
Proof
We may w.l.o.g. assume that V is complex. Then \(B := \sinh (A)/A\) is invertible on all generalized eigenspace corresponding to eigenvalues \(\lambda \not = \pi n i\), \(n\in {{\mathbb {Z}}}\setminus \{0\}\). We may therefore assume that V has only one eigenvalue \(\lambda = n \pi i\), \(n \not =0\). Then A is invertible, so that
Writing \(A = A_s + A_n\) for the Jordan decomposition of A, it follows that
As \(\ker (e^{2A_n}-\mathbf{1 }) = \ker (A_n)\) follows from \(2A_n = \log (e^{2A_n})\) as a polynomial in \(e^{2A_n}-1\), we see that \(\ker (B) = \ker (A_n)\) is the \(\lambda \)-eigenspace of A. \(\square \)
With similar arguments, or by replacing A by \(A - \frac{\pi i}{2}\mathbf{1 }\), we get:
Lemma C.2
\(\ker (\cosh (A)) = \ker (e^{-2A} + \mathbf{1 }) = \bigoplus _{n \in {{\mathbb {N}}}}\ker \Big (A^2 + \big (n + \frac{1}{2}\big )^2 \pi ^2 \mathbf{1 }\Big )\).
1.2 C.2 Fine points on polar maps
In this subsection, we consider two commuting involutions \(\sigma \) and \(\tau \) on a connected, not necessarily reductive, Lie group G and an open \(\sigma \)-invariant subgroup \(H \subseteq G^\tau \). We shall study the polar map
and its applications.
As H is invariant under \(\tau \) and \(\sigma \), both define commuting involutions on M and their fixed point manifolds intersect transversally in eH. The map
is a diffeomorphism onto the normal bundle \(N(G^\sigma /H^\sigma )\) of the subspace \(G^\sigma .eH \cong G^\sigma /H^\sigma \) and \(\Phi = \mathop {\mathrm{Exp}}\nolimits \circ \Psi \), where \(\mathop {\mathrm{Exp}}\nolimits : T(M) \rightarrow M\) is the exponential map.
First, we determine the regular points of \(\Phi \). As \(\Phi \) is \(G^\sigma \)-equivariant, it suffices to determine for which points [e, x] the tangent map \(T_{[e,x]}(\Phi )\) is injective, hence bijective for dimensional reasons. In the following calculation, we shall use the formula
for the differential of \(\mathop {\mathrm{Exp}}\nolimits \) [12, Lemma 4.6], where \({{\mathfrak {q}}}\rightarrow T_{\mathop {\mathrm{Exp}}\nolimits _{eH}(x)}, v \mapsto \exp x.v\), is the linear isomorphism induced by the action of \(\exp x \in G\) on M. For \(a \in {{\mathfrak {g}}}^\sigma , x,b \in {{\mathfrak {q}}}^{-\sigma }\), we obtain
Note that \(e^{-\mathop {\mathrm{ad}}\nolimits x}a = \cosh (\mathop {\mathrm{ad}}\nolimits x) a - \sinh (\mathop {\mathrm{ad}}\nolimits x) a\). If \(a\in {{\mathfrak {h}}}^{\sigma }\) then \(p_{{\mathfrak {q}}}(e^{-\mathop {\mathrm{ad}}\nolimits x}a) = - \sinh (\mathop {\mathrm{ad}}\nolimits x)a\), and if \(a \in {{\mathfrak {q}}}^\sigma \), then \(p_{{\mathfrak {q}}}(e^{-\mathop {\mathrm{ad}}\nolimits x}a) = \cosh (\mathop {\mathrm{ad}}\nolimits x)a\). Writing \(a = a_{{\mathfrak {h}}}+ a_{{\mathfrak {q}}}\) with \(a_{{\mathfrak {h}}}\in {{\mathfrak {h}}}^\sigma \) and \(a_{{\mathfrak {q}}}\in {{\mathfrak {q}}}^\sigma \), we thus obtain
The following lemma provides a characterization of the regular points.
Lemma C.3
For \(x \in {{\mathfrak {q}}}\), the following assertions hold:
-
(a)
\(\mathop {\mathrm{Exp}}\nolimits _{eH}\) is regular in x if and only if the map \(\frac{\sinh (\mathop {\mathrm{ad}}\nolimits x)}{\mathop {\mathrm{ad}}\nolimits x} : {{\mathfrak {q}}}\rightarrow {{\mathfrak {q}}}\) is invertible, which is equivalent to
$$\begin{aligned} \mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits x\vert _{{{\mathfrak {q}}}_L}) \cap {{\mathbb {Z}}}\pi i \subseteq \{0\}, \quad \text{ where } \quad {{\mathfrak {q}}}_L := {{\mathfrak {q}}}+ [{{\mathfrak {q}}},{{\mathfrak {q}}}]. \end{aligned}$$(C.5) -
(b)
If \(\mathop {\mathrm{Exp}}\nolimits _{eH}\vert _{{{\mathfrak {q}}}^{-\sigma }}\) is regular in \(x \in {{\mathfrak {q}}}^{-\sigma }\), then the polar map \(\Phi \) in (C.2) is regular in [g, x] if and only if, in addition, \(\cosh (\mathop {\mathrm{ad}}\nolimits x) : {{\mathfrak {q}}}^\sigma \rightarrow {{\mathfrak {q}}}^\sigma \) is invertible, which is equivalent to
$$\begin{aligned} \mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits x\vert _{{{\mathfrak {q}}}_L}) \cap \Big (\frac{\pi }{2} + {{\mathbb {Z}}}\pi \Big ) i = \emptyset . \end{aligned}$$(C.6)
Proof
(a) follows from the spectral theoretic description of the kernel of \(\frac{\sinh (\mathop {\mathrm{ad}}\nolimits x)}{\mathop {\mathrm{ad}}\nolimits x}\big |_{{{\mathfrak {q}}}}\) as the intersection of \({{\mathfrak {q}}}\) with the sum of the eigenspaces of \(\mathop {\mathrm{ad}}\nolimits x\) in \({{\mathfrak {g}}}_{{\mathbb {C}}}\) for the eigenvalues \(\lambda \in \pi i {{\mathbb {Z}}}\setminus \{0\}\) (Lemma C.1).
(b) Suppose that the restriction of \(\mathop {\mathrm{Exp}}\nolimits _{eH}\) to \({{\mathfrak {q}}}^{-\sigma }\) is regular, i.e., that \(\frac{\sinh (\mathop {\mathrm{ad}}\nolimits x)}{\mathop {\mathrm{ad}}\nolimits x} : {{\mathfrak {q}}}^{-\sigma } \rightarrow {{\mathfrak {q}}}^{-\sigma }\) is invertible. Then (C.4) shows that \(\Phi \) is regular in [e, x] if and only if \(\cosh (\mathop {\mathrm{ad}}\nolimits x) : {{\mathfrak {q}}}^\sigma \rightarrow {{\mathfrak {q}}}^\sigma \) is invertible, and this is equivalent to the condition on \(\mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits x\vert _{{{\mathfrak {q}}}_L})\) stated in (b). \(\square \)
The following lemma contains a wealth of information on singular points of the polar map \(\Phi \).
Lemma C.4
Let \(\Omega \subseteq {{\mathfrak {q}}}^{-\sigma }\) be an open \(H^\sigma \)-invariant subset consisting of \(\mathop {\mathrm{Exp}}\nolimits \)-regular elliptic elements, and consider the polar map
Let \(x \in \Omega \), write \(m := \mathop {\mathrm{Exp}}\nolimits _{eH}(x) \in M = G/H\), \({\mathcal {O}}_m := G^\sigma _e.m\) for its orbit, and put
Then the following assertions hold:
-
(a)
\({{\mathfrak {g}}}^{-\sigma _x} := \{ y \in {{\mathfrak {g}}}: \sigma _x(y) = - y \} = \ker (\cosh (\mathop {\mathrm{ad}}\nolimits x))\).
-
(b)
\({{\mathfrak {q}}}^{\sigma , -\sigma _x}\) complements the subspace \(\exp (-x).\mathop {\mathrm{im}}\nolimits (T_{[e,x]}(\Phi )) = \cosh (\mathop {\mathrm{ad}}\nolimits x){{\mathfrak {q}}}^\sigma \oplus {{\mathfrak {q}}}^{-\sigma }\) in \({{\mathfrak {q}}}\).
-
(c)
The eigenspaces \({{\mathfrak {g}}}^{\pm \sigma _x}\) are \(\tau \)-invariant, and on the Lie subalgebra \({{\mathfrak {g}}}^{\sigma _x^2} = {{\mathfrak {g}}}^{\sigma _x} \oplus {{\mathfrak {g}}}^{-\sigma _x}\), the involution \(\tau \) commutes with \(\sigma _x\).
-
(d)
The eigenspaces \({{\mathfrak {g}}}^{\pm \sigma _x}\) are \(\zeta _x\)-invariant, on \({{\mathfrak {g}}}^{\sigma _x}\) the automorphisms \(\zeta _x\) and \(\tau \) commute, and on on \({{\mathfrak {g}}}^{-\sigma _x}\) the complex structure \(\zeta _x\) and \(\tau \) anticommute. In particular, we have
$$\begin{aligned} \zeta _x({{\mathfrak {h}}}^{\sigma _x}) = {{\mathfrak {h}}}^{\sigma _x}, \quad \zeta _x({{\mathfrak {q}}}^{\sigma _x}) = {{\mathfrak {q}}}^{\sigma _x}, \quad \zeta _x({{\mathfrak {h}}}^{-\sigma _x}) = {{\mathfrak {q}}}^{-\sigma _x}, \quad \zeta _x({{\mathfrak {q}}}^{-\sigma _x}) = {{\mathfrak {h}}}^{-\sigma _x}.\end{aligned}$$ -
(e)
The eigenspaces \({{\mathfrak {g}}}^{\pm \sigma _x}\) are \(\sigma \)-invariant, and on the Lie subalgebra \({{\mathfrak {g}}}^{\sigma _x^2}\), the involution \(\sigma \) commutes with \(\sigma _x\). On \({{\mathfrak {g}}}^{\sigma _x}\), the automorphisms \(\zeta _x\) and \(\sigma \) commute, and on on \({{\mathfrak {g}}}^{-\sigma _x}\) the complex structure \(\zeta _x\) anticommute with \(\sigma \). In particular, we have
$$\begin{aligned} \zeta _x({{\mathfrak {g}}}^{\sigma ,\sigma _x}) = {{\mathfrak {g}}}^{\sigma ,\sigma _x}, \quad \zeta _x({{\mathfrak {g}}}^{-\sigma ,\sigma _x}) = {{\mathfrak {g}}}^{-\sigma , \sigma _x}, \quad \zeta _x({{\mathfrak {g}}}^{\pm \sigma , -\sigma _x}) = {{\mathfrak {g}}}^{{\mp }\sigma , -\sigma _x}.\end{aligned}$$ -
(f)
The stabilizer Lie algebra of m in \({{\mathfrak {g}}}\) is
$$\begin{aligned} {{\mathfrak {g}}}_m = {{\mathfrak {g}}}^{\tau \sigma _x}= \{ y \in {{\mathfrak {g}}}: \sigma _x(y) = \tau (y)\} \quad \text{ and } \quad {{\mathfrak {g}}}_m^{\sigma _x^2} = {{\mathfrak {h}}}^{\sigma _x} \oplus {{\mathfrak {q}}}^{-\sigma _x}.\end{aligned}$$The stabilizer Lie algebra in \({{\mathfrak {g}}}^\sigma \) is
$$\begin{aligned} {{\mathfrak {g}}}_m^\sigma ={{\mathfrak {h}}}^{\sigma ,\sigma _x} \oplus {{\mathfrak {q}}}^{\sigma , -\sigma _x}. \end{aligned}$$(C.7)The stabilizer group \(G_m\) acts on
$$\begin{aligned} T_m(M) = \exp x.{{\mathfrak {q}}}\, by \, g.(\exp x.y) = \exp x.\big (\mathop {\mathrm{Ad}}\nolimits (\zeta _x^G(g))y\big ). \end{aligned}$$ -
(g)
The tangent space of the orbit \({\mathcal {O}}_m\) is
$$\begin{aligned} T_m({\mathcal {O}}_m) = \exp (x).\big (\cosh (\mathop {\mathrm{ad}}\nolimits x){{\mathfrak {q}}}^\sigma + [x,{{\mathfrak {h}}}^\sigma ]\big ).\end{aligned}$$ -
(h)
Let \({{\mathfrak {q}}}_x := {{\mathfrak {q}}}^{\sigma ,-\sigma _x} + {{\mathfrak {q}}}^{-\sigma ,\sigma _x}\) and \({{\mathfrak {h}}}_x := [{{\mathfrak {q}}}_x,{{\mathfrak {q}}}_x]\). If the group \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}({{\mathfrak {h}}}_x)\) acts as a relatively compact group on \({{\mathfrak {q}}}_x\) and \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}({{\mathfrak {h}}}_x){{\mathfrak {q}}}^{-\sigma ,\sigma _x} = {{\mathfrak {q}}}_x\), then m is an interior point of \(\mathop {\mathrm{im}}\nolimits (\Phi )\).
Proof
(a) follows directly from \(2\cosh (\mathop {\mathrm{ad}}\nolimits x) = e^{\mathop {\mathrm{ad}}\nolimits x} + e^{-\mathop {\mathrm{ad}}\nolimits x}\).
(b) With (C.3) we see that the image of \(T_{[e,x]}(\Phi )\) is the subspace
As \(\mathop {\mathrm{ad}}\nolimits x\) is semisimple, a complement of \(\exp (-x).\mathop {\mathrm{im}}\nolimits (T_{[e,x]}(\Phi ))\) in \({{\mathfrak {q}}}\) is
(c) In view of \(\tau \sigma _x^2 \tau = \sigma _x^{-2}\), the fixed point space \({{\mathfrak {g}}}^{\sigma _x^2}\) is \(\tau \)-invariant. On this subspace \(\sigma _x= \sigma _x^{-1} = \tau \sigma _x \tau \), so that \(\tau \) preserves the two eigenspaces \({{\mathfrak {g}}}^{\pm \sigma _x}\) of \(\sigma _x\) on \({{\mathfrak {g}}}^{\sigma _x^2}\).
(d) As \(\zeta _x\) commutes with \(\sigma _x\), the eigenspaces \({{\mathfrak {g}}}^{\pm \sigma _x}\) are \(\zeta _x\)-invariant. Further, (c) implies that, on \({{\mathfrak {g}}}^{\sigma _x}\), we have \(\tau \zeta _x \tau = \zeta _x^{-1} = \zeta _x\).
(e) is shown with similar arguments as (c) and (d).
(f) In \(M = G/H\), the point \(m = \mathop {\mathrm{Exp}}\nolimits _{eH}(\exp x) = \exp x H\) is obtained by acting with \(\exp x\) on the base point eH. Therefore its stabilizer group is \(G_m = \exp x H \exp (-x)\) with the Lie algebra
Now the \(\tau \)-invariance of \({{\mathfrak {g}}}^{\sigma _x^2}\) implies that
To verify (C.7), let \(y = y_{{\mathfrak {h}}}+ y_{{\mathfrak {q}}}\in {{\mathfrak {g}}}^\sigma \) with \(y_{{\mathfrak {h}}}\in {{\mathfrak {h}}}^\sigma \) and \(y_{{\mathfrak {q}}}\in {{\mathfrak {q}}}^\sigma \). Then the corresponding vector field \(Y_M\) on M satisfies
Therefore \(Y_M(m) = 0\) is equivalent to
Thus both summands have to vanish, which is equivalent to
This implies (C.7).
To complete the proof of (f), we note that, for \(v \in {{\mathfrak {q}}}\) and \(g \in G_m\), we have
where \(\zeta ^G_x(G_m) = H\) acts on \(T_{eH}(M) \cong {{\mathfrak {q}}}\) by the adjoint representation.
(g) From (C.9) it follows that
Here the last equality follows from \((\mathop {\mathrm{ad}}\nolimits x)^2 {{\mathfrak {h}}}^\sigma \subseteq {{\mathfrak {h}}}^\sigma \) and the \(\mathop {\mathrm{Exp}}\nolimits \)-regularity of x, which, by Lemma C.3, is equivalent to the invertibility of \(\frac{\sinh (\mathop {\mathrm{ad}}\nolimits x}{\mathop {\mathrm{ad}}\nolimits x}\) on \({{\mathfrak {q}}}\).
(h) By (C.8), a natural complement of
in \({{\mathfrak {q}}}\) is the subspace
where the last equality follows from
As x is \(\mathop {\mathrm{Exp}}\nolimits _{eH}\)-regular and
the subset
contains a 0-neighborhood \(U_0\) in \({{\mathfrak {q}}}^{-\sigma ,\sigma _x}\). Now our assumption \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}({{\mathfrak {h}}}_x) {{\mathfrak {q}}}^{-\sigma ,\sigma _x} ={{\mathfrak {q}}}_x\) and the relative compactness of the group \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}({{\mathfrak {h}}}_x)\) on \({{\mathfrak {q}}}_x\) imply that \(U_1 := \mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}({{\mathfrak {h}}}_x) U_0\) is a 0-neighborhood in \({{\mathfrak {q}}}_x\).
Finally, we obtain from (C.7)
so that
This means that \(\Omega \) is a 0-neighborhood in \({{\mathfrak {q}}}_x\), and hence that \(\mathop {\mathrm{im}}\nolimits (\Phi )\) is a neighborhood of m because the map
has surjective differential in (e, 0). \(\square \)
Remark C.5
If \(\rho _i(\mathop {\mathrm{ad}}\nolimits x) < \pi /2\), then the polar map \(\Phi \) in the preceding lemma is regular in [e, x] (Lemma C.3). In this case \(\ker (\cosh (\mathop {\mathrm{ad}}\nolimits x)) = {{\mathfrak {g}}}^{-\sigma _x}\) is trivial, so that
follows from Lemma C.4(f).
The next lemma is useful to verify condition (h) in the preceding lemma.
Lemma C.6
Suppose that \(({{\mathfrak {g}}},\tau )\) is a reductive symmetric Lie algebra such that \({{\mathfrak {q}}}\) consist of elliptic elements, \(\sigma \tau = \tau \sigma \), and that \({{\mathfrak {a}}}\subseteq {{\mathfrak {q}}}^{-\sigma }\) is a maximal abelian subspace in \({{\mathfrak {q}}}\). Then
and the group \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}([{{\mathfrak {q}}},{{\mathfrak {q}}}])\) is compact.
Proof
Let \({{\mathfrak {a}}}\subset {{\mathfrak {q}}}\) be maximal abelian in \({{\mathfrak {q}}}\). Then \({{\mathfrak {a}}}\) contains \({{\mathfrak {z}}}({{\mathfrak {q}}})\). We then note that \({{\mathfrak {z}}}_{{\mathfrak {g}}}({{\mathfrak {a}}}) = {{\mathfrak {a}}}\oplus {{\mathfrak {z}}}_{{\mathfrak {h}}}({{\mathfrak {a}}})\) is \(\tau \)- and \(\sigma \)-invariant. Further \({{\mathfrak {g}}}= {{\mathfrak {z}}}_{{\mathfrak {g}}}({{\mathfrak {a}}}) \oplus [{{\mathfrak {a}}},{{\mathfrak {g}}}]\) implies that \({{\mathfrak {q}}}= {{\mathfrak {a}}}\oplus [{{\mathfrak {a}}},{{\mathfrak {h}}}].\) As
this leads to
The second assertion follows from \(\mathop {\mathrm{Inn}}\nolimits _{{\mathfrak {g}}}([{{\mathfrak {q}}},{{\mathfrak {q}}}]){{\mathfrak {a}}}= {{\mathfrak {q}}}\), and the third from the fact that the reductive Lie algebra \({{\mathfrak {q}}}+ [{{\mathfrak {q}}},{{\mathfrak {q}}}]\) (it is an ideal of \({{\mathfrak {g}}}\)) is compact. \(\square \)
1.3 C.3 Fibers of the polar map
For the polar map we have to analyze the relation
Applying the quadratic representation yields
For \(x,y \in {{\mathfrak {q}}}^{-\sigma }\) and \(g \in G^\sigma \), we also have \(gg^\sharp \in G^\sigma \), so that
If x and y are sufficiently small (imaginary spectral radius \(< \frac{\pi }{4}\)), we thus obtain \(\mathop {\mathrm{Ad}}\nolimits (g)y = x\), and thus \(gg^\sharp = e\), i.e., \(\tau (g) = g\).
See [16, Cor. 7.35] for similar arguments.
1.4 C.4 Fibers of \(\mathop {\mathrm{Exp}}\nolimits \)
Suppose that \(x, y \in {{\mathfrak {q}}}\) have the same exponential image \(\mathop {\mathrm{Exp}}\nolimits (x) = \mathop {\mathrm{Exp}}\nolimits (y)\) in M. We further assume that \(\mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits x) \cap i \pi {{\mathbb {Z}}}\subseteq \{0\}\), so that \(\mathop {\mathrm{Exp}}\nolimits \) is regular in x. Then we obtain in G the identity
and since \(\mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits (2x)) \cap 2 \pi i {{\mathbb {Z}}}\subseteq \{0\}\), \(\exp \) is regular in x. Therefore [17, Lemma 9.2.31] implies that
We conclude that \(\exp (x-y) = \exp (y-x)\), which leads to
so that \(\mathop {\mathrm{Exp}}\nolimits (y-x) \in M^\tau \), and \(\mathop {\mathrm{Exp}}\nolimits ({{\mathbb {R}}}(x-y)) \subseteq M\) is a closed geodesic.
We also conclude that \(x-y\) is elliptic with \(\mathrm{Spec}(\mathop {\mathrm{ad}}\nolimits (x-y))\subseteq \pi i {{\mathbb {Z}}}\).
Lemma C.7
If \(\rho _i(\mathop {\mathrm{ad}}\nolimits x), \rho _i(\mathop {\mathrm{ad}}\nolimits y) < \pi ,\) then \(\exp (x) = \exp (y)\) implies \(x-y \in {{\mathfrak {z}}}({{\mathfrak {g}}})\).
If, in addition, G is simply connected or \({{\mathfrak {g}}}\) is semisimple, then \(x = y\).
Proof
The preceding discussion implies that \([x,y] = 0\). Now \(\rho _i(\mathop {\mathrm{ad}}\nolimits (x-y)) < 2\pi \) leads to \(\mathop {\mathrm{ad}}\nolimits (x-y) = 0\), i.e., to \(x-y \in {{\mathfrak {z}}}({{\mathfrak {g}}})\). \(\square \)
D Quadrics as symmetric spaces
In this appendix we discuss an important example of a non-compactly causal symmetric spaces: d-dimensional de Sitter space \(\mathop {\mathrm{dS}}\nolimits ^d\), realized as a hyperboloid in Minkowski space.
1.1 D.1 Quadrics as symmetric spaces
Definition D.1
[24] (a) Let M be a smooth manifold and
be a smooth map with the following properties: each \(s_x\) is an involution for which x is an isolated fixed point and
Then we call \((M,\mu )\) a symmetric space.
(b) A morphism of symmetric spaces M and N is a smooth map \(\varphi : M\rightarrow N\) such that \(\varphi (x\cdot y)= \varphi (x)\cdot \varphi (y)\) for \(x, y \in M\).
(c) A geodesic of a symmetric space is a morphism \(\gamma : {{\mathbb {R}}}\rightarrow M\) of symmetric spaces, i.e.,
Any geodesic is uniquely determined by \(\gamma '(0) \in T_{\gamma (0)}(M)\), and, conversely, every \(v \in T_p(M)\) generates a unique geodesic \(\gamma _v\) with \(\gamma _v(0) = p\) and \(\gamma _v'(0) = v\). Accordingly, geodesics are encoded in the exponential functions
We then have
Example D.2
Let \((V,\beta )\) be a finite dimensional real vector space, endowed with a non-degenerate symmetric bilinear form \(\beta \). Then every anisotropic element \(x \in V\) defines an involution
fixing x and satisfying \(\mathop {\mathrm{Fix}}\nolimits (-s_x) = x^\bot = \{ y \in V: \beta (x,y) = 0\}\).
For \(c \in {{\mathbb {R}}}^\times \), let
denote the corresponding quadric in \((V,\beta )\). Then \((Q_c, \mu )\) with \(\mu (x,y) = s_x(y)\) is a symmetric space (Definition D.1) and \(\mathop {\mathrm{dim}}\nolimits Q_c = \mathop {\mathrm{dim}}\nolimits V -1\). Note that dilation by \(r \in {{\mathbb {R}}}^\times \) is an isomorphism of symmetric spaces from \(Q_c\) to \(Q_{r^2c}\).
For \(p \in Q_c\), the tangent space is \(T_p(Q_c) = p^\bot = \{ v \in V : \beta (p,v) = 0\}\). To describe the exponential function of the symmetric space \(Q_c\), we use the entire functions \(C, S : {{\mathbb {C}}}\rightarrow {{\mathbb {C}}}\) defined by
which satisfy
and
Note that
follows from \(1 = \cos ^2 z + \sin ^2 z\) and the surjectivity of the square map on \({{\mathbb {C}}}\).
Proposition D.3
The exponential function of the symmetric space \(Q_c\) is given by
Proof
To verify this claim, abbreviate \(\varepsilon := \frac{\beta (v,v)}{\beta (p,p)}\). By (D.5), on the right hand side of (D.6), the quadratic form \(\beta (\cdot ,\cdot )\) has the value
Therefore the right hand side of (D.6) is contained in \(Q_c\).
It remains to show that, for \(\varepsilon \in \{-1,0,1\}\), the curve
which satisfies \(\gamma _v'(0) = v\), actually is a geodesic, i.e.,
We consider three cases:
\(\varepsilon = 0\): Then \(\gamma _v(t) = p + tv\) and
\(\varepsilon = 1\): Then
by (D.3), and
\(\varepsilon = -1\): Then
by (D.4), and
\(\square \)
1.2 D.2 Minkowski space
On \(V := {{\mathbb {R}}}^{1,d}\), we consider the Lorentzian form
the future light cone
and the tube domain
The standard boost vector field \(X_h(v) = hv\) is defined by \(h \in \mathop {{\mathfrak {so}}}\nolimits _{1,d}({{\mathbb {R}}})\), given by
It generates the flow
and defines the involution
that we extend to an antilinear involution \(\overline{\tau }_h\) on \(V_{{\mathbb {C}}}\). It also defines a Wick rotation
satisfying \(\kappa _h^2 = \tau _h\).
Lemma D.4
The following subsets of V are equal:
-
(a)
The standard right wedge \(W_R := \{ x \in V : x_1 > |x_0| \}\).
-
(b)
The positivity domain \(W_V^+(h) := \{ v \in V : X_h(v) \in V_+ \}\) of \(X_h(v) = hv\).
-
(c)
The KMS domain of \(\alpha \): \(\{ x \in V : (\forall t \in (0,\pi ))\ \alpha _{it}(x) \in {\mathcal {T}}_V\}\).
-
(d)
The Wick rotation \(\kappa _h({\mathcal {T}}_V^{\overline{\tau }_h})\) of the fixed point set \({\mathcal {T}}_V^{\overline{\tau }_h} = {\mathcal {T}}_V \cap (V^{\tau _h} + i V^{-\tau _h})\) of the antiholomorphic involution \(\overline{\tau }_h\) on \({\mathcal {T}}_V\).
Proof
The equality of \(W_R\) and \(W_V^+(h)\) follows immediately from (D.7). For \(0< t < \pi \), we have
This implies the equality of the sets in (a) and (c). Finally, we observe that
implies that \(\kappa _h({\mathcal {T}}_V^{\overline{\tau }_h}) = W_R\). \(\square \)
Remark D.5
For the closed light cone \(C := \overline{V_+}\) and the h-eigenspaces
so that we obtain the following description of the standard right wedge
1.3 D.3 De Sitter space \(\mathop {\mathrm{dS}}\nolimits ^d, d \ge 2\)
We write \(G := \mathop {\mathrm{SO}}\nolimits _{1,d}({{\mathbb {R}}})^{\uparrow }\) for the connected Lorentz group acting on Minkowski space \((V,[\cdot ,\cdot ])\) and consider the de Sitter space
The purpose of this appendix is derive our main results for the example \(\mathop {\mathrm{dS}}\nolimits ^d\) by direct calculations without the elaborate structure theory developed for the general case. By Proposition D.3, the exponential function of the symmetric space \(\mathop {\mathrm{dS}}\nolimits ^d\) is given by
so that spacelike vectors v generate closed geodesics.
De Sitter space inherits the structure of a non-compactly causal symmetric space from the embedding into Minkowski space \((V,V_+)\). In particular, the positive cone in a point \(x \in \mathop {\mathrm{dS}}\nolimits ^d\) is given by
We keep the notation h, \(\alpha _t\) and \(X_h\) from our discussion of Minkowski space in Subsection D.2. As \(\mathop {\mathrm{dS}}\nolimits ^d\) is \(\alpha \)-invariant, the vector field \(X_h\) is tangential to \(\mathop {\mathrm{dS}}\nolimits ^d\), and Lemma D.4 immediately implies that its positivity domain is given by
The centralizer of h in G is the subgroup
(cf. [31, Lemma 4.12]).
Lemma D.6
For
we have
Proof
First we observe that both sides of (D.9) are G-invariant.
“\(\supseteq \)”: We have to show that any \(x \in V_+^\pi ({\mathbf{e}}_1)\) satisfies \(\mathop {\mathrm{Exp}}\nolimits _{{\mathbf{e}}_1}(ix) \in V + i V_+\). Since the orbit of x under \(G_{{\mathbf{e}}_1}\) contains an element in \({{\mathbb {R}}}{\mathbf{e}}_0\), we may assume that \(x = x_0 {\mathbf{e}}_0\) with \(0< x_0 < \pi \). Then
follows from \(\sin (x_0) > 0\).
“\(\subseteq \)”: If \(z = x + i y\in \mathop {\mathrm{dS}}\nolimits ^d_{{\mathbb {C}}}\cap (V + i V_+)\), then \([z,z] = -1\) and \(y \in V_+\). Acting with G, we may thus assume that \(y = y_0 {\mathbf{e}}_0\) with \(y_0 > 0\). Then \(x \in y^\bot = {\mathbf{e}}_0^\bot \) follows from \(\mathop {\mathrm{Im}}\nolimits [z,z] = 0\), so that we may further assume that \(x = x_1 {\mathbf{e}}_1\) for some \(x_1 \ge 0\). Now \(z = i y_0 {\mathbf{e}}_0 + x_1 {\mathbf{e}}_1 \in \mathop {\mathrm{dS}}\nolimits ^d_{{\mathbb {C}}}\) implies that
Hence there exists a \(t \in (0,\pi )\) with \(y_0 = \sin t\) and \(x_1 = \cos t\). This leads to
\(\square \)
Definition D.7
The complex manifold
is called the tube domain of \(\mathop {\mathrm{dS}}\nolimits ^d\). If coincides with \(G.\mathop {\mathrm{Exp}}\nolimits _m(i V_+^\pi (m))\) for every \(m \in \mathop {\mathrm{dS}}\nolimits ^d\).
Starting from the relation \({\mathcal {T}}_{\mathop {\mathrm{dS}}\nolimits ^d} = G.\mathop {\mathrm{Exp}}\nolimits _{{\mathbf{e}}_2}(i V_+^\pi ({\mathbf{e}}_2))\), the invariance of \({\mathbf{e}}_2\) under \(\tau _h = e^{\pi i h}\) permits us to obtain a nice description of the fixed point set of \(\overline{\tau }_h\) on \({\mathcal {T}}_{\mathop {\mathrm{dS}}\nolimits ^d}\).
Lemma D.8
\({\mathcal {T}}_{\mathop {\mathrm{dS}}\nolimits ^d}^{\overline{\tau }_h} = G^h.\mathop {\mathrm{Exp}}\nolimits _{{\mathbf{e}}_2}(i V_+^\pi ({\mathbf{e}}_2)^{-\tau _h}).\)
Proof
We clearly have
We also note that
Next we observe that
Hence the elements \(z \in {\mathcal {T}}_{\mathop {\mathrm{dS}}\nolimits ^d}^{\overline{\tau }_h}\) are of the form
Therefore the orbit of z under \(G^h = \exp ({{\mathbb {R}}}h) \mathop {\mathrm{SO}}\nolimits _{d-1}({{\mathbb {R}}})\) (see (D.8)) contains an element of the form \(y = (iy_0, 0, y_2,0,\ldots ,0)\) with \(y_0 > 0\) and \(y_0^2 + y_2^2 = 1\). Hence there exists a \(t \in (0,\pi )\) with \(y_0 = \sin (t)\) and \(y_2 = \cos (t)\). Then
This implies that
\(\square \)
Proposition D.9
The following subsets of de Sitter space \(M =\mathop {\mathrm{dS}}\nolimits ^d\) are equal:
-
(a)
\(W_R \cap \mathop {\mathrm{dS}}\nolimits ^d = \{ x \in \mathop {\mathrm{dS}}\nolimits ^d : x_1 > |x_0|\}\).
-
(b)
\(W_M^+(h) := \{ x \in \mathop {\mathrm{dS}}\nolimits ^d : X_h(x) \in V_+(x) \}\) (the positivity domain of \(X_h\)).
-
(c)
\(W_M^{\mathrm{KMS}}(h) = \{ x \in \mathop {\mathrm{dS}}\nolimits ^d : (\forall t \in (0,\pi )) \ \alpha _{it}(x) \in {\mathcal {T}}_M\}\) (the KMS domain of \(\alpha \)).
-
(d)
\(\kappa _h({\mathcal {T}}_M^{\overline{\tau }_h})\).
-
(e)
\(W_M(h) = G^h.\mathop {\mathrm{Exp}}\nolimits _{{\mathbf{e}}_2}((C_+^\circ + C_-^\circ )^\pi )\).
Proof
The equality of the sets under (a) and (b) follows from Lemma D.4.
As \(\alpha _{it}(m) \in M_{{\mathbb {C}}}\) for every \(t \in {{\mathbb {R}}}\) and \(m \in M\), Lemma D.6 shows that the set under (c) coincides with \(W_V^+(h) \cap M = W_M ^+(h)\).
From the proof of Lemma D.8, we recall that
Now \(\kappa _h(x) = (-i x_1, -i x_0, x_2, \ldots , x_d)\) implies
Combining Lemma D.8 with (d), we finally obtain
for \(C_\pm ^\circ = {{\mathbb {R}}}_+ ({\mathbf{e}}_1 \pm {\mathbf{e}}_0).\) \(\square \)
Remark D.10
For the \(\alpha \)-fixed base point \({\mathbf{e}}_2 \in \mathop {\mathrm{dS}}\nolimits ^d\) the cone
is an infinitesimal version of the wedge domain in \(M = \mathop {\mathrm{dS}}\nolimits ^d\).
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Neeb, KH., Ólafsson, G. Wedge domains in non-compactly causal symmetric spaces. Geom Dedicata 217, 30 (2023). https://doi.org/10.1007/s10711-022-00755-x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10711-022-00755-x