Dynamics On Lorentz Manifolds 1st Scot Adams
download
https://ebookbell.com/product/dynamics-on-lorentz-manifolds-1st-
scot-adams-2148120
Explore and download more ebooks at ebookbell.com
Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Dynamics On Lorentz Manifolds Scot Adams
https://ebookbell.com/product/dynamics-on-lorentz-manifolds-scot-
adams-2632152
Dynamics On The Riemann Sphere A Bodil Branner Festschrift Hjorth P
https://ebookbell.com/product/dynamics-on-the-riemann-sphere-a-bodil-
branner-festschrift-hjorth-p-2044090
Dynamics On And Of Complex Networks Volume 2 Applications To
Timevarying Dynamical Systems 1st Edition Przemyslaw A Grabowicz
https://ebookbell.com/product/dynamics-on-and-of-complex-networks-
volume-2-applications-to-timevarying-dynamical-systems-1st-edition-
przemyslaw-a-grabowicz-4230480
Dynamics On And Of Complex Networks Iii Machine Learning And
Statistical Physics Approaches 1st Ed Fakhteh Ghanbarnejad
https://ebookbell.com/product/dynamics-on-and-of-complex-networks-iii-
machine-learning-and-statistical-physics-approaches-1st-ed-fakhteh-
ghanbarnejad-10486534
Dynamics On And Of Complex Networks Applications To Biology Computer
Science And The Social Sciences 1st Edition Sitabhra Sinha Auth
https://ebookbell.com/product/dynamics-on-and-of-complex-networks-
applications-to-biology-computer-science-and-the-social-sciences-1st-
edition-sitabhra-sinha-auth-1345258
Ecological Dynamics On Yellowstones Northern Range Committee On
Ungulate Management In Yellowstone National Park
https://ebookbell.com/product/ecological-dynamics-on-yellowstones-
northern-range-committee-on-ungulate-management-in-yellowstone-
national-park-4591028
Holomorphic Dynamics On Hyperbolic Riemann Surfaces Marco Abate
https://ebookbell.com/product/holomorphic-dynamics-on-hyperbolic-
riemann-surfaces-marco-abate-47604214
Holomorphic Dynamics On Hyperbolic Riemann Surfaces Marco Abate
https://ebookbell.com/product/holomorphic-dynamics-on-hyperbolic-
riemann-surfaces-marco-abate-50378734
Vegetation Dynamics On The Mountains And Plateaus Of The American
Southwest 1st Edition John L Vankat Auth
https://ebookbell.com/product/vegetation-dynamics-on-the-mountains-
and-plateaus-of-the-american-southwest-1st-edition-john-l-vankat-
auth-4290070
ynamcs on
lonenrz Manifolds
SCOT Adams
World Scientific
ynawics on
loKenrz Mampolds
ynaancs on
lonemz Mamplds
SCOT Adams
University of Minnesota
USA
V f e World Scientific
WH New Jersey • London •'Sim
Singapore • Hong Kong
Published by
World Scientific Publishing Co. Pte. Ltd.
P O Box 128, Farrer Road, Singapore 912805
USA office: Suite IB, 1060 Main Street, River Edge, NJ 07661
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
DYNAMICS ON LORENTZ MANIFOLDS
Copyright © 2001 by World Scientific Publishing Co. Pte. Ltd.
All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright
Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to
photocopy is not required from the publisher.
ISBN 981-02-4382-0
Printed in Singapore by World Scientific Printers
To IOJIHH and AjieKceft
Preface
We begin with a quote from a biography of Felix Klein taken from the site
http://www-history.mcs.st-andrews.ac.uk/history/Mathematicians/Klein.html
on 8 November 1999:
Klein's synthesis of geometry as the study of the prop-
erties of a space that are invariant under a given group of
transformations, known as the Erlanger Programme (1872),
profoundly influenced mathematical development . . . The
Erlanger Programme gave a unified approach to geometry
which is now the standard accepted view.
Transformations play a major role in modern mathe-
matics and Klein showed how the essential properties of a
given geometry could be represented by the group of trans-
formations that preserve those properties.
In other words, central to Klein's Erlanger program is the idea of study-
ing geometry through symmetry. This is the main theme of this book,
applied specifically to a class of geometric objects called Lorentz manifolds.
Contents
Preface vii
Chapter 1 Introduction, History and Outline 1
1.1 Lorentz manifolds and relativity . . 1
1.2 Symmetries of Lorentz manifolds 5
1.3 Outline of succeeding chapters 7
1.4 Notation 8
1.5 Acknowledgements 9
Chapter 2 Basic Results and Definitions 11
2.1 Some set-theoretic notions 11
2.2 Some group-theoretic notions 14
2.3 Some topological notions 15
2.4 Some notions from linear algebra 18
2.5 Matrix concentration lemmas 22
2.6 First results on expansive sequences 23
2.7 Topological groups 25
2.8 Discrete groups 29
2.9 Proper actions 31
2.10 Bilinear and quadratic forms 32
2.11 Root systems 35
2.12 Minkowski forms - basic definitions 39
Chapter 3 Basic Differential Topology 41
3.1 Some differential topological notions 41
ix
x Contents
3.2 Inheritability of continuity and smoothness to leanike submani-
folds 47
3.3 Definition of prefoliation and foliation 49
3.4 Preliminary results to the Frobenius Theorem 51
3.5 Uniqueness in the Frobenius Theorem 52
3.6 Passage from local to global in the Frobenius Theorem 54
3.7 The Frobenius Theorem 55
3.8 Potential submersions 57
3.9 Lorentz metrics - basic definitions 62
Chapter 4 Basic Lie Theoretic Results 65
4.1 Some Lie theoretic definitions and notation 65
4.2 Dynamical consequences of the Frobenius Theorem 70
4.3 exp, Ad and ad 75
4.4 The Lie group Lie algebra correspondence 80
4.5 Some facts about Lie subgroups 83
4.6 The Lie algebra oi[A,B] 90
4.7 Lie groups and Lie algebras from bilinear and quadratic forms 91
4.8 Abelian Lie groups 93
4.9 Miscellaneous results 96
4.10 Generalities on semisimple groups and algebras 112
4.11 Real Jordan decomposition 128
4.12 Consequences of results on real Jordan decomposition 131
4.13 Generalities on algebraic groups 134
4.14 Generalities on nilpotent groups and algebras 138
4.15 Generalities on the nilradical 141
4.16 Relationships between representation theories 149
Chapter 5 More Lie Theory 151
5.1 Connection-preserving diffeomorphisms form a Lie group . . . . 151
5.2 The isometry group of a pseudoRiemannian manifold is a Lie
group 153
5.3 More results on expansive sequences 154
5.4 Lie groups densely embedded in other Lie groups 156
5.5 Generalities on the Levi decomposition 160
5.6 Large normalizers and centralizers 163
5.7 Representation theory 166
Contents xi
Chapter 6 Minkowski Linear Algebra 175
6.1 Notations for important elements and Lie subalgebras of so(Qa) 175
6.2 Linear algebra of Minkowski vector spaces 176
6.3 Basic calculations 184
6.4 Embeddings of Lorentz Lie algebras 192
Chapter 7 Basic Dynamical Results 197
7.1 Kowalsky's Lemma 197
7.2 Higher jets of vector fields and metrics - notation 198
7.3 Matrix realizations of jets and calculus on jets 202
7.4 Miscellaneous results 206
7.5 A basic collection of rigidity results 221
7.6 Strongly lightlike and nontimelike vectors 224
7.7 Basic results on degenerate orbits 225
7.8 More on strongly lightlike and nontimelike vectors 227
7.9 Nonproperness and cocompact subgroups 235
7.10 Kowalsky subsets 236
7.11 Types of chaotic actions 238
7.12 Induction of actions: Definition 249
7.13 Induction of actions: Basic results 250
7.14 Riemannian dynamics 251
Chapter 8 Examples of Actions on Compact Lorentz Mani-
folds 261
8.1 The isometry group of a compact quotient of SL2(M) 261
8.2 Twisted Heisenberg groups 262
8.3 SL 2 (E), twisted Heisenberg and closure 268
Chapter 9 Examples of Nonproper Actions 271
9.1 Restriction-Induction 272
9.2 Examples constructed using a quadratic form in the Lie algebra 276
9.3 A nilpotent Lie group without nonproper Lorentz dynamics . . 278
9.4 Groups with SO(n, 1) or SO(n,2) as a local direct factor . . . . 280
9.5 Groups with a normal subgroup isomorphic to K 284
Chapter 10 Semisimple Groups Admitting a Nonproper Ac-
tion 285
10.1 Locally free actions of SL3(K) 285
xii Contents
10.2 Locally faithful actions of SL3(R) 287
10.3 Locally faithful actions of Sp3(K) C R 6 x 6 288
10.4 Locally faithful actions of SU(2,1) 290
10.5 Kowalsky's theorems 291
Chapter 11 Groups with Action on a Compact Lorentz Man-
ifold 293
11.1 Local freeness 293
11.2 Nilpotent Lie groups 297
11.3 Solvable nonnilpotent Lie groups with Abelian nilradical . . . . 298
11.4 Solvable nonnilpotent Lie groups with nonAbelian nilradical -
preliminaries 300
11.5 Solvable nonnilpotent Lie groups with nonAbelian nilradical -
final results 306
11.6 Semisimple Lie groups 310
11.7 The general case 311
Chapter 12 The Isometry Group of a Compact Lorentz Man-
ifold 315
12.1 Oseledec splitting 316
12.2 Preliminaries 317
12.3 A candidate for the new Killing field 319
12.4 Final classification result 326
Chapter 13 Highly Symmetric Compact Lorentz Manifolds 327
13.1 SL2(K)-actions on compact Lorentz manifolds 327
13.2 Actions of twisted Heisenberg groups on compact Lorentz man-
ifolds - a sketch 331
Chapter 14 Locally Free Orbit Nonproper Lorentz Actions 335
14.1 Preliminary result to degeneration of Adg^j 336
14.2 Degeneration of Ad0<7j 337
14.3 Locally free actions of simply connected Lie groups 341
14.4 Compact subgroups not in the radical 345
14.5 Sequences with no large Kowalsky subsets 347
14.6 Compact connected stabilizers 350
Chapter 15 Orbit Nonproper Lorentz Actions 357
15.1 Preliminaries 357
Contents xiii
15.2 A strongly vanishing element 363
15.3 A strongly lightlike element 366
15.4 Nilradical is free 367
15.5 Nilradical is not free, but its center is free, Part I 372
15.6 Nilradical is not free, but its center is free, Part II 375
15.7 The main result 377
Appendix A The Borel Density Theorem 381
A.l Notation and terminology 381
A.2 Preliminaries 382
A.3 The Borel Density Theorem 383
Appendix B Tameness of Algebraic Actions in Characteristic
Zero 387
Bibliography 391
Index 395
Chapter 1
Introduction, History and Outline
We begin with a very elementary motiviation of Lorentz geometry, to be
followed by a more technical description of the contents of this book.
1.1 Lorentz manifolds and relativity
To understand the importance of Lorentz manifolds in physics, assume
that the speed of light is 1 and imagine, at time t = — 1, a two-dimensional
universe which is empty except for the following:
(1) A particle called "he", located at (0,0) and not moving.
(2) A particle called "she", located at (—1/2,0) and moving at veloc-
ity 1/2 toward him.
(3) For each point on the circle x2 + y2 = 1, a particle of light moving
at speed 1 toward him.
In this male-centric universe, he perceives that all things are moving
toward him and that he is located at the center of all the action. This
proceeds until the great romantic moment t — 0, when he and she come
together and are bathed in light from all sides. Then, alas, our two main
players separate as do the particles of light. All objects in this universe
then move away from him for the rest of time.
So, for example, at time t = 8, he is still located at (0,0) and she
is at (4,0). The light is located at the locus of points (x,y) satisfying
x2 + y2 = 8 2 , i.e., the light forms a circle of radius 8 about the origin.
To complete the picture, one should also imagine that all of these par-
ticles existed infinitely far in the past. So, for example, at time t = —2, she
l
2 Introduction, History and Outline
would be located at (—1,0). Also, at t = - 2 , the light would form a circle
x2 + y2 = 2 2 twice as large as at time t = — 1.
To see everything at once, we can imagine a three-dimensional "space-
time", with axes labeled x, y and i. Our original universe is obtained by
slicing our space-time with the plane t = — 1. In this slice, we see a circle
of light located at points {(x,y, - 1 ) | x2 + y2 = 1}, plotted, say, in yellow.
We see him at (0,0, —1), plotted, say, in blue. We see her at (—1/2,0, —1),
plotted, say, in pink.
At time t = 0, the yellow, blue and pink are all plotted at (0,0,0).
At time t = 1, the yellow forms the circle {(x,y, 1) | x2 + y2 = 1}, the
blue is plotted at (0,0,1) and the pink is plotted at (1/2,0,1).
At time t = 2, the yellow forms the circle {(x,y,2) \x2 + y2 = 4}, the
blue is plotted at (0,0,2) and the pink is plotted at (1,0,2).
Et cetera.
The set of yellow points is called the "light cone". It is the cone
{(x,y,t) | x2 + y2 = t2}. The set of blue points is called his "lifeline". It is
the i-axis. The set of pink points, her lifeline, is the line {(t/2,0, i) 1i € E}.
Note that there is one particular particle of light which is chasing her
toward him. This is the particle with lifeline {(t,0,t) 1i 6 R}. Let us call
this particle LQ.
Changing perspective, let us try to imagine what his happening from
her point of view. She perceives herself at the center of the universe, so,
from this point of view, her lifeline should be the i-axis. According to 19th
century physics, from her point of view, he is traveling toward her with
velocity —1/2; she would say that his lifeline is {(—i/2,0,i) \t € E}.
What of the light cone? Let us focus for a moment on the light par-
ticle LQ that is chasing her toward him. According to 19th century ideas,
from her point of view, she should see L0 as moving a velocity of 1/2 toward
her. That is, from her point of view, the particle L0 should have lifeline
{(i/2,0,i) | i € E}. More generally, under a 19th century understanding of
physics, if he perceives an event which he plots at (x,y,t), then she will
plot it at (a; — (t/2),y,t); in particular, she would plot the light cone as
{(x-(t/2),y,t)\x2 + y2 = t2}.
This need to constantly switch between perspectives of various partici-
pants is highly invidious, and we propose to resolve all this by postulating
the existence of a single space-time M, on which all events occur, with no
particular coordinates attached to it. What he does is to choose a preferred
Lorentz manifolds and relativity 3
diffeomorphism $ : M ->• E 3 . To describe an event at a point m € M, he
would say that it occurred at $(m). Similarly, she prefers her own diffeo-
morphism (or "coordinates") * : M -> R 3 , and she would say that this
same event occurred at \t(m). Yet another observer might choose yet an-
other coordiatization of the "real" space-time M. These coordinate systems
(like $ and * ) are sometimes called "inertial reference frames". Changing
coordinates from his to hers is given by the map F := \t o $ - 1 :M3 ->M 3 ,
and, in the preceding paragraph, we asserted that, under pre-relativistic
mechanics, it would be reasonable to have F(x,y,t) = (x — (t/2),y,t).
Here is the catch: Experimental evidence indicates that this is simply
not correct. She does not perceive L0 as moving at speed 1/2. In fact,
amazingly, the empirical evidence is that she perceives all the particles of
light as moving at speed 1. Her plot of the light cone is exactly the same as
his; it is £ := {(x,y,t)\x2 + y2 = t2}. Then F(£) = £, i.e., F preserves £.
In particular, the formula F(x,y,t) — (x — (t/2),y,t) simply cannot be
true, so 19th century physics fails to describe accurately our universe.
In our modern understanding of the "overlap" map F, we know more
than just that F(C) — £: Let Q : I 3 -> E be the quadratic form defined
by Q(x,y,t) — x2 + y2 — t2, so that the light cone £ is the 0-level set
of Q, i.e., £ = <5 -1 (0). Then the basic axiom of special relativity is
that F preserves all the level sets of Q, not just the 0-level set £. That
is, Q o F = Q. In other words, F preserves Q. The collection of linear
transformations which preserve Q is called the orthogonal group of Q,
and is denoted by O(Q), so the basic axiom asserts that F € O(Q).
We now set this aside for a moment. The 18th and 19th centuries
saw the beginnings of what is now called Riemannian geometry. In its
simplest form, one imagines a smooth surface X C E 3 to which particles
are constrained to move. Under this constraint, minimizing travel distance
between two points p, q € X is a basic problem in the calculus of variations.
For any tangent vector v to X, let \v\ denote the length of v. For any path
7 : [0,1] -> X from p to q, define the length of 7 by £(7) = / 0 |-y(t) | <it.
The variational problem, then, is to minimize the functional L. For each
point 1 € I , we have a quadratic form gx : TXX -> E given by the formula
gx(v) = \v\2. Abbreviating gx as g for readability, we define the length of 7
to be £(7) := / 0 (fKiM)) 1 / 2 dt; we wish to find 7 from p to q minimizing
length among all competing paths. If the surface X is a flat plane in E 3 ,
then, as usual, the distance minimizing curve is a straight line. However, if
4 Introduction, History and Outline
the surface is curved, then the answer is likely to be more involved.
Note that we do not really require the ambient space R3 containing X.
All one needs is a Riemannian manifold, i.e., a manifold X together
with a system of positive definite quadratic forms {gx : TXX —• R}xex,
one on each tangent space to X. These forms should vary smoothly in x.
Again define L(-y) := J 0 (giJit)))1^2 dt, and we look to minimize L over
paths connecting a pair of points.
It is important to note that there is not a unique parametric path that
minimizes distance, even in a fiat plane P in R3. Given two points p,q £ P,
the path 70 : [0,1] —> P defined by 70(t) — p(l — t) + qt minimizes the
distance, but so does the path 71 : [0,1] -»• P defined by 71 (t) = 70(t 2 ). This
is because L is invariant under reparameterization. Given a Riemannian
manifold X and two points p, q € X, a better quantity to study is the
energy of a path 7 : [0,1] -¥ X defined by E(j) :— J0 g(j(t)) dt. It turns
out that energy minimizing paths are the same as constant speed length
minimizing paths. (See §12 of [Mi73].)
The positive definiteness of the quadratic forms ensures that the square
root in the definition of L makes sense. However, for manifolds carrying
other kinds of forms, one could still propose either to minimize E or to
minimize L over paths for which <?(7(£)) is always positive. In particular,
we may require not that {gm} is positive definite on each tangent space, but
only that it is nondegenerate on each tangent space. Under this weakened
assumption, the manifold X together with the system {gx}x€X is called a
pseudoRiemannian manifold. The family {gx}X£x of quadratic forms
is called a pseudoRiemannian metric. A natural question is why one
should study such objects.
To answer this, let us return to our space-time M. Recall the quadratic
form Q(x,y,t) = x2 + y2 — t2. Because R3 is a vector space, we may
identify each tangent space of R3 with R3 itself, and, in this way, the form
Q identifies with a nondegenerate quadratic form on each tangent space,
i.e., with a pseudoRiemannian metric g on R3. A pseudoRiemannian metric
coming from a quadratic form in this way is said to be flat.
The basic axiom of special relativity then ensures that F preserves g.
Under the diffeomorphism $ : M -¥ R 3 , g corresponds to a pseudoRieman-
nian structure j $ on M, Similarly, under the diffeomorphism V& : M -> R3,
g corresponds to a pseudoRiemannian structure j $ on M . Since F pre-
serves g, we see that g$ = g^. That is, while different observers will give
different coordinates to the same point in M, they all agree on a particular
Symmetries of Lorentz manifolds 5
pseudoRiemannian metric on M.
Let d > 2 be an integer and let V be a real d-dimensional vector space.
If Q : V -> R is a quadratic form, then we say that Q is Minkowski if
there is an isomorphism / : Rd —> V such that, for all (x\,..., xj) € E d , we
have Q(f(xi,...,Xd)) — x\ + • • • + x\_x — xd. If g is a pseudoRiemannian
metric on a manifold X, then we say that g is Lorentz if, for all a; € X,
the quadratic form gx on TXX is Minkowski.
With this terminology, special relativity, then, is the study of a flat
Lorentz manifold, i.e., a vector space on which a single Minkowski form
is used to create a pseudoRiemannian metric by identifying V with each of
its tangent spaces. In our example, we described a space-time with three
dimensions, but it is more typical, within special relativity, to analyze four-
dimensional space-times.
By contrast, general relativity considers a more general class of Lorentz
manifolds, those whose metric comes from a certain system of differential
equations. Once this metric is determined, the equations of motion of
particles are denned by minimizing energy, in analogy with Riemannian
geometry.
The collection of symmetries of a pseudoRiemannian manifold (i.e.,
the collection of self-diffeomorphisms preserving the metric) forms a group
called its isometry group. Typically, the kinds of Lorentz manifolds ap-
pearing in general relativity are four-dimensional and have compact isome-
try groups. Their isometry groups therefore act in an uncomplicated man-
ner (e.g., with closed orbits), from a dynamical point of view.
In this work, we set aside the motivations of special and general rel-
ativity, and focus on Lorentz manifolds with large, complicated isometry
groups, and in any dimension.
1.2 Symmetries of Lorentz manifolds
In any kind of dynamics of groups, it is basic to determine the collection of
groups that have "interesting" actions. As the definition of "interesting" is
varied, one obtains many different areas within dynamical systems. Work-
ing within a particular area, once the list of groups is complete, a second
(and usually harder) problem is to figure, for each one, the collection of
interesting actions that it admits. If this can be completed, then one can
reasonably claim to finished an area of dynamical systems.
6 Introduction, History and Outline
It has been known since early work of Cartan that the isometry group of
any "structure of finite type" (a class of structures which includes Lorentz
metrics) has a unique manifold structure that makes it into a Lie group. The
basic structure theory of Lie groups asserts that any connected Lie group
decomposes into two parts: its "semisimple" part and its "solvable" part.
(See the first two paragraphs of §4.9, p. 96.) So, in tackling a problem
in Lie theory, it is quite common to break the problem into semisimple
and solvable parts. Semisimple groups break up into simple groups. (See
Corollary 4.10.3, p. 113.) The best understood of the solvable groups are
the nilpotent groups. Thus one typically tries to do a nilpotent form of the
problem and a "simple" form (which, terminology notwithstanding, is often
very complicated). If successful, one is emboldened to study solvable and
semisimple problems. Finally, one may aim to solve the general problem.
The basic goal of much of this book is to determine the collection of Lie
groups admitting an interesting action on a Lorentz manifold. In different
places we define "interesting" in different ways. We sometimes focus on the
semisimple form of the problem, in which we wish to find all semisimple
Lie groups with an interesting action. In other places we focus on simple
groups, nilpotent groups and solvable groups.
What might one require for an action to be interesting? Since manifolds
are smooth objects, it is natural to require that the action be smooth. For
actions on Lorentz manifolds, another basic requirement is that the action
be by isometries, i.e., that each element act as an isometry of the Lorentz
manifold. We will say that an action of a Lie group on a Lorentz manifold
is isometric if it is smooth and by isometries. An isometric action of a
Lie group G on a Lorentz manifold M is smooth iff the action is given by
a smooth homomorphism from G to the isometry group Isom(M) of M.
Any group can act trivially on any manifold, and such an action is
probably not very interesting. By definition, the kernel of an action is
the collection of all group elements which fix all points of the space; it is a
normal subgroup. So, to be interesting, one might require that an action be
faithful, which is to say that its kernel is trivial. Any non-faithful action
descends to a faithful action of the group mod the kernel, and so the original
action is simply lifted from a faithful action.
A slightly less stringent requirement is that the action be locally faith-
ful, which is to say that its kernel is discrete. If a Lie group G acts locally
faithfully on a set M and if G ->• G is a covering homomorphism, then
the lifted action of G on M is again locally faithful. So local faithfulness
Outline of succeeding chapters 7
is preserved on passing to covers; faithfulness is not. The Lie algebra of a
Lie group is the same as the Lie algebra of any of its covering groups, so,
from a Lie theoretic point of view, it is not uncommon to think of a group
as being essentially unchanged by passing to a covering. Extending this
philosophy, one might want to lift actions to coverings without considering
the lifted action to be essentially different from the original action. From
this point of view, local faithfulness is a reasonable requirement.
The basic requirements that an action be isometric and locally faithful
appear nearly everywhere in this book. However, every Lie group acts faith-
fully on itself by left-translation, preserving any left-invariant metric. This
action has only one orbit, so the orbit structure is clearly as uncomplicated
as can be. In the remainder of the book, we try out various additional
dynamical requirements on our actions. The general philosophy of much of
what follows is that, while any Lie group can act faithfully and isometrically
on a Lorentz manifold, if one imposes even small additional requirements,
then the list of groups is dramatically reduced.
1.3 Outline of succeeding chapters
In this book, we assume, on the part of the reader, a good basic under-
standing of measure theory, topology, and differential topology. A typical
first-year graduate course in these subjects will cover the needed material.
We require basic facts from Riemannian geometry and Lie theory, but we
develop most of the needed definitions, and we liberally cite results from
standard texts so that a student who is not versed in these areas should find
the arguments complete. We should point out that, while we have made
an effort to prove or give careful references for all of the results herein,
it is nevertheless true that we rely on a significant level of mathematical
maturity, especially in Chapter 15.
In Chapter 2, we collect together the miscellany of elementary definitions
and results that we will be using. Chapter 3 concerns basic results in
differential topology, i.e., basic facts about smooth manifolds. In Chapters
4 and 5, we set up the facts from Lie Theory that we will need. In Chapter
4, one can find an introduction to many basic topics in the theory of real
Lie groups, although, to keep the exposition within reasonable bounds,
we do not start from scratch, and one will find a number of references to
basic texts. We sometimes reprove some well-know facts if we feel that the
8 Introduction, History and Outline
exposition in the literature can readily be improved, or is not available. For
the most part, Chapter 4 develops results that are generally of interest in
Lie Theory, whereas Chapter 5 collects results that are of special interest
in this book.
Basic information on the structure of Minkowski vector spaces and their
isometry groups can be found in Chapter 6. Chapters 2, 3, 4 and 6, as well
as Appendices A and B, contain material that is well-known. Some of the
material in Chapter 5 may also be described as well-known.
In Chapter 8, we give a collection of examples of actions of Lie groups
on compact Lorentz manifolds. Eventually, in Chpaters 11 and 12, we show
that the collection of Lie groups analyzed in Chapter 8 is, in some sense,
exhaustive. In Chapter 13, following M. Gromov and A. Zeghib, we show
that, for some of these groups, we can describe their possible actions on
compact Lorentz manifolds.
In Chapter 9, we consider noncompact Lorentz manifolds. We give ex-
amples of groups admitting actions on such manifolds that exhibit some
level of complicated orbit structure. In Chapters 10, 14 and 15, we show
that the list of groups analyzed in Chapter 9 is, again, exhaustive. Chap-
ter 10 focuses on semisimple groups acting on Lorentz manifolds, and de-
scribes work originally due to N. Kowalsky. Chapter 14 focuses on free
actions and actions with compact stabilizers. Chapter 15 studies general
isometric actions of general simply connected Lie groups.
The problem of proving completeness of the various lists of groups ob-
tained requires some new basic dynamical techniques, which are found
in Chapter 7, which may be thought of as a kind of "toolbox" to be used
in proving the results of Chapters 10-15.
Appendix A contains an elementary proof of the Borel Density Theorem.
Appendix B contains an elementary proof that algebraic actions are smooth.
1.4 Notation
There is a significant amount of notation in this book, but, to aid the reader,
we have tried to reference it in the index, under "ZZZ Notation".
Some notation is introduced and used repeatedly in many different chap-
ters, and the reader would be well-advised to become familiar with it as
quickly as possible. This includes Qd, </>B, I/>B, ^>{d\ ft(d), N(d]', V{d\
d (d m m 1 1
Nl \ E\f, e \ XB, Xg , X% , OMV, OUT *, MV, AfV *, vt -» v, Sm
Acknowledgements 9
and g|h. All these can be found under "ZZZ Notation" in the index.
Throughout this book, if a Lie group is denoted by a capital Latin letter
(perhaps with some adornments like asterisk and/or subscript), then the
Lie algebra of that Lie group will always be denoted by the corresponding
small fraktur character (with the same adornments). So, for example, if
H£ is some Lie group, then its Lie algebra is denoted \j%.
1.5 Acknowledgements
This book was written with the support of the National Science Foundation
(DMS-9703480 and DMS-0072165). I also acknowledge with gratitude the
support of my department, which generously gave me a paid one-semester
leave to provide time for the writing of this book. D. Witte has answered
innumerable questions over the years, and, without this help, the results
appearing here might be far less sharp. I also thank my wife, IOJIHH,
who provided constant support and who tolerated, with good humor, my
occasional weekend absences to work on this manuscript.
Chapter 2
Basic Results and Definitions
Throughout this text, Z denotes the set of integers, Q denotes the set of
rational numbers E denotes the set of real numbers and C denotes the set
of complex numbers. Let M+ := (0, oo) be the set of positive real numbers.
If S and T are sets, then S\T := {s e S\s $T}.
A Borel space is a set with a a-algebra. The sets in the u-algebra are
referred to as Borel sets. A Borel measure on a Borel space (X, B) is a
countably additive, <r-finite function / i : B —> [0, oo] such that n(X) ^ 0. If
X and Y are Borel spaces, then a map / : X ->• Y is Borel if, for all Borel
subsets B C Y, we have that f~x(B) is a Borel subset of X.
If X is a topological space, then, unless otherwise specified, X will be
given the smallest cr-algebra containing its topology, so that X is also a
Borel space.
2.1 Some set-theoretic notions
Let X and Y be sets and let / : X ->• Y be a function. A subset A C I i s
called a fiber of / if, for some y € Y, we have A = f~x (y). A subset AC. X
will be said to be /-saturated if, for some B C Y, we have f~1(B) = A.
If / : X -4 Y is surjective, then a cross-section to / is a map g : Y -> X
such that, for all y £ Y, we have f(g(y)) — y-
L E M M A 2.1.1 Let X and Y be sets and let f : X ->Y be a function. Let
A,B C X. Assume that A is f-saturated. Thenf(AnB) = (f (A))n{f (B)).
11
12 Basic Results and Definitions
Proof. We have f(AC\B) C f(A) and f(Ar\B)Cf(B), so it suffices
to show that (f(A)) n (/(B)) C / ( 4 n B). Choose ,4' C y such that
f~1(A') = A. Let ye (f(A))n(f(B)). We wish to show that y £ f(AnB).
Choose b £ B such that f(b) = ?/. We have
f(b) = yef(A) = f(f-1(A'))CA',
so b £ / _ 1 ( ^ ' ) = ^- So
> since
& £ B , we get 6 6 A n B. It follows that
y = f(b)ef(AnB). a
Let X be a set, let v, w £ X and let y be a, collection of subsets of X .
For any integer n > 1, we define a 3^-chain of length n from v to w to be
an (n + l)-tuple (XQ, • • •, xn) £ Xn+1 such that xo — v, such that xn = w
and such that for all integers k £ [l,n], there exists 7 £ ^ such that both
:Efc_i € y and Xk € y . We say that v and u; are 3^-chain connected if,
for some integer n > 1, there exists a ^-chain of length n from u t o w .
Let X be a set and let y be a collection of subsets of X. We say that
X is [y-connected if any two points of X are y-chain connected.
Let R be an equivalence relation on a set X. For all x € X, we let
R(x) := {y e X \ (x,y) £ R} denote the equivalence class of x. For all
subsets X0 C X, we define R(X0) := [J R(x).
x£X0
Let X be a set and let S C X x X. Then the equivalence relation
on X generated by S is the smallest equivalence relation on X contain-
ing S. Let 5* := {(y,x) | (z,2/) € S} and let £> := {{x,x)\x £ X}. If R is
the equivalence relation generated by S, then (a;, y) £ R iff there exists an
integer n > 1 and there exist zo, • • •, zn £ X such that
• zo = x;
• zn = y; and
• for all integers i £ [1,n], we have (ZJ_I,Zi) £ SU Sl ISD.
Let X be a set and let S C X x X. For all a; € X, let
S(x) :={y£X\ (x, y) £ S}, S\x) := {y £ X \ (y, x) £ 5}
and E(x) := (S(x)) U (Sl(x)) U {x}. Let £ := {£(a;) | a; £ X } . Let i? be the
equivalence relation on X generated by S. Then, for all x,y £ X, we have:
(a;, y) £ R iff a; and y are £-chain-connected.
Let R and 5 be two equivalence relations on a set X. Then we say that
R has countable index in S if every equivalence class of 5 is a countable
Some set-theoretic notions 13
union of equivalence classes of R. In particular, if R has countable index
in S, then Jf? is a sub-equivalence relation of S, i.e., R C S.
L E M M A 2.1.2 Let X be a set and let Xi be a sequence of subsets of X.
Assume thatX = ( J Xt. For all i, let Ri be an equivalence relation on Xi.
i
Assume, for all i,j, for all x £ Xi, that there is a countable subset C C Xj
such that (Ri(x)) D Xj C Rj(C). Let R be the equivalence relation on X
generated by\\ Ri. Then, for all i, Ri has countable index in R\Xi.
i
Proof. For all i, let Si be the set of equivalence classes of Ri and let
Si := £i x {i}. Let S := [j Si. For all 5 = (E,i) £ S, for all j , let
Aj(S) := {F £ £j | EHF ^ 0} C £j; then A'j{S) is countable. For all S £ S,
for all j , let Aj(S) := ( ^ ( 5 ) ) x {j} C Sj C S; then A,(S) is countable.
For all S e S, let A(S) := ( J Aj(S) C 5; then A(S) is countable. For all
3
So C <S, let yl*(<So) := M ^4(5). Then, for all countable So C <S, we have
se5 0
that .4*(<So) is countable. From the above definitions, we have: for all i,
for all T = (F,j) 6 <S, for all y 6 -F n X i ; we have {Ri(y),i) G «4(T).
For all S € 5 , let Ci(S) := ^1(5); then Ci(5) is countable. We make
the following recursive definition: For all S £ S, for all integers n > 1,
C n +i(5) = ,4*(C n (S)). By induction, we see, for all S £ S, for all integers
oo
n > 1, that Cn{S) is countable. For all S eS, let C(5) := ( J Cn(S); then
n=l
C(5) is countable.
Fix i and fix a; € Xj. We wish to show that there is a countable subset
BCSi such that (R\Xi)(x) = \J B. Let B := {B e £t\B C (R\Xi)(x)}.
Then, as Ri C R\Xi, we conclude that (R\Xi)(x) = M B. We wish to
BeB
show that B is countable.
Let E := Ri(x) £ £u S := (E,i) £ Si and V := {D ££i\(D,i) eC(S)}.
Then C(S) is countable, so T> is countable, so it suffices to show that B C T>.
Fix B £ B. We wish to show that B £ V.
We have B £ B C £{, so B ^ 0. Fix y 6 5 . Then y S S € £",, so
B = Ri(y). As B £ B, we get B C (i?|Xi)(x). Then y G (R\Xi)(x) C i?(a;).
14 Basic Results and Definitions
Let £ := \^J £j. Since R is generated by M Rj and since y £ R(x), it
i j
follows that x and y are f-chain connected. Choose an integer m > 1 and
an £ -chain (zo,...,zm) of length m from x to y.
For all integers fc £ [l,m], choose Ek £ £ such that Zk-i,Zk £ i?*.
For all integers A; € [l,m], choose jk such that .E^ 6 £jk and then define
Sk •= (Ek,jk) e Sjh. We have x 6 £ n ^ , so E0 £ >IJ 0 (5), so 5 0 £ -4j 0 (5).
For all integers k £ [l,m], we have Zk-i £ £ * - i nEfc, so Ek £ -/^(Sfc-i),
so Sk £ Ajh (Sk-i). Let n := m + 1. Then 5 m € C„(5).
Let j := j m , F := Em and T := (F, j) = Sm. Then y = zm £ £Jm = F.
Moreover, T £ C„(5). so A(T) C ^*(C„(5)). As y £ B n F , we see that
B e A\(T), so (B,t) € A ( T ) C A{T). Then
(B,i) 6 A{T) C ^*(C n (5)) = C„+i(S) C C(5).
So, by definition of V, we conclude that B £ V. •
2.2 Some group-theoretic notions
In any group G, the identity element of G will be denoted \Q and the center
of G will be denoted by Z(G). A group G is centerfree if Z(G) = {1G}-
For any group G, for any H,K C G, we denote by [H, K] to be the
subgroup of G generated by {/ifc/i_1fc_1 \h £ H,k £ K}.
Let G be a group. If S C G, then we define
C G (S) := {<? G G | Vs £ S, gs = sg}, NG(S) := {g £ G \ gSg'1 = S}.
For all g £ G, we define CG(g) := CG({g}) •
Let G be a group and let S, T C G. We say that 5 normalizes T if,
for all s £ S, we have sT = Ts. We say that 5 and T centralize each
other if, for all s £ S, for all* £ T, we have st — ts. Alternatively, in this
situation, we will say that S centralizes T or that T centralizes S.
If G is a group, then Aut(G) denotes the automorphism group of G.
If G is a group and N is a normal subgroup of G, then we define a
homomorphism Intjy : G -¥ Aut(iV) by (IntAr(<7))n = gng~l.
Unless otherwise specified, by "action" we always mean left action. If
a group G acts on a set X, then the set of orbits of this action is denoted
G\X. For a right action, the set of orbits is denoted X/G.
Some topological notions 15
Let a group G act on a set X . For all x € X,
StabGO*;) '•— {g & G \ gx = x}
denotes the stabilizer in G of x. The G-action on X is free if, for all
a; € A", we have Stabc(a;) = {1G}- The kernel of the G-action on X is
| ) StabG(z). The action is faithful if its kernel is { 1 G } .
xex
If a group G acts on two sets X and Y and if / : X -> Y is a function,
then we say that / is G-equivariant if, for all g € G, for all a; € X , we
have /(gz) = g(f(x)).
2.3 Some topological notions
Following [SS78], we adopt the point-set topological convention that normal
implies regular implies Hausdorff. We will use lcsc as an abbreviation for
the phrase "locally compact, second countable and Hausdorff". Note that,
with our conventions, "lcsc" implies Hausdorff.
Let S be a subset of a topological space X . We say that S is p r e c o m -
pact in X if the closure in X of S is compact. We leave it as an exercise
for the reader to prove: S is closed in X iff, for all a; € X , there exists an
open neighborhood N of x in X such that S n N is closed in the subspace
N of X. We say that 5 is locally closed in X if, for all s € S, there
exists an open neighborhood N of s in X such that 5 n iV is closed in the
subspace N of X. Note that S is locally closed in X iff there exist subsets
U, C C X such that U is open in X , such that G is closed in X and such
that 5 = U n G.
Let X and Y be topological spaces and let / : X —» Y be a function.
We say that / : X —¥ Y is open if, for any open subset U of X , we have
that f(U) is open in Y. We say that / : X —> Y is closed if, for any closed
subset G of X , we have that / ( G ) is closed in Y.
If xi is a sequence in an lcsc topological space X, then "xi —> oo in X"
means "xi leaves compact subsets of X " . That is, x% -*• oo in X iff for all
compact sets X C X, the set {i | X{ € X } is finite.
Let X and Y be lcsc topological spaces and let / : X —> Y be continuous.
Then / is said to be proper if, for every compact K C Y, we have that
f~x(K) is compact. Equivalently, / : X —• Y is closed and every fiber of /
is compact. A map is nonproper if it is not proper. The map / : X —> Y
16 Basic Results and Definitions
is nonproper iff there is a sequence xt in X such that x, -» oo in X, but
such that f(xt) is convergent in Y.
A topological space is Polish if it is separable and admits a complete
metric compatible with its topology. By basic arguments in point-set topol-
ogy, any lcsc topological space is Polish (see, e.g., i) =J> iv) of Theorem
5.3, p. 29 of [Kec95]). A topological space X is called a Baire space if, for
any sequence Xi, X2,... of closed subsets of X, if X = Xi U X2 U • • •, then
there exists an nonempty open subset U of X and there exists an integer
i > 1 such that U C Xi. In other words, X cannot be written as a count-
able union of closed sets each of which has empty interior in X. The Baire
category theorem asserts that any locally compact Hausdorff topological
space is Baire and that any Polish space is Baire.
A topological space is discrete if all of its subsets are open. A topo-
logical space X is zero-dimensional if, for every x € X, for every open
neighborhood U of a; in X, there is an open neighborhood V of x in X
such that V is closed in X and such that V C U. A topological space X
is totally disconnected if, for any subspace S of X, if S is connected,
then S has at most one element. A topological space X is totally path-
disconnected if every path in X is constant, i.e., if, for every continuous
7 : M. -> X, there exists x £ X such that j(R) = {x}. A discrete topo-
logical space is zero-dimensional. A zero-dimensional topological space is
totally disconnected. A totally disconnected topological space is totally
path-disconnected. There is a connected, totally path-disconnected sub-
space of the Euclidean plane E 2 (see the space denoted " F " under Ex-
ample 128 on p. 145 of [SS78]). A subset of a topological space is dis-
crete (resp. zero-dimensional, resp. totally disconnected, resp. to-
tally path-disconnected) if, when endowed with its inherited topology,
it is discrete (resp. zero-dimensional, resp. totally disconnected, resp. to-
tally path-disconnected). Note that, under this definition, {1/n | n > 1} is
a discrete subset of E, so a discrete subset may not be closed. However, a
discrete subset of a T\ topological space is always locally closed.
By Example 75 on p. 93 of [SS78], there is an infinite countable Haus-
dorff topological space which is connected. However, in non-pathological
situations, countable topological spaces are highly disconnected. For ex-
ample, since E cannot be written as a countable union of proper closed
subsets, it follows that any countable Ti topological space is totally path-
disconnected. Thus, any countable subset of a 7\ topological space is totally
path-disconnected. Moreover, by Lemma 2.3.1, p. 17 below, any countable
Some topological notions 17
regular topological space is zero-dimensional. Consequently, any countable
subset of a regular topological space is zero-dimensional.
L E M M A 2.3.1 Any countable regular topological space is metrizable and
zero-dimensional.
Proof. Let X be a regular topological space. Assume that X is countable
as a set. We wish to show that X is metrizable and zero-dimensional.
Since X is Lindelof, by Lemma 1 of Chapter 4 on p. 113 of [Kel55],
X is normal. Then, by Theorem 16 of Chapter 4 on p. 125 of [Kel55], X is
metrizable. Let x 6 X and let Y be a closed subset of X. Assume that
x £ Y. We wish to show that there is a subset C of X which is both open
and closed in X such that x € C C X\Y.
By Lemma 4 of Chapter 4 on p. 115 of [Kel55], choose a continuous
function / : X ->• E such that f(X) C [0,1], such that f(x) = 0 and
such that f(Y) = {1}. Since X is countable, f(X) is countable. Choose
t e (0,1) such that t £ f(X). Let C := /"^([O.i]). Then C is closed in X,
x e C and C D Y = 0. Moreover, as C = / _ 1 ( ( - 1 , * ) ) , we see that C is
open in X. •
L E M M A 2.3.2 Let X be a topological space. Let y be a collection of
subsets of X. Assume, for all Y £ y, that Y is connected in the relative
topology inherited from X. Assume that X is y-connected. Then X is
connected.
Proof. Let V, W be nonempty open sets in X such that V f~l W = 0 and
such that V U W = X. We aim for a contradiction.
Choose v € V and w 6 W. Choose an integer n > 1 and choose an
(n + l)-tuple (z0,..., zn) £ Xn+1 such that ZQ = v, such that zn = w and
such that for all integers k € [l,n], there exists Y 6 y such that both
Zk-i 6 y and Zk € V. Choose an integer i G [l,n] such that Zj_i € V
and such that ^ € W. Choose Y ey such that Zi_i,Zj G V. Give F the
relative topology inherited from X. Then Y is connected. Let A :=V C\Y
and let £? := W n V. Then A and B are open in Y. Moreover, A D B = 0
and A U B = 7 . Finally, as Zj_i € A and Zi £ J3, we see that A ^% ^ B.
Then F is not connected, contradiction. •
The next result is a partial converse to Lemma 2.3.2, p. 17.
L E M M A 2.3.3 Let X be a connected topological space and let U be an
open cover of X. Then X is U-connected.
18 Basic Results and Definitions
Proof. For all x,y E X, write x ~ y if x and y are W-chain connected.
Then ~ is an equivalence relation on X. For all x 6 X, let Vx be the
equivalence class of x, so T4 = {«/ £ X | x ~ y}. We wish to show, for all
a; € X, that K = X.
For all a; 6 X, by definition of ~ , we see that there is a subset Ux C U
such that Vx = M [/. In particular, for all x 6 X, we see that Vx is open.
J7€M»
Since {Vx | a; 6 X } is a partition of the connected topological space X
by nonempty open sets, we conclude, for all x 6 X, that Vx — X. D
2.4 Some notions from linear algebra
By vector space, we will always mean a finite-dimensional real vector
space, unless otherwise specified. By complex vector space we will
always mean a finite-dimensional complex vector space, unless otherwise
specified. If V is a vector space, then V* denotes the dual vector space
consisting of all linear transformations V -» R.
For any real vector space V, by Vc we denote the complex vector space
C ® R V; yc is the complexification of V. For any complex vector space V,
by VR we denote the underlying real vector space V, after forgetting the
complex structure; VR is the reallocation of V. By construction, for any
complex vector space V, as sets, we have VR = V.
Let V be a complex vector space and let J : VR —> VR be defined by
J{v) — \f^\v. Let V be the complex vector space whose underlying real
vector space is the same as that of V, but on which multiplication by y/— 1 is
defined by \/—Tu = — (J(v)). Then scalar multiplication in V is defined by
(a + &\/—l)v = av — b(J(v)). The complex vector space V is the conjugate
of V. We define VR := (V) R ; by construction, as sets, we have VR = V.
For any vector space V, for all v, w € V, we define
v±iw := (1 ® v) ± {y/^l <g> w) € VC.
For any vector space V, the map
(x, y) H> x + iy : V ©V ->• (V C ) R
is an isomorphism of real vector spaces, so V©V = ( V C ) R . For any complex
vector space V, the identity map x H+ X : VR -> VR is an isomorphism of
Some notions from linear algebra 19
real vector spaces, so VR = VR. For any complex vector space V, the map
(x,y) ^ (x + y/=ly) - i(y + y/=lx) : V©V -> ( V R ) C
is an isomorphism of complex vector spaces, so V © V = (VR)C. For any
vector space V, the map
v + iw M- v — iw : Vc —>• Vc
is an isomorphism of complex vector spaces, so Vc = Vc. All of the
isomorphisms described in this paragraph are natural.
If Vi is a sequence in a vector space V and if v € V, then we will say
that Vi converges in direction to v and write Vi —>• v if v ^ 0 and there
exists a sequence ti in E such that
• U -> 0 in R; and
• tiVi —> v in V.
Let V be a vector space and let T; be a sequence in End(V). An element
t) e V i s said to be Kowalsky for T; if there is a sequence Vi in V such
that Vi —> 0 in V and such that Ti(vi) —• v in V. We say that a subset
S C V i s Kowalsky for T; if all of its elements are Kowalsky for T;. Note
that, if we denote by VQ the set of all v e V such that v is Kowalsky for Ti,
then VQ is a vector subspace of V.
For all integers m, n > 1, let Wnxn denote the set of all m x n matrices
with real entries. For all integers d > 1, for all integers i,j £ [l,d], let i5>.'
denote the d x d matrix with a one in the (i, j) entry and zeroes elsewhere.
For all integers d > 1, let e[',..., ey denote the standard basis of E d x l .
Let V be a vector space, let d := dim(V) and let B be an ordered basis
of V. Let ipB : E d ~> V be the vector space isomorphism which carries the
standard ordered basis of Rd to B. Let 4>B '• R d x l -> V denote the vector
space isomorphism which carries the standard ordered basis of E d x l to B.
For any v € V, we denote (p^iv) by [V]B-
Let V and V be vector spaces and let T : V -> V be a linear transfor-
mation. Let d := dim(V) and let d' := dim(V). Let B be an ordered basis
of V and let £ ' be an ordered basis of V. Then [T]Jf G E d ' x d is defined
by: for all v € E d x l , we have ( ^ o T o <j>B)(v) = [T\$v. We call [T]f the
matrix of T with respect to $ and B' .
Let V be a vector space and let T : V -t V be a linear transformation.
Let B be an ordered basis of V. Then the matrix of T with respect
to B, written [T] B , is defined by [T]B := [T]%.
20 Basic Results and Definitions
If V is a complex vector space and if T : V -» V is a complex linear
map, then we say that T is complex diagonalizable if there is an ordered
basis of V with respect to which the matrix of T is diagonal.
If V is a vector space, then a linear map T : V —> V is real diagonaliz-
able if there is an ordered basis of V with respect to which the matrix of T is
diagonal. The complexification of T is denoted T c : Vc -» Vc. We say that
T is complex diagonalizble if Tc : Vc -»• Vc is complex diagonalizable.
We say that T : V -» V is elliptic if {exp(tX)} t6 R is precompact in GL(g).
A linear transformation T is elliptic iff both T is complex diagonalizable
and all characteristic roots of T are pure imaginary.
If T : V -> V is an endomorphism of a vector space V, then there exists
a unique triple D, E, N : V —> V of endomorphisms of V satisfying:
• D : V —> V is real diagonalizable;
• E : V -¥ V is elliptic;
• N : V -> V is nilpotent;
• [D, E] = [D, N] = [E, N] = 0; and
• T = D + E + N.
The equation T = D + E + N is then said to be a real Jordan decom-
position on V. We call D the real diagonalizable part of T. We call
E the elliptic part of T. We call N the nilpotent part of T.
Let T : V —>• T^ be an endomorphism of a vector space V. Choose
D,E,N e End(V) such that T = £> + .E + J V i s a real Jordan decompo-
sition on V. Let S C End(F) be the M-linear span of T,T2,T3,.... Then
D,E,N £S.
Let V be a vector space. A subset <S of End(V) is said to be real Jordan
closed if, for every T € S, for every D,E,N € End(V), i f T = £> + jE;-|-A/'
is a real Jordan decomposition on V, then D,E,N £ S.
Let F b e a real vector space and let / : V -+ V be the identity map. A
complex structure on V is a linear map J :V —> V such that J 2 = —I.
If W is a vector subspace of a vector space V, then we define
codim v (W0 := dim(V) - dim(W).
L E M M A 2.4.1 Let V and V be vector spaces and let f,D,E,N :V -> V
and f',D',E',N' : V —> V be linear transformations. Assume that both
f — D + E + N and f = D' + E' + N' are real Jordan decompositions. Let
T : V —> V be a linear transformation. Assume that r o / = / ' o r . Then
1
T o D = D o T and r o E = E' o r and r o N = N' o r .
Some notions from linear algebra 21
The preceding lemma is basic and is left as an exercise for the interested
reader, with the following hint: If r is either surjective or injective, then
Lemma 2.4.1, p. 20 can be argued from uniqueness of the real Jordan de-
composition. A general r may be written as the composition of a surjective
map followed by an injective map, so the general case can be seen to follow
from the surjective and injective cases.
L E M M A 2.4.2 Let V be a vector space and let S : V ->• V be complex
diagonalizable. Assume that every characteristic root of S : V —> V is
contained in E U v^TK. For all A 6 M, let V\ := {v G V \ S(v) = Xv}. Let
VD '•= / J ^A- Let T : V —> V be the real diagonalizable part of S. Let
AeR\{0}
VE be an S-invariant vector subspace ofV. Assume that VD + VE = V and
that VD l~l VE = {0}. Then T\VD = S\VD and T\VE = 0.
Proof. We have S(VD) C VD and S(VE) C VB. The transformation
S\VE : VE —> VE is complex diagonalizable and all of its characteristic
roots are contained in y/^lR. Then S\VE '• VE —> VE is elliptic. The
transformation S\VD -VD~>VD is real diagonalizable. Define D : V -> V
by D\VD = S\VD and D\VE = 0. Define E : V -¥ V by E\VD = 0 and
E\VE = S\VE- Then S = D + E + 0isa, real Jordan decomposition on V.
Then, since T is the real diagonalizable part of S, we get T = D. Then
T\VD = D\VD = S\VD and T\VE = D\VE = 0. •
The next result appears as Lemma 4.4 in [A99b].
L E M M A 2.4.3 Let k>2 be an integer and let
V = V0 2 Vi 2 V2 D • • • D Vk = {0}
be a chain of vector subspaces of V. Assume, for all i £ { 0 , . . . , k}, that
T(Vi) C Vi. Let V. := (F 0 /Vi) © • • • © (Vi-i/Vi). Let T„ : K -*• V, 6e tte
mop induced 6y T : V -^ V. Then dim(T,(K)) < dim(T(V)).
Proof. It suffices to prove this in the case where k = 2. Let V := V/Vi
and let 7r : V ->• V be the canonical map. Identifying V\ with Vi/{0}, we
have V, =V@Vi. Then d = dim(V") = dim(V„).
Let K denote the kernel of T : V —> V and let L denote the kernel
of T. : V, -> V*. Let 7? := 7r(if). Let d K := dim(JiT), let dL := dim(L)
and let n := (dim(ii')) + (dim(.K" n Vi)). The surjective homomorphism
22 Basic Results and Definitions
•K\K : K -> K~ has kernel K C\Vi,so dK = n. We have ~K © (if n Vi) C I,,
so n < d i . Then da = n < d^.
Then dim(T,(K)) = d-dL<d-dK= dim(T(V)). D
The following two results are Lemma 6.4 and Corollary 6.5 on pp. 466-
467 of [AOOa].
L E M M A 2.4.4 Let V be a finite dimensional vector space and let Bi be a
sequence in GL(F). Let Y G End(V) be Kowalsky for the sequence of maps
A i-> BiAB'1 : End(V) -> End(V). Then Y : V -» V is nilpotent.
Proof. Let Yi be a sequence in End(V) such that Y* —> 0 in End(V) and
such that BiYiB^1 -> Y. Let d := dim(V). We may assume that V = Md.
Let K := SO(d) and let A+ := { o i ^ j f + • • • + adE$ | oi > • • • > a d > 0}.
Since we have GL(V) = KA+K, we may assume, for all i, that Bi G A+.
Then V is strictly upper triangular, and, in particular, is nilpotent. D
2.5 Matrix concentration lemmas
Let d > 1 be an integer and let J := { 1 , . . . , d}. For all j G J, let ej := e), '.
For all j , A; € J, let Ejk := E<.f. Let e j , . . . , e ^ be the basis of (M d x l )* dual
to e i , . . . , e d , so that, for all j,Z € J, we have e*(ej) = fy. Let {-Ej*m}j,me.J
be the basis of (E d x d )* dual to {.Ejjjj.kej, so that, for all j,k,l,m G J,
we have Efm{Ejk) = SjiSkm. For all E C J, a matrix u € Rdxl is said to
be c o n c e n t r a t e d on E if, for all j G <7\E, we have e*Av) = 0. For all
II C J x J, a matrix M G K d x d is said to be c o n c e n t r a t e d on II if, for
all (j, k) G (J x J ) \ n , we have £<-A(Af) = 0.
L E M M A 2.5.1 Let D := EEu H 1- REdd be the set of diagonal dxd
real matrices. Let Ai be a sequence in D. Let v G Rdxl be Kowalsky
for Rdxl -> W*xl. Let E denote the set of j G J such that
{Ejj(Ai)}i is unbounded in R. Then v is concentrated on E.
Proof. Choose a sequence vt in R d x l such that Vi -> 0 and such that
AiVi -»• v in M d x l . Then, by matrix multiplication, for all j G J \ E , we
have e^AiUi) ->• 0, so e*^(v) = 0 . D
L E M M A 2.5.2 Let D* ~ {aiEn + ••• + adEdd \ au... ,ad G K\{0}} be
the set of invertible diagonal dxd real matrices. Let Ai be a sequence
First results on expansive sequences 23
in D*. Let M £ K d x l be Kowalsky for X h+ AiXA'1 : Rdxl ->• R d x l .
Lei LT denote the set of (j,k) £ J x J such that {(Ejj(Ai))/(E%k(Ai))}i is
unbounded in R. Then M is concentrated on II.
Proof. Choose a sequence M» in ffidxl such that Mt ->• 0 and such that
AiMiA^1 -> M in R d x l . Then, by matrix multiplication, we see, for all
0', k)e(Jx J ) \ n , that E^AiMiAr1) -> 0, so £* fc (M) = 0. D
2.6 First results on expansive sequences
Let W be a vector space and let V» be a sequence of vector subspaces of W.
Let Vi be a sequence in W. Then we say that v» is a sequence in V; or
that i>j is in V{ if, for all i, we have t>j € V;.
Let W be a vector space and let Tj be a sequence in End(W). Let Vi be
a sequence of vector subspaces of W. Then we say that Ti is expansive
on Vi if, for every sequence Vi in Vi, we have: if {Ti(vi)}i is precompact
in W, then u, -» 0 in W. Note that we do not assume, for any i, that
Ti{Vi) C Vj. Note that, if W is a vector subspace of W, if V C W and
if, for all i, we have Ti(W) C W , then Tj is expansive on V* iff Ti\W is
expansive on Vi. In particular, if G is a Lie group, if H is a Lie subgroup of G
if 5i is a sequence in H and if Vj is a sequence of vector subspaces of h, then
Adg(ffj) is expansive on V, iff Ad[,((;t) is expansive on V*. Consequently, we
may simply say that Adgi is expansive on Vi, without ambiguity. Similarly,
if G is a Lie group, if N is a normal Lie subgroup of G, if Vj is a sequence
of vector subspaces of g/n and if gt is a sequence in G, then we say that
Adgt is expansive on Vi if Adg/n(gi) is expansive on Vi and, again, there is
no worry of ambiguity.
The connection between expansiveness and Kowalsky subspaces is:
L E M M A 2.6.1 Let W be a vector space. LetTi be a sequence in End(TV).
Let Ui be a sequence of vector subspaces ofW. Let V be a vector subspace
ofW. Let k := dim(V). Assume, for all i, that dim(T;({/;)) = k. Assume
that Ti is expansive on Ui- Assume that Ti(Ui) —> V in the topological space
of k-dimensional vector subspaces of W. Then V is Kowalsky for T{.
Proof. Let v £ V. We wish to show that v is Kowalsky for Tj.
Choose a sequence u^ in Ui such that Ti{ui) —>• v. Because Ti is expan-
sive on U, we see that Ui -> 0 in U. Thus v is Kowalsky for Tj. •
24 Basic Results and Definitions
L E M M A 2.6.2 Let W be a vector space and let Ti be a sequence in End(W)
Let Xi and yi be precompact sequences in W. Let p,q 6 W. Assume that
both Ti{xi) —^ p and Ti(yi) —*• q in W. Assume that p and q are linearly
independent in W. Then Tt is expansive onffircj+ Ryt.
Proof. Let Vi be a sequence in Vi :— Mxi + Ryi and assume, for a con-
tradiction, that {Ti(vi)} is precompact in W, but that Vi •/> 0 in W.
Passing to a subsequence, we may assume that Vi is bounded away
from 0 in W. Since p and q are linearly independent, passing to a tail, we
may assume, for all i, that Xi and yt are linearly independent. For all i,
choose (o,i,bi) £ E 2 such that Vi = diXi + biyi. Since Vi is bounded away
from 0 in W and since {xi}i and {?/;}« are both precompact in W, it follows
that (ai,bi) is bounded away from (0,0) in E 2 .
For all i, let p\ := Ti(xt) and let q\ :=Ti(yi). Choose sequences s; and ti
in E such that Sj —• 0 and ti -» 0 in E and such that Sip[ —> p and Uq[^ q
in W. Since p ^ 0 ^ q, by passing to a tail, we may assume, for all i, that
Si ^ 0 7^ t». For all i, we have Ti(vi) = aip[ + biq\. For all i, let p; := Sip\,
let gi := Uql, let a* := aj/s, and let ft := &*/*»; then Ti(vt) = aiPi + fcqi.
Consequently, {ctiPi+/3iqi}i is precompact in W. Moreover, we havep; —>• p
and qi -» q in W.
Since (aj,6j) is bounded away from (0,0) in E 2 and since st —t 0 and
ii —>• 0 in E, it follows that (aj,/3j) —• oo in K2. Choose a sequence Mj
in E such that Ui —>• 0 in E and such that {ui(a;i,/3i)}i is precompact in
E2 \{(0,0)}. Passing to a subsequence, assume that Ui(ati, ft) is convergent
in E 2 \{(0,0)}. Let («',/?') := lim Wi(ai,ft); then (a',ft) € E 2 \{(0,0)}.
i—>oo
Since u, —> 0 in E and {aiPi + Piqi}i is precompact in W, we conclude that
Ui{ctiPi + Piqi) —> 0 in V. On the other hand, Ui(aiPi + /3iqi) -> a'p + /3'g
in V. Then a'p + ft'q' = 0, so, by linear independence of p and q, we have
(a', ft) = (0,0), contradiction. •
Let W be a vector space and let T, be a sequence in End(W). Let V be a
sequence of vector subspace of W. Then we say that Ti is expansive on V
if, for every sequence v^ in V, we have: if {Ti(vi)}i is precompact in W, then
«j —> 0 in V. That is, Tj is expansive on the constant sequence V, V,V,....
Note that we do not assume, for any i, that Tj(V) C V.
COROLLARY 2.6.3 Let W be a vector space and let Ti be a sequence in
End(W). Let x,y,p,q G W. Assume that both Ti(x) -*• p and Ti(y) —v q
Topological groups 25
in W. Assume that p and q are linearly independent in W. Then Ti is
expansive on Kx +My.
L E M M A 2.6.4 Let W be a vector space and let Ti be a sequence in
End(W). LetV andW be a vector subspaces ofW. Assume VnW = {0}.
Let Vi be a sequence of vector subspacse of V. Let W := W/W. Let
n : W -> W be the canonical linear map. Let Ti be a sequence in End(W).
Assume, for all i, that 7roTj = T,o7r. Assume that Ti is expansive on 7r(V;).
Then Ti is expansive on Vi.
Proof. Let Vi be a sequence in V; and assume that Ti(vi) is precompact
in V. We wish to show that Vi —> 0 in V.
Let V := 7r(V). For all i, let V := ir(Vi). For all i, let vi := ir(yi). Then
for all i, we have 7r(Ti(uj)) = Ti{vt). So, since TJ(UJ) is precompact in W,
it follows that Tt{vi) is precompact in W. So, as Ti is expansive on Vi, we
see that v\ -* 0 in W. Then 7r(uf) ->• 0 in V. AsVnW = {0}, it follows
that the map v •-»• 7r(u) : V —> V is a vector space isomorphism. So, as
7T(DJ) —> 0 in V, we conclude that Vi —> 0 in V. D
2.7 Topological groups
By topological group, we will always mean a Hausdorff topological group.
Let G be a topological group. The complement of an open subgroup
of G is a union of cosets and is therefore open. Consequently, any open
subgroup of G is closed. Consequently, if G is connected, then no proper
subgroup of G can be open. Consequently, if G is connected, if S C G and
if S contains a nonempty open subset of G, then 5 generates G, i.e., no
proper subgroup of G contains S.
If G is a topological group, then G° denotes the connected component
of G containing 1Q. If J? is a closed subgroup of a topological group G,
then H° denotes the connected component of H (in the relative topology
inherited from G) containing la- If G is a topological group, then we define
Z°(G) := (Z(G))°.
Let G be a topological group. For S CG, let C%(S) := (CG(S))° and
AT° (S) := (NG(S))°. For g G G, let C°G(g) •= (CG(<?))°.
If G is a locally compact topological group, then we give Aut(G) the
compact-open topology and we define Aut°(G) := (Aut(G)) 0 .
Let G be an lcsc group. A Borel measure /i on G is left-invariant
26 Basic Results and Definitions
(resp. right-invariant) if, for all Borel S C G, for all g € G, we have
fi(gS) = fi(S) (resp. n(Sg) = l*(S)). A left-invariant (resp. right-invariant)
positive Borel measure on G is called a left Haar measure (resp. right
Haar measure) on G.
Let G be an lcsc group. Let A denote either the set of all left Haar
measures on G or the set of all right Haar measures on G. Then A ^ 0.
Moreover, if A, p, € A, then there exists a > 0 such that A = ap,. Moreover,
for all A € A, A is positive on all nonempty open subsets of G and finite on
all compact subsets of G. Moreover, if, for some A 6 A, we have \(G) < oo,
then G is compact. Conversely, if G is compact, then, for all A 6 A, we
have A(G) < oo.
Let G be a topological group acting on a set X. Then (Stabc(a;)) 0 is
called the connected stabilizer in G of x, and is denoted Stab^(a;). The
G-action on X action is locally free if, for all x € X, Stabc(x) is discrete
in G. The action is locally faithful if its kernel is discrete in G.
Let G be a topological group acting on a topological space X. Then
the G-action on X is continuous if the map (g, x) t-> gx : G x X —> X is
continuous.
L E M M A 2.7.1 Let G be an lcsc topological group and let G act contin-
uously on topological spaces X and Y. Assume that Y is Baire. Assume
that the G-action on Y is transitive. Let f : X —> Y be a continuous
G-equivariant map. Then f : X —>Y is open.
Proof. For all x £ X, let hx : G —>• X be defined by hx(g) = gx; then hx
is continuous. For all x € X, let px :— f o hx : G -¥ Y. For any U C X,
for any a; e I , we have PxiK1^)) = f(U fl (Gx)). So, for any U C X,
we have f(U) = M Pxih^iU)). We wish to show, for any open subset U
of X, that f(U) is open in Y. It therefore suffices to show, for all x € X,
that px : G -» Y is open. Fix x £ X and fix an open subset W of G. We
wish to show that px(W) is open in Y.
Let p := px : G -> Y. For any subset A C Y, let A° denote the
interior in Y of A and let A denote the closure in Y of A. Fix w € W.
We wish to show that p{w) € {p(W))°. Let W := w~lW. Then W is
an open neighborhood of \Q in G. Let y0 := P(1G)- Then y0 = / ( ^ ) -
Since wy0 — p(w) and since w(p(W')) — p(W), it suffices to show that
</o e (P(W'))°.
Topological groups 27
For all A C G, let A denote the closure in G of A. Choose an open neigh-
borhood V of la in G such that V is compact and such that V V C W.
Choose a sequence <?, in G such that MffiV = G. Then M<7JVJ/O = Gyo,
i i
so, by transitivity of G on Y, we conclude that MffiVyo = Y.
i
Since Y is Baire, choose io such that (gi0Vyo) ^ 0. So, since we have
ffj^1 {OioVyo) = {Vyo) J w e conclude that (Vy0) ^ 0. Since V is compact,
it follows that Vyo is compact, so Vyo is closed in y . So, as Vyo 2 V2/o> we
conclude that Vy0 D Vy~o~. Then (Vy0)° 2 (Vy^)° ^ 0. Choose x £ (Vy0)°.
Then x £ Vyo, so choose v £ V such that x = i;?/o. Then
2/o = v~xx e w-MCVyo)"] = (v-Wyo)0 C ( V ^ F i / o ) 0 C (W"i/ 0 )°.
For all g £G,we have p(g) = S(P(1G)) = ffj/o- Then p(W) = W'j/o- Thus
2/o € (p(W))°. •
A subgroup B of an lcsc topological group G is said to be cocompact
in G if there is a compact subset K of G such that G = BK. Equivalently,
there is a compact subset K of G such that G — KB. If if is a closed
subgroup of an lcsc topological group G, then B is cocompact in G iff G/H
is compact in the quotient topology.
L E M M A 2.7.2 Let B be an lcsc topological group and let G be a normal
closed subgroup of B. Let IT : B —> BjG be the canonical homomorphism.
Let T be a closed subgroup of B. Assume thatTtlG is cocompact in G and
that 7r(r) is cocompact in B/G. Then T is cocompact in B.
Proof. Let e := l j j . Let K be a compact subset of G such that we have
K(T n G) = G. Then G C KT. Let L be a compact subset of B/G such
that L • 7r(r) — B/G. Let L be a compact'subset of B such that L C ir(L).
Then L r G = n^^Lr)) = ^ ( Z - T r C r ) ) = ^(B/G) = B.
Since G is normal in if, we have G r = TG. Then
H = LYG = LGT C L ( i i T ) r = LAT C if,
so B = LKT. So, since LK is a compact subset of B, it follows that T is
cocompact in if. D
L E M M A 2.7.3 Let G be a lcsc topological group. Let A and B be closed
subgroups of G. Assume that AC\ B = {1G} and that AB = G. Assume
28 Basic Results and Definitions
that A normalizes B. Let AQ be a cocompact closed subgroup of A and let
BQ be a cocompact closed subgroup of B. Assume that AQ normalizes BQ.
Then AQBQ is a cocompact closed subgroup of G.
Proof. The multiplication map F : A \x B —> G defined by F(a, b) = oh
is a continuous bijective homomorphism. By Lemma 2.7.1, p. 26, the map
F : A K B -> G is open. Then F : A tx B -» G is an isomorphism of lcsc
topological groups.
Since A0 is closed in A and since BQ is closed in B, it follows that
A0 x B0 is closed in AxB, and therefore that ^o-Bo = F(A0 * B0) is closed
in G = F(A K B). It remains to show that AQB0 is cocompact in G.
Let Co := AoB0. Choose a compact subset K of G such that BQK = B.
Then C0K = A0B0K = A0B. Then KC0 — A0B. Choose a compact
subset K' of G such that K'A0 = A. Then
K'K(AoBo) = K'KCo = K'A0B = AB = G.
Then AQBQ is cocompact in G. O
L E M M A 2.7.4 Let G and H be lcsc topological groups. Let f : G -* H
be a continuous surjective homomorphism. Let K denote the kernel of f
and let n : G —> G/K be the canonical homomorphism. Then there is an
isomorphism f : G/K —tHof lcsc topological groups such that f on = f.
Proof. By abstract group theory, fix an isomorphism / ' : G/K —> H of
abstract groups such that f on = f. Since / : G -> H is surjective, we see
that / ' : G/K —> H is surjective. Since / : G —> H is continuous, we see
that / ' : G/K -> H is continuous. It remains to show that / ' : G/K -> H
is open.
Let G act on H by g • h — (f(g))h. By surjectivity of / : G -> H, this
action is transitive. So, as / ' : G/K —> H is G-equivariant, we see, from
Lemma 2.7.1, p. 26, that / ' : G/K -> H is open. •
COROLLARY 2.7.5 Let G and H be lcsc topological groups. Then any
continuous surjective homomorphism G —> H is open.
L E M M A 2.7.6 Let G and H be lcsc topological groups and let f : G -> H
be a continuous, bijective homomorphism. Then f is an isomorphism of
topological groups.
Discrete groups 29
Proof. Let K := ker(/). Then, as / is injective, we have K = {1G}- Let
•n and / ' be as in Lemma 2.7.4, p. 28. Then 7r and / ' are both isomorphisms,
so, as / = / ' o 7r, we conclude that / is an isomorphism, as well. D
L E M M A 2.7.7 Let G and H be Icsc topological groups and let f : G -» H
be a continuous homomorphism. Then f : G —>• H is proper iff both ker(/)
is compact and f(G) is closed in H.
Proof. "Only if" follows from the fact that / is proper iff both / is closed
and every fiber of / is compact. Assume that K := ker(/) is compact and
that Ho := f(G) is closed in H. We wish to show that / : G —> H is proper.
Since Ho is a closed subgroup of H, the inclusion Ho —>• H is proper.
Let n : G —> GjK be the canonical homomorphism. By Lemma 2.7.4,
p. 28, choose an isomorphism / ' : GjK -> H0 of topological groups such
that / = / ' o 7r. Since / ' and TT are both proper, it follows that / : G —> Ho
is as well. So, since the inclusion Ho -» H is also proper, it follows that
/ : G -> H is proper. •
L E M M A 2.7.8 Let G be an Icsc topological group, let K be a compact
subset ofG and let A be a closed subset ofG. Then AK and KA are closed
subsets of G.
Proof. The maps (a, k) \-t ak : A x K -» G and (k, a) *-* ka : K x A -> G
are both proper, and so have closed image. •
2.8 Discrete groups
Any subspace of a second countable topological space is again second count-
able, and a discrete, second countable topological space is countable. Thus,
a discrete subset of a second countable topological space is countable. By
the Baire Category Theorem, any countable Polish topological space is dis-
crete. By Theorem 3.11, p. 17 of [Kec95], a Gs subspace of a Polish space
is again Polish; then any countable, Gs subset of a Polish topological space
is discrete. In particular, a countable, closed subgroup of a Polish topo-
logical group is discrete. A locally closed subgroup of a topological group
is closed, so a discrete subgroup of a topological group is closed. If G is a
Polish topological group and if T is a subgroup of G, the T is discrete in G
iff r is a countable closed subgroup of G.
30 Basic Results and Definitions
If G is a topological group and if V is a subgroup of G, then T is discrete
in G iff 1 Q is not an element of the closure in G of T \ { 1 G } iff there exists
an open neighborhood U of 1Q in G such that U (~\ V = {1G}-
L E M M A 2.8.1 Let H be a topological group and let G be a closed subgroup
of H. Let 7r : H —> H/G be the canonical map. Let T be a subgroup of H.
Assume that TC\G is discrete in G and that TT(T) is discrete in H/G. Then
T is discrete in H.
Proof. Let e := l j j . Let U be an open neighborhood of n(e) in H/G such
that U n 7r(r) = {n(e)}. Then ( T T " 1 ^ ) ) D F = F n G, Let V be an open
neighborhood of e in if such that V n ( r n G ) = {e}. Let W := (n'1 (U))nV.
Then W nT CW n(TnG) = {e}. Thus F is discrete in H. O
A subgroup r of an lcsc group G is said to be a lattice in G if both T
is discrete in G and the transitive G-action on G/T preserves a finite Borel
measure.
L E M M A 2.8.2 Let H be a closed subgroup of an lcsc topological group G.
Then the following are equivalent:
(1) H is discrete.
(2) H is a zero-dimensional topological space.
(3) H is totally disconnected.
(4) H is totally path-disconnected.
(5) H° = {1G}.
Proof. The implications (1) =>• (2) = > (3) =>• (4) follow from the defi-
nitions. Since H° is a connected Lie group and is therefore path-connected,
we see that (4) =>• (5). We assume that H° = {1 G } and we wish to show
that H is a discrete subset of G.
Let T denote the discrete topology on H and let \x be the relative topol-
ogy on H inherited from G. Then the identity map (H,T) -> (H,fj.) is
a bijective continuous homomorphism of lcsc groups, so, by Lemma 2.7.6,
p. 28, we see that it is an isomorphism of topological groups. That is, the
discrete topology on H agrees with the relative topology. That is, H is a
discrete subset of G. D
Proper actions 31
2.9 Proper actions
Let G be an lcsc topological group acting on an lcsc topological space X.
The G-action on X is proper if (g,x) H-> (a;,gx) : GxX —> X xX is proper.
Equivalently, for any compact subset K C X, {g £ G \ gK n K ^ 0} is a
compact subset of G. When G is a discrete group, many writers prefer the
term properly discontinuous instead of proper. The G-action on X is
nonproper if it is not proper. Equivalently, there are sequences gi in G
and x, in X such that gi —> oo in G, such that a;, converges in X and such
that giXi converges in X.
L E M M A 2.9.1 Let G be an lcsc topological group acting properly on an
lcsc topological space X. Then the quotient topology on G\X is Hausdorff.
Proof. Fix x, y £ X. Assume that Gx ^ Gy. We wish to show that
there are, in X, G-invariant open neighborhoods UQ and Vo of x and y,
respectively, such that Uo n Vo = 0. It suffices to show that there are,
in X, open neighborhoods U and V of x and y, respectively, such that
(GU) n (GV) = 0.
Let V\ be a precompact open neighborhood of y in X. By properness of
the action, choose a compact subset C CG such that, for all g £ G\C, we
have gx £Vi. For all S C X, let S denote the closure in X of S. Choose
an open neighborhood V of y such that V C Vi\(Ga;). Then (Gx) n V = 0.
Choose an open neighborhood U\ of x in X such that U\ is compact.
By properness of the action, choose a compact subset K CG such that, for
all g £ G\K, for all z £ V, we have gz £ U\.
As (Gx) n V = 0, it follows that x £ GV. In particular, x £ KV. Since
K is compact and since V is closed in X, it follows that KV is closed in X.
Let U := E/i\(ifV). Then U is an open neighborhood of x. Moreover,
U C C/i and [7 n (2fF) = 0, so, by definition of K, we get U n (GV) = 0.
Then C7 n (GV) = 0. Then (GU) n (GV) = 0. •
Let G be an lcsc topological group acting properly on an lcsc topological
space X. Because the canonical map X -* G\X is open, by Lemma 2.9.1,
p. 31, it follows that the quotient topology on G\X is lcsc. (Recall that,
under our definitions, lcsc implies Hausdorff.)
32 Basic Results and Definitions
2.10 Bilinear and quadratic forms
Let F b e a vector space and let B : V x V -> E be a function. We say that
the map B is a bilinear form on V if, for all v € V,
w t-> B(v, w) : V -> E and w H-> B(IU, u) : V ->• E
are both linear transformations. We say that B is symmetric if, for all
v,w G V, B(v,w) — B(w,v). We say that B is antisymmetric or alter-
nating if, for all v,w G V, we have B(v,w) = —B{w,v). If B is a bilinear
form, then there exist unique symmetric and antisymmetric bilinear forms
S and A such that B = S + A.
The kernel of a symmetric bilinear form B : V x V —> E is
ker(B) := {v € V | Vw G V, £(u,u;) = 0}.
We say that B is degenerate if ker(B) ^ {0}. We say that B is nonde-
generate otherwise. If {vi,... ,Vd} is a basis of V, then B is degenerate
iff the matrix [B(vi,Vj)] has determinant = 0.
Let V be a vector space. If B : V x V -4 E and B ' : V x V - • K
are bilinear forms, then we say that B and B' are equivalent if there is a
vector space isomorphism f :V ->V such that, for all v, w € V, we have
B'{f(v),f(w)) — B(v,w). Let d :— dim(V). There is a nondegenerate
antisymmetric bilinear form on V iff d is even. If d is even, then any two
nondegenerate antisymmetric bilinear forms on V are equivalent.
A symmetric bilinear form B : V x V —> E is positive definite if, for
all v G ^ \ { 0 } , we have B(v,v) > 0. It is negative definite if, for all
v £ V\{0}, we have B(v,v) < 0. It is positive semidefinite if, for all
v G V, we have B(v,v) > 0. It is negative semidefinite if, for all v &V,
we have B(v, v) < 0. It is definite if it is either positive definite or negative
definite. It is indefinite if it is not definite.
An inner product space is a vector space V together with a positive
definite symmetric bilinear form B :V xV -> R.
Let V be a vector space and let B : V x V -> E be a symmetric bilinear
form. A vector v G V is isotropic with respect to B if B(v,v) = 0. A
subset VQ C V is isotropic with respect to B if, for all v,w G Vo, we have
B(v,w) = 0. A vector subspace VQ of V is positive definite (resp. nega-
tive definite, resp. positive semidefinite, resp. negative semidefinite,
resp. definite, resp. indefinite, resp. degenerate, resp. nondegenerate)
Bilinear and quadratic forms 33
with respect to B if B\(VQ X Vo) is positive definite (resp. negative defi-
nite, resp. positive semidefinite, resp. negative semidefinite, resp. definite,
resp. indefinite, resp. degenerate, resp. nondegenerate). Let V denote the
collection of all positive definite subspaces of V. Let M denote the collection
of all negative definite subspaces of V. Let p := max{dim(HA) | W € V}
and let q := max{dim(W) | W E N}- Then (p, q) is the signature of B.
Let V be a vector space and let B : V x V -> E be a symmetric
bilinear form of signature (p, q). Then dim(ker(B)) = (dim(V)) — (p + q).
In particular, B is nondegenerate iff p + q = dim(y).
Let d > 1 be an integer, let V and V be d-dimensional vector spaces
and let B : V x V -> E and B' : V x V -» E be symmetric bilinear forms.
Let (p, q) be the signature of B and let (p1, q') be the signature of B'. Then
B and B' are equivalent iff (p,q) = (p',q')-
Let V be a vector space and let d := dim(F). Let I := { 1 , . . . , d} and
let I2 := I x I. Let A := {(i,i) \i 6 / } C J 2 . For any nondegenerate
symmetric bilinear form i? : V x V —> E, there is a basis {vi,..., v^} of V
such that, for all i € / , we have B(vi,Vi) € { + 1 , - 1 } and such that, for
all (i, j) 6 I 2 \ A , we have B(v,,Vj) = 0. That is, [B(vi,Vj)] is a diagonal
matrix and every diagonal entry is either + 1 or — 1. If p is the number of
diagonal entries equal to + 1 and q is the number of diagonal entries equal
to —1, then B has signature (p, q).
Let V be a vector space. A function Q : V —> E is said to be a quadratic
form if there exists a bilinear form B :VxV —> R such that, for all v 6 E d ,
Q(v) = B(v,v).
For any quadratic form Q on a vector space V, there exists a unique
symmetric bilinear form B on V such that, for all v £ V, Q(v) = B(v,v).
The symmetric bilinear form B is said to be the polarization of Q. The
signature of a quadratic form is, by definition, the signature of its polar-
ization. A quadratic form is positive definite (resp. negative definite,
resp. positive semidefinite, resp. negative semidefinite, resp. definite,
resp. indefinite, resp. degenerate, resp. nondegenerate) if its polariza-
tion is. We define the kernel of Q to be the kernel of the polarization
of Q; it is denoted ker(Q). A vector v € V is isotropic with respect to
Q if Q(v) — 0. A subset VQ C V is isotropic with respect to Q if it is
isotropic with respect to the polarization of Q. A vector subspace Vo of V is
positive definite (resp. negative definite, resp. positive semidefinite,
resp. negative semidefinite, resp. definite, resp. indefinite, resp. de-
34 Basic Results and Definitions
generate, resp. n o n d e g e n e r a t e ) with respect to Q if <2|Vb is positive
definite (resp. negative definite, resp. positive semidefinite, resp. negative
semidefinite, resp. definite, resp. indefinite, resp. degenerate, resp. nonde-
generate).
Two quadratic forms are equivalent if their polarizations are. Let V
and V be vector spaces and let Q : V -¥ E and Q' : V -»• E be quadratic
forms. Then Q is equivalent to Q' iff there is a vector space isomorphism
/ : V -> V such that, for all v € V, we have Q'(f(v)) = Q{v).
A function Q : Rd —> E is a quadratic form iff Q is a homogeneous poly-
nomial of degree two, i.e., iff there is asystem {cij}i<i<j<d of real numbers
such that, for all xi,...,Xd € E, we have Q{x\,... ,xd) = T J CijXiXj.
l<i<j<d
In this case, the polarization B : M.d x E d -* M. of Q is given by the rule:
For all x = (xu...,xd)e Rd, for all y = (ylt... ,yd) € W,
B{x,y) = \ y ^ ckkXkVk
/ j
l<i<j<d
^ijy^iVj T XjUi)
+
.*=!
Let V be a vector space, let d := dim(F) and let Q : V -> R be a
quadratic form. Then <5 is degenerate iff there exists a linear transformation
L : V -> E d _ 1 and a quadratic form F : IT*-1 -> E such that Q = F o L.
Let V be a vector space. Then SBF(F) denotes the vector space of all
symmetric bilinear forms on V, and QF(V) denotes the vector space of all
quadratic forms on V.
Let V be a vector space and let B £ SBF(V). A linear transformation
T : V —¥ V is said to be symmetric with respect to B (resp. anti-
symmetric with respect to B) if, for all v, w 6 V, we have B(T(v),w) —
B(v,T(w)) (resp. B(T(v),w) = -B(v,T(w))). If T : V -» V is symmet-
ric (resp. anti-symmetric) with respect to B, then the finite-dimensional
spectral theorem asserts that T is real diaongalizable (resp. elliptic).
Let V be a vector space and let B € SBF(V). For any X,Y CV, we
define B(X,Y) := B(X xY) = {B(x,y)\x £ X,y e Y}. For any X CV,
for any v eV,we define B(X,v) := S(X,{u}) and B(u,X) := B({v},X),
so, by symmetry, B(X, v) = B(w,X).
L E M M A 2.10.1 Let V be a vector space and let Q e QF(V). Assume
that Q is positive semidefinite. Then ker(Q) = Q _ 1 (0).
Proof. Let B £ SBF(V^) be the polarization of Q. By definition of
Root systems 35
ker(Q), we have ker(Q) C <2 _1 (0). Fix v e Q _ 1 ( 0 )- We wish to show
that v E ker(Q). Fix w £ V and let o := B(v,w). We wish to show that
a = 0. Assume, for a contradiction, that o ^ O .
We have B(v,v) = Q(u) = 0. Let b := B(w,w). Let * := -(b + l)/(2a).
Then Q(tu + to) = B(t« + WJ, tu + w) = 0 + 2ta + 6 = - 1 < 0, contradicting
the assumption that Q is positive semidefinite. D
C O R O L L A R Y 2.10.2 Let V be a vector space and let Q 6 QF(V) be
nondegenerate. Then
(1) Q is positive definite iff Q is positive semidefinite; and
(2) Q is negative definite iff Q is negative semidefinite.
Proof. For both (1) and (2), "only if" follows from the definitions.
Proof of "if" in (1): Since Q is nondegenerate, ker(Q) = {0}, so, by
Lemma 2.10.1, p. 34, we have Q - 1 (0) = {0}, so Q is definite. Then, as Q
is definite and positive semidefinite, we see that Q is positive definite. End
of proof of "if" in (1).
Proof of "if" in (2): Since — Q is nondegenerate and positive semidef-
inite, by "if" of (1) of Corollary 2.10.2, p. 35, we see that — Q is positive
definite, and so Q is negative definite. End of proof of "if" in (2). •
The following is a consequence of the Chain Rule.
L E M M A 2.10.3 Let V be a vector space. For all v e V, let iv : V -> TVV
be defined by iv(w) — (d/dt)t=o(v + tw). Let 6 > 0. Let I := (—6,6). Let
a,b : I —> V be smooth. Let x := a(0) and y := 6(0). Let
x' := (d/dt)t=0(a(t)) e TXV and y' := (d/dt)t=0(b(t)) € TyV.
Let F € SBF(V). Define f : I -> E by f(t) = F(a{t), b{t)). Let f : I -> R
denote the derivative of f. Then /'(0) = F(i~l(x'),y) + F(x, i~1 (?/'))•
2.11 Root systems
Let (V, B) be an inner product space. For all v €V, define
V1 := {w e V\B(v,w) =0}.
For all v € ^ \ { 0 } , define rv : V -> V be the reflection in vx defined by
rv(w) =w- 2[(B(v,w))/(B(v,v))]v; then
Other documents randomly have
different content
darum zu ersuchen; wenn nun zwei es verweigern, mit darum
anzuhalten, so kann Sjnetkoff die Ballotage verweigern, und dann
wird die Partei der Alten einen anderen von den Ihrigen wählen,
sodaß unser ganzer Plan verloren ist. Wenn aber nur der eine Kreis
Swijashskiys nicht mit bittet, so wird Sjnetkoff ballotieren. Man wird
ihn dann selbst wählen und ihm absichtlich das Amt wieder
übertragen, sodaß sich die gegnerische Partei verrechnet, und wenn
sie einen Kandidaten von den Unseren aufstellen, diesem das Amt
überträgt.“
Lewin verstand, aber nicht vollständig, und wollte soeben noch
einige Fragen stellen, als plötzlich alle durcheinander zu sprechen
begannen, lärmten und sich nach dem großen Saale in Bewegung
setzten.
„Was ist das? Wie? Wen wird man wählen? Vertrauen? Zu wem?
Was ist? Verwirft man? Es giebt kein Vertrauen! Man läßt Phleroff
nicht zu. Was; unter Anklage! So läßt man niemand zu! Das ist
niedrig! Das Gesetz!“ hörte Lewin von verschiedenen Seiten rufen,
und begab sich zusammen mit der Menge, die sich drängte, und zu
fürchten schien, daß sie etwas versäumte, in den großen Saal. Er
näherte sich in dem Gedränge der Adligen dem Gouverneurstisch, an
welchem der Gouvernementsvorsteher, Swijashskiy und andere
Wortführer eifrig miteinander debattierten.
28.
Lewin stand ziemlich entfernt. Ein schwer und heiser atmend
neben ihm stehender Adliger und ein zweiter mit knarrenden, dicken
Stiefelsohlen, störten ihn, deutlich zu hören. Aus der Ferne vernahm
er nur die weiche Stimme des Gouvernementsvorstehers, darauf das
pfeifende Organ des scharfzüngigen Adligen und dann die Stimme
Swijashskiys. Sie stritten, soviel er zu verstehen imstande war, über
die Bedeutung eines Paragraphen des Gesetzes und den Sinn der
Worte, „wer sich unter gerichtlicher Untersuchung befindet“.
Der Haufe teilte sich, um dem zum Tisch herantretenden Sergey
Iwanowitsch Raum zu geben. Sergey Iwanowitsch sagte, nachdem
er die Beendigung der Rede des scharfzüngigen Adligen abgewartet
hatte, ihm scheine, daß es am richtigsten sei, sich nach dem
Paragraphen des Gesetzes zu richten und bat den Sekretär, den
Paragraphen aufzusuchen. In demselben war gesagt, daß man im
Falle der Meinungsverschiedenheit zu ballotieren habe.
Sergey Iwanowitsch verlas den Paragraphen, und begann den
Sinn desselben zu erörtern, aber da unterbrach ihn ein großer, dicker
und krummer Gutsherr mit roten Ohren, in enger Uniform mit im
Nacken hinten hochstehendem Kragen. Er trat an den Tisch und rief
laut, mit einem Finger darauf schlagend:
„Ballotieren! Zu den Kugeln greifen! Weg da mit dem Geschwätz!
Zu den Kugeln!“
Mehrere Stimmen erhoben sich jetzt plötzlich zusammen, und der
große Adlige mit seinem Finger, mehr und mehr in Zorn geratend,
schrie lauter und lauter. Es ließ sich jedoch nicht unterscheiden, was
er sagte. Er sagte das Nämliche, was Sergey Iwanowitsch vorschlug,
aber offenbar haßte er diesen und dessen ganze Partei, und dieses
Gefühl des Hasses teilte sich nun der ganzen Partei mit und rief den
Widerstand einer gleichen, wenn auch gemäßigteren Erbitterung auf
der anderen Seite hervor. Rufe erschallten und eine Minute lang
wogte alles durcheinander, sodaß der Gouvernementsvorsteher
genötigt war, um Ordnung zu bitten.
„Ballotieren! Ballotieren! Wer ein Edelmann ist, der sieht das ein!
Wir vergießen unser Blut! Das Vertrauen des Monarchen! Nicht den
Gouvernementsvorsteher achten; er ist kein Amtmann! Darum
handelt es sich nicht! Bitte, zu den Kugeln! Es ist eine Schande!“
vernahm man zornige, sinnlose Schreie von allen Seiten.
Die Blicke und Gesichter wurden immer zorniger und die Reden
immer ungebärdiger. Sie drückten einen unversöhnlichen Haß aus.
Lewin begriff nicht im geringsten, worum es sich handle, und war
erstaunt über die Leidenschaftlichkeit, mit welcher die Frage, ob
man über Phleroff ballotieren solle oder nicht, behandelt wurde. Er
hatte, wie ihm später Sergey Iwanowitsch erklärte, jenen
Syllogismus vergessen, daß im Interesse des allgemeinen Wohls der
Gouvernementsvorsteher entfernt werden müsse; zu der Entfernung
desselben aber war eine Majorität der Kugeln erforderlich; zur
Erlangung dieser Majorität weiterhin mußte man Phleroff Stimmrecht
erteilen, und zur Anerkennung Phleroffs als eines Stimmberechtigten
mußte man erklären, wie der Paragraph des Gesetzes aufzufassen
sei.
„Also eine Stimme kann die ganze Angelegenheit entscheiden, und
man muß daher ernst und konsequent sein, wenn man der
gemeinsamen Sache dienen will,“ schloß Sergey Iwanowitsch.
Lewin hatte dies jedoch vergessen, und es wurde ihm schwer ums
Herz, diese von ihm geachteten braven Männer in einer so
unangenehmen schlimmen Erregung sehen zu müssen.
Um sich von diesem beklemmenden Gefühl frei zu machen, ging
er, ohne das Ende des Streites abzuwarten, in den Saal, in welchem
sich niemand befand, als einige Lakaien beim Buffet. Als er die mit
dem Abwischen von Geschirr, dann Aufstellen von Tellern und
Gläsern beschäftigten Diener, ihre ruhigen, aber lebhaften Gesichter
sah, empfand Lewin ein unerwartetes Gefühl der Erleichterung, als
sei er aus einem Zimmer voll widrigen Geruchs in die reine Luft
hinausgetreten. Er begann auf und abzuschreiten, mit Befriedigung
auf die Diener blickend. Es gefiel ihm sehr, als einer derselben, ein
Mann mit grauem Backenbart, den jüngeren, die diesen zum besten
hatten, voll Geringschätzung lehrte, wie man Servietten falten
müsse. Lewin hatte sich gerade mit dem alten Diener in ein
Gespräch eingelassen, als der Sekretär der Adelsvormundschaft, ein
alter Herr, der die spezielle Eigenschaft besaß, alle Adligen des
Gouvernements dem Namen und der Herkunft nach zu kennen, ihn
davon abzog.
„Bitte gefälligst, Konstantin Dmitritsch,“ sagte er zu ihm, „Euer
Bruder sucht Euch. Es wird ballotiert.“
Lewin trat in den Saal, erhielt eine weiße Kugel und begab sich
hinter seinem Bruder Sergey Iwanowitsch zum Tische, an welchem
mit wichtiger, ironischer Miene Swijashskiy stand, der seinen Bart in
die volle Hand genommen hatte und daran roch.
Sergey Iwanowitsch steckte die Hand in den Kasten, legte seine
Kugel hinein und blieb, Lewin Platz machend, am Orte stehen. Lewin
trat heran, wandte sich aber, da er vollständig vergessen hatte, um
was es sich handle und in Verlegenheit geraten war, an Sergey
Iwanowitsch mit der Frage, wohin er die Kugel legen solle? Er frug
leise, während man in seiner Nähe sprach, sodaß er hoffen konnte,
es habe niemand seine Frage vernommen. Aber die Sprechenden
verstummten und seine unziemliche Frage wurde gehört. Sergey
Iwanowitsch runzelte die Stirn.
„Das ist Sache der Überzeugung für einen jeden,“ sagte er
gemessen. Einige lächelten. Lewin errötete, streckte hastig die Hand
unter das Tuch und legte die Kugel nach rechts, da sie sich gerade in
seiner rechten Hand befand. Nachdem er dies gethan hatte, besann
er sich, daß er auch die linke Hand hineinstecken müsse und steckte
sie hinein, doch schon zu spät, und zog sich dann, noch mehr in
Verlegenheit geraten, schnell in die hintersten Reihen zurück.
„Einhundertsechsundzwanzig dafür! Achtundneunzig dagegen!“
klang die Stimme des Sekretärs, welcher den Buchstaben r nicht
aussprechen konnte. Gelächter erschallte: ein Knopf und zwei Nüsse
waren in dem Kasten gefunden worden. Der Adlige war zugelassen
und die Jungpartei hatte gesiegt. Aber die Partei der Alten hielt sich
noch nicht für besiegt. Lewin vernahm, daß man Sjnetkoff bat, zu
ballotieren, und gewahrte, daß ein Trupp der Edelleute den
Gouvernementsvorsteher umringte, welcher sprach. Lewin trat
näher. Sjnetkoff sprach, indem er den Edelleuten antwortete, vom
Vertrauen des Adels, von dessen Liebe zu ihm, deren er nicht würdig
sei, da sein ganzes Verdienst nur in seiner Ergebenheit für den Adel
bestehe, dem er zwölf Jahre des Dienstes geweiht hätte. Mehrmals
wiederholte er die Worte „ich habe gedient, soviel es meine Kräfte
gestatteten, im Glauben und in der Wahrheit; ich schätze euch hoch
und danke euch!“ und plötzlich hielt er, von Rührung überwältigt,
inne und verließ den Saal.
Mochten nun diese Thränen von dem Gefühl einer Ungerechtigkeit
gegen ihn, von der Liebe zum Adel, oder von der Gespanntheit der
Situation herrühren, in welcher er sich befand, indem er sich von
Gegnern umgeben fühlte — genug, die Erregung teilte sich weiter
mit, die Majorität des Adels war gerührt und auch Lewin empfand
ein Gefühl von Zärtlichkeit für Sjnetkoff.
In der Thür stieß der Gouvernementsvorsteher mit Lewin
zusammen.
„Entschuldigt, bitte, entschuldigt,“ sagte er wie zu einem
Unbekannten, lächelte aber, als er Lewin erkannt, und diesem schien
es, als ob er etwas hätte sagen wollen, aber vor Erregung nicht
sprechen könne.
Der Ausdruck seines Gesichts und seiner ganzen Gestalt in der
Uniform, mit den Ordenskreuzen und den weißen galonnierten
Beinkleidern, erinnerte Lewin, als er so hastig dahinschritt, an ein
gemästetes Schlachtvieh, welches sieht, daß es mit seiner Sache
übel bestellt ist.
Der Gesichtsausdruck des Mannes hatte etwas eigentümlich
Rührendes für Lewin, welcher erst am Tage vorher in
Vormundschaftsangelegenheiten in seinem Hause gewesen war und
ihn in der ganzen Größe eines guten und geselligen Menschen
kennen gelernt hatte. Das große Haus mit den altertümlichen
Familienmeubles; keine geschniegelten oder unsauberen, sondern
ehrerbietige alte Diener, offenbar noch aus der Zeit der ehemaligen
Leibeigenschaft stammend, die ihren Herrn nicht geändert hatten;
eine wohlbeleibte, gutmütige Hausfrau im Häubchen mit Spitzen und
einem türkischen Spitzenshawl, welche ihr liebes Enkelchen, die
Tochter ihrer Tochter liebkoste; ein jugendlicher Sohn, Gymnasiast
der sechsten Klasse, welcher von der Schule gekommen war und
den Vater begrüßte, indem er ihm die große Hand küßte; die
ermahnenden, freundlichen Reden und Gebärden des Hausherrn;
alles dies hatte in Lewin gestern unwillkürlich Hochachtung und
Sympathie erweckt. Jetzt erschien ihm dieser alte Herr rührend und
beklagenswert und er wünschte, ihm einige angenehme Worte zu
sagen.
„Ihr werdet vielleicht wieder unser Vorsteher werden,“ sprach er.
„Kaum,“ versetzte jener, erschreckt aufblickend, „ich bin
abgespannt, schon alt; es giebt würdigere und jüngere als ich,
mögen die nun dienen,“ und der Vorsteher verschwand durch eine
Seitenthür.
Es trat nun der feierlichste Augenblick ein; man mußte zur Wahl
schreiten. Die Wortführer der einen und der anderen Partei zählten
an den Fingern die weißen und die schwarzen Kugeln nach.
Die Debatten wegen Phleroff hatten der Jungpartei nicht nur die
eine Kugel Phleroffs mehr verschafft, sondern auch einen Gewinn an
Zeit herbeigeführt, sodaß noch drei Adlige herbeigeholt werden
konnten, welchen es durch Intriguen der Alten unmöglich gemacht
worden war, an den Wahlen teilzunehmen. Zwei der Adligen, die eine
Schwäche für den Wein besaßen, hatte man durch Kumpane
Sjnetkoffs trunken gemacht, dem dritten die Uniform entwendet.
Als die Jungpartei dies erfahren hatte, sandte sie sogleich,
während der Debatten über Phleroff, in einer Mietkutsche Freunde
fort, um den Einen uniformieren zu lassen, und Einen der beiden
Berauschten zur Sobranje zu bringen.
„Den Einen habe ich gebracht, ich habe ihn mit Wasser begossen,“
sagte der nach ihm gesandte Gutsherr, zu Swijashskiy tretend, „doch
es ist nicht gefährlich, er taugt schon noch dazu.“
„Ist also nicht zu sehr berauscht, daß er nicht etwa umfällt?“ sagte
Swijashskiy kopfschüttelnd.
„Nein; er ist ganz munter; doch hatten sie ihn bald völlig
niedergetrunken. Ich habe dem Buffetier gesagt, er soll ihm auf
keinen Fall Wein geben.“
29.
Der enge Saal, in welchem man rauchte und aß, war gefüllt von
den Edelleuten. Die Aufregung stieg stetig, und auf allen Gesichtern
war die Unruhe bemerkbar. Namentlich waren die Wortführer in
Aufregung, da sie alle Einzelverhältnisse und die Berechnung aller
Kugeln kannten. Sie waren die Ordner der bevorstehenden Schlacht.
Die Übrigen suchten, wie die Krieger vor dem Kampfe, obwohl sie
sich zu diesem vorbereiteten, noch Zerstreuungen. Die Einen
nahmen am Tische stehend oder sitzend einen Imbiß; die Anderen
gingen Cigaretten rauchend, in dem langen Zimmer auf und nieder,
und unterhielten sich mit lange nicht gesehenen Freunden.
Lewin verspürte keine Eßlust, er rauchte nicht; zu den Seinigen,
mit Sergey Iwanowitsch, Stefan Arkadjewitsch, Swijashskiy und
anderen gehen, wollte er nicht, da bei ihnen in lebhaftem Gespräch
Wronskiy in Stallmeisteruniform stand. Schon gestern hatte ihn
Lewin bei den Wahlen erblickt, und geflissentlich vermieden, da er
nicht wünschte ihm zu begegnen. Er trat ans Fenster und setzte sich
nieder, auf die Gruppen blickend, und auf das horchend, was um ihn
herum gesprochen wurde.
Es bedrückte ihn namentlich, daß alle, wie er sah, aufgeregt,
besorgt und geschäftig waren, und nur er allein nebst einem alten,
zahnlosen Greis in Marineuniform, der neben ihm saß, ohne
Interesse und ohne Beschäftigung war.
„Das ist ein Betrug! Ich habe ihm gesagt, daß es nicht so ist.
Gewiß. Er konnte auf drei Jahre nicht wählen,“ sprach energisch ein
etwas verwachsener Gutsherr von kleiner Gestalt mit pomadisierten
Haaren, die auf dem gestickten Kragen seiner Uniform lagen, stark
mit den Absätzen seiner offenbar für die Wahlen neugefertigten
Stiefeln aufstampfend, und wandte sich dann, einen mißvergnügten
Blick auf Lewin werfend, kurz ab.
„Ja, eine unsaubere Sache, was man auch sagen mag,“ fuhr ein
kleiner Gutsherr mit dünner Stimme fort.
Hinter ihnen näherte sich Lewin eilig ein ganzer Trupp von
Gutsherrn, einen dicken General umringend. Die Herren suchten
offenbar einen Platz, wo sie so sprechen konnten, daß man sie nicht
hörte.
„Wie kann er sich unterstehen, zu sagen, ich hätte ihm die
Beinkleider entwenden lassen! Er hat sie vertrunken, denke ich! Ich
mache mir den Teufel aus ihm und seinem Fürstenrang! Er soll es
nicht wagen, das zu äußern. Eine Schweinerei ist es!“
„Aber erlaubt doch! Die berufen sich auf den Gesetzparagraphen,“
sprach man in einer anderen Gruppe, „die Frau muß als Adlige
eingetragen werden!“
„Zum Teufel mit dem Paragraphen! Ich spreche wie es mir ums
Herz ist! Darauf stützen sich brave Edelleute. Man soll Vertrauen
haben!“
„Wollen Ew. Excellenz mitkommen, fine champagne.“
Ein anderer Trupp ging zu einem laut schreienden Edelmann; es
war dies Einer der drei Berauschten.
„Ich habe der Marja Ssemionowna stets geraten zu verpachten,
weil sie ihren Vorteil nicht zu wahren versteht,“ sagte ein
graubärtiger Gutsherr mit angenehmer Stimme, in der
Oberstenuniform des alten Generalstabs. Es war dies der nämliche
Herr, welchem Lewin bei Swijashskiy begegnet war. Lewin erkannte
ihn sofort, auch der Gutsherr schaute nach Lewin, und sie begrüßten
sich.
„Sehr angenehm. Gewiß, ich erinnere mich noch recht wohl; wir
sahen uns im vergangenen Jahre bei Nikolay Iwanowitsch.“
„Nun, wie steht es mit Eurer Ökonomie?“ frug Lewin.
„Man arbeitet stets mit Schaden,“ antwortete der Gutsherr mit
höflichem Lächeln, aber mit einem Ausdruck von Ruhe und der
Überzeugtheit, daß es so sein müsse, neben Lewin stehen bleibend;
„aber wie kommt Ihr denn in unser Gouvernement?“ frug er dann.
„Ihr seid wohl gekommen, um an unserem coup d'état
teilzunehmen?“ sagte er, sicher, aber schlecht die französischen
Worte aussprechend. „Ganz Rußland ist zusammengekommen; auch
die Kammerherren und beinahe selbst die Minister.“ Er wies auf die
repräsentierende Gestalt Stefan Arkadjewitschs in den weißen
Pantalons und der Kammerherrenuniform, welcher mit einem
General ging.
„Ich muß Euch gestehen, daß ich die Bedeutung der Adelswahlen
recht wenig verstehe,“ sprach Lewin.
Der Gutsherr schaute ihn an.
„Was giebt es denn da zu verstehen? Eine Bedeutung liegt darin
gar nicht. Es ist das eine hinfällige Institution, welche ihr Fortleben
nur dem Trägheitsgesetz verdankt. Seht doch hin; die Uniformen
sagen Euch das ja schon. Dies ist eine Sobranje von
Friedensrichtern, dauernden Mitgliedern und so fort, aber nicht von
Edelleuten.“
„Aber weshalb seid Ihr denn gekommen?“ frug Lewin.
„Aus Gewohnheit; das ist das Eine. Ferner, um die Verbindungen
aufrecht zu erhalten. Darin liegt eine moralische Verpflichtung in
gewisser Hinsicht. Dann aber, wenn ich die Wahrheit sagen soll,
auch aus meinem eigenen Interesse. Mein Schwager wünscht als
Mitglied gewählt zu werden; die Familie ist arm und man muß sie
vorwärts bringen. Weshalb sind wohl diese Herren gekommen?“
sagte er, auf den bissigen Adligen zeigend, welcher hinter dem
Gouverneurstisch sprach.
„Das ist ein neues Adelsgeschlecht.“
„Neues hin, neues her; aber kein Adel! Das sind Landleute,
während wir Gutsherren sind. Sie legen als Adlige die Hand an sich
selbst.“
„Aber Ihr sagt doch, es sei dies eine abgelebte Institution?“
„Abgelebt hin, abgelebt her; man müßte sich aber doch
pietätvoller zu ihr stellen. Wäre wenigstens Sjnetkoff — — mögen
wir gut oder schlecht sein, aber tausend Jahre sind wir doch alt
geworden. Wißt Ihr, es kann einmal vorkommen, daß man vor
seinem Hause einen Garten anlegt und placiert. Da steht Euch nun
aber gerade auf dem Platze ein hundertjähriger Baum. Mag er gleich
knorrig und alt sein, für die Blumenbouquets wird man ihn doch
wohl nicht umhauen, sondern diese so anlegen, daß sie den Baum
umfangen;“ sagte er vorsichtig, und änderte darauf das Thema. „Wie
geht es denn mit Eurer Ökonomie?“ frug er.
„Auch nicht gut. Fünf Prozent wirft sie ab.“
„Da rechnet Ihr Euch aber noch nicht mit! Ihr seid doch auch
etwas wert! Ich kann auch von mir selbst so reden. Während der
Zeit, in der ich nicht Landwirtschaft trieb, habe ich im Dienst
dreitausend Rubel gehabt. Jetzt arbeite ich mehr, als im Dienst, und
habe ebenso wie Ihr, meine fünf Prozent, und damit ist es gut.
Meine Arbeit ist dabei noch umsonst.“
„Aber warum thut Ihr es denn, wenn Ihr nur Verlust habt?“
„Nun, man arbeitet eben. Was wollt Ihr sonst? aus Gewohnheit;
und man weiß, daß man so muß. Ich will Euch weiter sagen,“ und
der Gutsbesitzer stemmte sich mit den Ellbogen auf das Fenster und
fuhr fort, „mein Sohn hat keine Lust zur Landwirtschaft; er wird
offenbar einmal ein Gelehrter. Niemand wird somit die Sache einmal
fortführen, und doch arbeitet man. Jetzt habe ich einen Garten
angelegt.“
„Ja, ja,“ sagte Lewin, „das ist völlig richtig. Ich fühle stets, daß in
meiner Ökonomie keine rechte Berechnung liegt, man arbeitet aber
— man fühlt gleichsam eine Verpflichtung seinem Lande gegenüber.“
„Da will ich Euch noch etwas sagen,“ fuhr der Gutsbesitzer fort.
„Mein Nachbar, ein Kaufmann, war bei mir. Wir gingen das Land ab,
und den Garten. ‚Nein,‘ sagte er da, ‚Stefan Wasiljewitsch, bei Euch
ist alles in Ordnung, aber der Garten ist vernachlässigt.‘ Dabei
befindet er sich aber in vollständiger Ordnung. ‚Nach meiner Ansicht,
würde ich diese Linden anhauen. Man muß nur bis auf den Saft
schlagen. Es sind da ihrer tausend Linden, und jede von ihnen giebt
zwei gute Schaffe.‘“
„Für den Erlös daraus müßte er Vieh kaufen oder Land und es den
Bauern pachtweise verteilen,“ ergänzte Lewin, welcher offenbar
schon mehr als einmal mit ähnlichen Überschlägen zu thun gehabt
hatte. „Und er wird sich ein Vermögen begründen, während wir, Ihr
und ich, wenn Gott es giebt, nur das unsere zusammenhalten und
den Kindern zu hinterlassen streben müssen.“
„Ihr seid verheiratet, wie ich hörte?“ sagte der Gutsbesitzer.
„Ja,“ antwortete Lewin mit Stolz und Genugthuung. „Es ist aber
etwas Seltsames,“ fuhr er fort, „wir leben zusammen gerade wie die
alten Vestalinnen, und hüten unser Herdfeuer.“
Der Gutsbesitzer lächelte unter seinem weißen Bart.
„So ist es auch bei uns, da hat sich unser Freund Nikolay
Iwanitsch, oder jetzt Graf Wronskiy seßhaft gemacht, die eine
agronomische Industrie einführen wollen; aber das hat nicht weiter,
als bis zu Kapitalverlust geführt.“
„Aber weshalb machen wir es nicht, wie die Kaufleute? Warum
fällen wir nicht die Bäume im Garten der Rinde halber?“ sagte Lewin,
auf den Gedanken zurückgreifend, der ihn frappiert hatte.
„Nun, wie Ihr sagtet, man hütet sein Feuer. Dieses wäre ja auch
nicht ein adliges Verfahren. Unsere adlige Thätigkeit vollzieht sich
nicht hier, bei den Wahlen, sondern in unserem Winkel. Es giebt
auch einen Standesinstinkt bezüglich dessen, was man soll und was
man nicht soll. Bei den Bauern ist es ebenso; wenn ein Bauer gut ist,
so sucht er soviel Land zu pachten, als er kann. So schlecht dieses
nun sein mag, er pflügt es. Gleichfalls ohne Berechnung, und
geradezu zu seinem Schaden.“
„Wie wir“ — sagte Lewin. „Es ist nur außerordentlich angenehm,
Euch zu treffen,“ fügte er hinzu, indem er Swijashskiy zu ihm
herantreten sah.
„Ah, wir begegnen uns wohl zum erstenmal wieder, seit ich bei
Euch war,“ begrüßte ihn der Gutsbesitzer, „wir haben uns auch
verschworen.“
„Wie, schmäht Ihr neue Einrichtungen?“ frug Swijashskiy lächelnd.
„Ein wenig.“
„Man hat uns den Mut benommen.“
30.
Swijashskiy nahm Lewin unter den Arm und ging mit ihm zu
seinen Freunden.
Jetzt konnte Lewin Wronskiy nicht mehr vermeiden, welcher bei
Stefan Arkadjewitsch und Sergey Iwanowitsch stand und offen dem
herankommenden Lewin entgegenblickte.
„Sehr erfreut. Mir scheint, als hätte ich einmal das Vergnügen
gehabt, Euch begegnet zu sein — bei der Fürstin Schtscherbazkaja,“
sagte er, Lewin die Hand reichend.
„Ja; ich entsinne mich Eurer Begegnung recht wohl,“ sagte Lewin,
purpurrot werdend, und wandte sich sogleich, um mit seinem Bruder
zu sprechen.
Mit feinem Lächeln unterhielt sich Wronskiy mit Swijashskiy weiter,
offenbar ohne den geringsten Wunsch, in ein Gespräch mit Lewin zu
geraten; dieser hingegen blickte, mit dem Bruder redend,
unverwandt Wronskiy an und überlegte, wovon er wohl mit
demselben sprechen könnte, um seine Taktlosigkeit wieder
gutzumachen.
„Was verhandelt man jetzt?“ frug Lewin, Swijashskiy und Wronskiy
anblickend.
„Sjnetkoff. Er muß abschläglich antworten oder beistimmen,“
antwortete Swijashskiy.
„Nun, hat er denn beigestimmt oder nicht?“
„Darum handelt es sich ja eben; weder dies noch das ist der Fall,“
sagte Wronskiy.
„Und wenn er es verweigert, wer wird dann ballotieren?“ frug
Lewin, Wronskiy anschauend.
„Wer da will,“ sagte Swijashskiy.
„Werdet Ihr es thun?“ frug Lewin.
„Nur ich nicht,“ sagte Swijashskiy, in Verlegenheit geratend und
einen erschreckten Blick auf den neben ihm mit Sergey Iwanowitsch
stehenden, bissigen Herrn werfend.
„Nun wer denn; Njewjedowskiy?“ frug Lewin, im Gefühl, daß er
sich verwickelte.
„Dies wäre aber noch schlimmer. Njewjedowskiy und Swijashskiy
waren ja die zwei Kandidaten.“
„Ich werde es in keinem Falle thun,“ antwortete der sarkastische
Herr. Es war Njewjedowskiy selbst. Swijashskiy machte Lewin mit
demselben bekannt.
„Hat es auch deine Achillesferse getroffen?“ sagte Stefan
Arkadjewitsch, Wronskiy zublinzelnd, „das ist etwas nach Art der
Wettrennen. Man kann da eine Wette machen.“
„Ja; das berührt Einen bei der schwachen Seite,“ sagte Wronskiy.
„Und hat man sich einmal mit der Sache abgegeben, so will man sie
auch ausführen. Es ist ein Kampf!“ sagte er, stirnrunzelnd und die
starken Kinnbacken zusammenbeißend.
„Was für ein Kenner der Swijashskiy ist. Wie klar bei ihm alles ist!“
„Ach ja,“ versetzte Wronskiy zerstreut.
Ein Stillschweigen trat ein, während dessen Wronskiy so wie man
eben auf etwas Zufälliges blickt, auf Lewin, auf dessen Füße,
Uniform und Gesicht, schaute. Nachdem er die finster auf sich
gerichteten Augen bemerkt hatte, äußerte er, um doch wenigstens
etwas zu sagen:
„Wie kommt es denn — Ihr seid doch ständiger Dorfbewohner,
nicht aber Friedensrichter? Ihr seid ja nicht in der Uniform eines
Friedensrichters?“
„Das kommt daher, daß ich glaube, das Friedensgericht
repräsentiert eine thörichte Institution,“ antwortete Lewin finster, der
schon längst darauf gewartet hatte, mit Wronskiy ins Gespräch zu
kommen, um seine Taktlosigkeit bei Gelegenheit wieder
auszugleichen.
„Ich glaube dies nicht; im Gegenteil,“ sagte Wronskiy ruhig, aber
mit Verwunderung.
„Es ist doch nur eine Spielerei,“ unterbrach ihn Lewin. „Die
Friedensrichter sind uns nicht notwendig. Ich habe innerhalb acht
Jahren nicht eine einzige Klage gehabt, und was ich gehabt habe,
das wurde durch Ersatzleistung ausgeglichen. Der Friedensrichter
wohnt in einer Entfernung von vierzig Werst von mir. In einer Sache,
in welcher es sich um zwei Rubel handelt, muß ich dann einen
Vertrauensmann, welcher mich fünfzehn kostet, schicken.“
Er erzählte nun, wie ein Bauer einem Müller Mehl gestohlen habe,
und der Bauer, als der Müller es ihm mitgeteilt, gegen diesen
Verleumdungsklage eingereicht hätte. Alles das paßte nicht hierher
und war dumm; und Lewin fühlte dies auch, während er sprach.
„O, über dieses Original!“ sagte Stefan Arkadjewitsch mit seinem
mandelsüßesten Lächeln, „indessen gehen wir; man scheint zu
ballotieren.“
Sie gingen auseinander.
„Ich begreife nicht,“ sagte Sergey Iwanowitsch, den ungeschickten
Gang seines Bruders bemerkend, zu diesem, „ich begreife nicht, wie
es möglich ist, bis zu solchem Grade jeglichen politischen Taktes bar
zu sein. Das ist es eben, was wir Russen nicht haben. Der
Gouvernementsvorsteher ist unser Gegner, und du bist mit ihm ami
cochon und bittest ihn, zu ballotieren. Graf Wronskiy — ich mache
ihn mir ja auch nicht zum Freunde — hat mich zum Essen
eingeladen. Ich werde nicht zu ihm fahren, aber er ist auf unserer
Seite, weshalb soll ich uns deshalb einen Feind aus ihm machen?
Dann frägst du Njewjedowskiy, ob er ballotieren würde. Das geht
doch nicht an.“
„Ach, ich verstehe nichts davon! Alles das ist doch fades Zeug,“
antwortete Lewin mürrisch.
„Du sagst da, daß dies alles fades Zeug sei, befassest du dich aber
damit, so verwickelst du dich dennoch.“
Lewin blieb stumm und sie betraten zusammen den großen Saal.
Der Gouvernementsvorsteher hatte sich, obwohl er den ihm
bereiteten Verrat in der Luft liegen fühlte, und nicht alle ihn darum
gebeten hatten, gleichwohl entschlossen, zu ballotieren. Im Saal
wurde alles still, der Sekretär verkündete mit lauter Stimme, daß
Rittmeister der Garde, Michail Ljepanowitsch Snjetkoff zum
Gouvernementsvorsteher gewählt werden solle.
Die Kreisvorsteher kamen mit Tellern, auf welchen die Kugeln
lagen, von ihren Tischen zu dem Gouvernementstisch, und die
Wahlen begannen.
„Wirf rechts,“ flüsterte Stefan Arkadjewitsch Lewin zu, als er
zusammen mit dessen Bruder hinter dem Vorsteher zu dem Tische
schritt.
Lewin hatte indessen jetzt jenes Kalkul vergessen, das man ihm
erklärt hatte, und fürchtete, Stefan Arkadjewitsch möchte sich geirrt
haben, indem er sagte „rechts“. Sujetkoff war doch offenbar der
Gegner. Als er daher zum Kasten gekommen war, hielt er die Kugel
in der Rechten, überlegte sich jedoch, daß er irre, und nahm, dicht
vor dem Kasten, die Kugel in die linke Hand, um sie dann offenbar
links zu legen.
Ein Kenner der Sache, welcher an dem Kasten stand, und an der
bloßen Bewegung des Ellbogens erkannte, wohin jeder warf,
runzelte unwillig die Stirn. Er hatte keine Lust, seinen Scharfsinn
anzustrengen.
Alles war still geworden und man vernahm nur das Zählen der
Kugeln. Darauf rief eine einzelne Stimme die Zahl der Wähler und
der Nichtwählenden aus.
Der Vorsteher war mit beträchtlicher Majorität wieder gewählt
worden. Es erhob sich ein allgemeiner Lärm und man drängte nach
der Thür. Sjnetkoff trat ein, und der Adel umringte ihn, unter
Beglückwünschungen.
„Nun, jetzt ist es wohl zu Ende?“ frug Lewin Sergey Iwanowitsch.
„Es fängt eben erst an,“ versetzte für Sergey Iwanowitsch lächelnd
Swijashskiy; „der Kandidat des Vorstehers kann mehr Kugeln
erhalten.“
Lewin hatte dies vollkommen vergessen. Er entsann sich erst jetzt,
daß hier eine gewisse Feinheit verborgen lag, doch wurde es ihm
zuviel, sich darauf besinnen zu sollen, worin sie bestand.
Niedergeschlagenheit überkam ihn und er sehnte sich darnach, von
diesem Haufen wegzukommen.
Da ihn niemand beachtete, und er wie es schien, von niemand
vermißt wurde, begab er sich leise nach dem kleinen Saal, wo man
speiste, und fühlte große Erleichterung, als er die Diener wiederum
erblickte. Der alte Diener legte ihm die Speisenkarte vor, und Lewin
willigte ein. Nachdem er ein Kotelett mit Fasolen gegessen und sich
mit dem Diener über dessen frühere Herrschaft unterhalten hatte,
begab sich Lewin, der den Saal nicht wieder zu betreten wünschte,
in welchem es ihm so unangenehm war, auf die Tribünen.
Diese waren angefüllt von geputzten Damen, die sich über das
Geländer beugten, im Bemühen, nicht ein einziges Wort von dem zu
verlieren, was unten gesprochen wurde. Um die Damen herum
saßen und standen elegante Advokaten, Gymnasialschüler mit
Augengläsern, und Offiziere. Überall wurde von den Wahlen
gesprochen, und davon, wie der Vorsteher erschöpft sei, und wie
vortrefflich die Debatten gegangen wären; in der einen Gruppe
vernahm Lewin das Lob seines Bruders. Eine Dame sagte zu einem
Advokaten:
„Wie freue ich mich, daß ich Koznyscheff gehört habe! Da ist es
schon der Mühe wert, ein wenig zu hungern. Es war reizend! Wie
klar und verständlich alles! Bei uns im Gericht spricht niemand so.
Nur Maydel, und selbst der ist noch bei weitem nicht so
redegewandt.“
Als Lewin einen freien Platz an dem Geländer gefunden hatte,
beugte er sich darüber und begann Umschau zu halten und zu
lauschen.
Alle Adligen saßen in Spalieren, in ihren Kreisabteilungen. In der
Mitte des Saales stand ein Mann in Uniform, welcher mit klingender
lauter Stimme rief:
„Es wird ballotiert für die Kandidaten des
Gouvernementsvorstehers des Adels, Stabrittmeisters Evgeniy
Iwanowitsch Apuchtin!“
Totenstille trat ein, und man vernahm eine schwache
Greisenstimme:
„Ich verzichte!“
„Es wird ballotiert der Hofrat Peter Petrowitsch Bolj,“ begann
wiederum die Stimme.
„Ich verzichte!“ ertönte eine jugendliche pfeifende Stimme.
Nochmals ertönte das Gleiche und wieder erschallte das „ich
verzichte“; und so ging es eine Stunde lang fort. Auf das Geländer
gestemmt, schaute und lauschte Lewin. Anfangs wunderte er sich
und suchte zu erfassen, was dies alles bedeute; dann aber, nachdem
er sich überzeugt hatte, er könne nichts verstehen, fing er an, sich
zu langweilen. Als er sich hierauf all die Aufregung und Erbitterung,
die er auf den Gesichtern aller wahrgenommen, vergegenwärtigte,
wurde es ihm schwer ums Herz; er beschloß abzureisen, und ging
hinab.
Als er durch die Vorhalle der Tribünen schritt, begegnete er einem
auf- und niederschreitenden, bedrückt aussehenden Gymnasiasten
mit thränenschwimmenden Augen. Auf der Treppe begegnete ihm
ein Paar. Eine Dame, welche eilig auf den Absätzen lief, war es und
der gewandte Genosse des Prokurators.
„Ich habe Euch gesagt, daß Ihr nicht zu spät kommt,“ sagte der
Prokurator, gerade, als Lewin zur Seite trat, um die Dame
vorüberzulassen.
Lewin war schon auf der Ausgangstreppe und zog soeben aus der
Westentasche die Nummer seines Pelzes hervor, als ihn der Sekretär
abfing.
„Gestattet, Konstantin Dmitritsch, man ballotiert!“
Zum Kandidaten war Njewjedowskiy, der sich so entschieden
geweigert hatte, gewählt worden.
Lewin schritt zur Saalthür; sie war verschlossen. Der Sekretär
pochte, die Thür öffnete sich und er befand sich zwei Gutsbesitzern
mit geröteten Gesichtern gegenüber.
„In meiner Macht liegt es nicht,“ sagte der eine rotaussehende
Gutsbesitzer.
Hinter den beiden hob sich das Gesicht des
Gouvernementsvorstehers hervor. Dieses Gesicht erschien
furchterweckend mit seinem Ausdruck von Erschöpfung und Angst.
„Ich habe dir befohlen, niemand hinauszulassen!“ schrie er den
Thürhüter an.
„Ich habe nur eingelassen, Ew. Excellenz!“
„Mein Gott!“ schwer seufzend ging der Gouvernementsvorsteher,
müde in seinen weißen Pantalons, den Kopf gesenkt,
dahinschreitend, durch die Mitte des Saales nach dem großen
Tische.
Man hatte das Amt Njewjedowskiy übertragen, wie es auch vorher
geplant worden war, und dieser war jetzt Gouvernementsvorsteher.
Viele befanden sich in heiterer Stimmung, viele waren zufrieden und
glücklich, viele entzückt, viele unzufrieden und unglücklich. Der
Gouvernementsvorsteher war in einer Verzweiflung, die er nicht
verbergen konnte. Als Njewjedowskiy den Saal verließ, umringte ihn
die Menge, und folgte ihm begeistert nach, so, wie sie am ersten
Tage dem Gouvernementsvorsteher gefolgt war, als derselbe die
Wahlen eröffnet hatte, so, wie sie Sjnetkoff gefolgt war, als dieser
gewählt wurde.
31.
Der neugewählte Gouvernementsvorsteher und viele aus der
triumphierenden Partei der Jungen, speisten an diesem Tage bei
Wronskiy.
Wronskiy war einmal deshalb zu den Wahlen gekommen, weil es
ihm auf dem Dorfe langweilig geworden war, und er seine Rechte
auf Freiheit vor Anna geltend machen mußte, als auch zum Zwecke,
Swijashskiy mit seiner Unterstützung bei den Wahlen für alle
Bemühungen um Wronskiy bei den Semstwowahlen zu lohnen, und
vor allem deshalb, streng alle jene Pflichten der Stellung eines
Adligen und Gutsherrn zu erfüllen, die er sich auserwählt hatte.
Aber er hatte durchaus nicht erwartet, daß ihn diese Wahlen so
sehr interessieren, ihn so bei seinen Neigungen fassen würden, und
daß er der Sache so gewachsen sei. Er war in dem Kreise der
Adligen eine vollkommen neue Erscheinung, hatte aber offenbar
Erfolg und irrte nicht mit der Annahme, daß er bereits Einfluß unter
denselben gewonnen habe.
Zur Erringung dieses Einflusses unterstützte ihn sein Reichtum,
und sein vornehmer Rang, ein schönes Besitztum in der Stadt,
welches ihm ein alter Bekannter, Schirkoff abgetreten hatte, der sich
mit Finanzgeschäften befaßte und eine blühende Bank in Kaschin
besaß; ferner sein ausgezeichneter Koch, der vom Dorfe mit
hereingebracht worden war, dann seine Freundschaft mit dem
Gouverneur, der sein Kamerad, und zwar ein protegierter Kamerad
Wronskiys gewesen — vor allem aber sein einfaches, allen
gegenüber sich gleich bleibendes Wesen, welches sehr bald die
Mehrzahl der Edelleute veranlaßte, ihr Urteil über seinen
vermeintlichen Stolz zu ändern.
Er fühlte selbst, daß, mit Ausnahme jenes sonderlichen Herrn,
welcher an die Kity Schtscherbazkaja verheiratet war, und der ihm, à
propos de bottes, mit wahnsinniger Wut einen Haufen ungereimter
Dummheiten nachsagte, jeder Edelmann, mit welchem er sich
bekannt gemacht hatte, sein Anhänger wurde.
Er erkannte klar, und auch andere sahen dies ein, daß er zu dem
Erfolg Njewjedowskiys sehr viel beigetragen habe, und jetzt, an
seiner Tafel, verspürte er bei der Feier der Wahl Njewjedowskiys, die
angenehme Empfindung eines Triumphes über den Gewählten.
Die Wahlen selbst hatten ihn derart gefesselt, daß er, falls er im
Lauf der nächsten drei Jahre verheiratet sein würde, selbst daran
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com