Einstein’s Relativity

Fred I. Cooperstock Steven Tieu

•

Einstein’s Relativity
The Ultimate Key to the Cosmos

123
Dr. Fred I. Cooperstock Dr. Steven Tieu
University of Victoria Systemware Innovation
Victoria, BC Toronto, ON
Canada Canada

ISBN 978-3-642-30384-5 ISBN 978-3-642-30385-2 (eBook)

DOI 10.1007/978-3-642-30385-2
Springer Heidelberg New York Dordrecht London

Library of Congress Control Number: 2012942707

Ó Springer-Verlag Berlin Heidelberg 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief
excerpts in connection with reviews or scholarly analysis or material supplied specifically for the
purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the
work. Duplication of this publication or parts thereof is permitted only under the provisions of
the Copyright Law of the Publisher’s location, in its current version, and permission for use must always
be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright
Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.

Cover illustration by Travis Morgan, [email protected]

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

All physical theories, their mathematical
expressions notwithstanding, ought to lend
themselves to so simple a description that
even a child could understand them
Attributed to Albert Einstein by Louis de
Broglie in Nouvelles perspectives en
microphysique (trans. New York; Basic
Books, 1962), 184
In memory of Matt, Dani, and Jerry and
In honour of N. V. Tieu and V. Kha
Acknowledgments

We are grateful to our senior editor, Angela Lahee for her considerate guidance
and advice throughout the course of our progress. An anonymous reader affiliated
with the publisher carefully scrutinized an early draft of the manuscript and pro-
vided us with several useful points and corrections. Editorial assistant Claudia
Neumann kindly translated the text of the reader into English. We thank them both
for their kind efforts. As always, Peter Gary was a constant source of inspiration.
Ruth Cooperstock carefully examined this book for grammatical points and
provided us with valuable guidance as to explanations that would appeal to the lay-
reader. To the extent that we were able to do so, we followed her sage advice. Her
encouragement and devoted support were instrumental in seeing this work through
to its completion.

ix
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Preliminary Aspects of Classical Physics . . . . . . . . . . . . . . . . . . . 9

2.1 Newton and His Laws of Motion . . . . . . . . . . . . . . . . . . . . . 9
2.2 Frames of Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 Let There Be Light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4 The Whole Electromagnetic Spectrum . . . . . . . . . . . . . . . . . 14

3 A Simple Approach to Einstein’s Special Relativity . . . . . . . . . . . 17

3.1 Navigating in Spacetime . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2 The Relativity of ‘‘Now’’. . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Relativity Transformations for Space and Time . . . . . . . . . . . 26
3.4 The Twin or ‘‘Clock’’ Paradox . . . . . . . . . . . . . . . . . . . . . . . 28
3.5 Perceptions of Length and Time. . . . . . . . . . . . . . . . . . . . . . 32
3.6 More Paradoxes in Special Relativity . . . . . . . . . . . . . . . . . . 34
3.7 Exploring Spacetime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.8 Types of Intervals and the Light Cone . . . . . . . . . . . . . . . . . 42
3.9 Energy-Momentum—the Relativity Modifications . . . . . . . . . 47

4 Introducing Einstein’s General Relativity . . . . . . . . . . . . . . . . . . 51

4.1 What is Gravity? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2 Towards a New Theory of Gravity . . . . . . . . . . . . . . . . . . . . 57
4.3 Motion of Bodies in General Relativity. . . . . . . . . . . . . . . . . 59

5 Testing Einstein’s General Relativity. . . . . . . . . . . . . . . . . . . . . . 63

5.1 Motion of the Planets Around the Sun . . . . . . . . . . . . . . . . . 63
5.2 The Schwarzschild Solution . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.3 Other Classical Tests of General Relativity . . . . . . . . . . . . . . 69
5.4 Singularities and Black Holes . . . . . . . . . . . . . . . . . . . . . . . 72

xi
xii Contents

6 Gravitational Waves and Energy-Momentum . . . . . . . . . . . . . . . 79

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6.2 Gravitational Waves in Einstein’s Theory . . . . . . . . . . . . . . . 80
6.3 The Energy Issue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.4 Can Energy Be Localized? . . . . . . . . . . . . . . . . . . . . . . . . . 86

7 The Universe According to Relativity . . . . . . . . . . . . . . . . . . . . . 91

7.1 Getting Acquainted with the Elements of the Cosmos. . . . . . . 91
7.2 Evolving Views on the Universe . . . . . . . . . . . . . . . . . . . . . 92

8 Spacetime Diagrams for General Relativity . . . . . . . . . . . . . . . . . 101

9 The Motion of the Stars in the Galaxies . . . . . . . . . . . . ....... 103

9.1 The Dark Matter Paradigm: An Overview . . . . . . . . ....... 103
9.2 Galactic Dynamics: Newtonian and General
Relativistic Approach . . . . . . . . . . . . . . . . . . . . . . ....... 105
9.3 General Relativity Applied to the Observed Galactic
Velocity Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 108

10 The Motion of Galaxies in Galaxy Clusters . . . . . . . . . . . . . . . . . 115

10.1 Preliminary Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
10.2 Perceptions of Velocity: Free-Fall in Vacuum . . . . . . . . . . . . 116
10.3 Spherical Dust Collapse . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
10.4 Summing Up the Dark Matter Picture . . . . . . . . . . . . . . . . . . 119

11 Dark Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

12 Time Machines, the Multiverse and Other Fantasies . . . . . . . . . . 127

12.1 Closed Time-Like Curves . . . . . . . . . . . . . . . . . . . . . . . . . . 127
12.2 Topological Twists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
12.3 The Multiverse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

13 Concluding Commentary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

Appendix A: Proving That the Spacetime Interval is an Invariant . . . 135

Appendix B: Deriving the Einstein Field Equation . . . . . . . . . . . . . . . 137

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Chapter 1
Introduction

If you were to stop the first three people that you meet on the street and ask them what
they know of Einstein’s Relativity, you would likely get answers of the following
form:
“I don’t know anything about physics. But Einstein was this world-famous genius
with a big mop of hair, who wandered around in his slippers and crumpled sweatshirt.”
The second might get more technical and say: “I know: E = mc2 and this made
the atomic bomb.”
The third might reply with
“Oh, there is this rhyme:
There once was a girl named Bright/ Whose speed was much faster than light./
She set out one day/ In a relative way/ And returned on the previous night. It sounds
crazy but what do I know?”
Later, we will delve into the real story of the remarkable Ms Bright and her equally
remarkable sisters.
Since you are reading this book, you wish to know more about Relativity and
well you should. The essential aim of this book is to bring you, the reader, to a
level of appreciation of Einstein’s Relativity that you had never previously believed
was possible without a great deal of additional training. This goal is in line with
Einstein’s exhortation above, but at a necessarily higher level since our aim is more
ambitious.
Einstein’s Relativity has changed the face of physics and it is unfortunate that
so little is known about Relativity beyond the professional scientific community.
His theories of Relativity (yes, plural, there are two, there is the general theory
that blends into the special theory under the right conditions as we will discuss
later) embody science at its very best. They are bold, elegant, imaginative, logically
structured and encompass so much of what we do in physics. Above all for science,
the Special Theory of Relativity is so well tested as to be accepted almost universally
and the general theory, while far more limited in experimental verification, is almost
universally regarded as our very best theory of gravity. Part of our goal is to show you

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 1

DOI: 10.1007/978-3-642-30385-2_1, © Springer-Verlag Berlin Heidelberg 2012
2 1 Introduction

that Einstein’s General Relativity has an even greater degree of applicability than is
generally believed by many.
Most people have an intuitive notion of what physics is but cannot define it pre-
cisely. By definition, physics is the scientific study of the properties and interactions
of matter and energy. Given its extraordinary scope, physicists are often inclined to
wonder how anyone would not wish to study physics. We will take you through a
journey that will not even require any physics background to appreciate a goodly
part of this book, possibly even more. At the same time, if you still remember or
are willing to dust off the simple algebra that you had learned in junior high-school,
you will gain a still better appreciation of this gem of science. Insofar as possible, we
will use illustrations to bring out the essential ideas. For Special Relativity, diagrams
will make otherwise bewildering concepts crystal-clear. As well, such diagrams will
be used in a couple of suggestive forms for the curved space of General Relativity.
This will reveal in a simple picture, the big bang and the potential for a big crunch of
the universe. We will have succeeded in our task if your journey is pleasurable and
you come away from it feeling that you have a real appreciation of Einstein’s stellar
achievements.
We will begin with that other paramount genius in the entire history of physics,
Isaac Newton. We will discuss his laws of motion which form the foundation of
classical mechanics, laws which served us entirely for our primary endeavours up
until the twentieth century and continue to do so to this day.
As a lead-in to Special Relativity, we will take a brief excursion through the basics
of electromagnetism since electromagnetic radiation plays such a fundamental role
in Relativity as well as in our basic existence. Our approach to Special Relativity will
be that of H. Bondi who, under the influence of J.L. Synge, described the essentials of
Special Relativity in terms of “spacetime” diagrams. These diagrams are analogous
to the blueprints that are used in the construction of a building. Just as blueprints
map out pictorially the locations of rooms and fixtures in the space of a building, the
spacetime diagrams of Bondi map out the locations and the motions of observers and
light rays in spacetime. We will be referring to spacetime frequently as we proceed.
Spacetime is that essential new structure with which we will become familiar in
Relativity. In Relativity, space and time become unified in a very elegant manner.
However, it is not as if, in Relativity, that time is placed on the same footing as space,
a mistaken notion that many have come to believe. Space and time remain as different
concepts but they link in a new mathematical structure, a marriage if you will, that
we call “spacetime”, and with this union, we will see that time no longer carries that
absolute character that it had maintained in physics over the earlier centuries.
Bondi’s approach serves as an excellent introduction towards a full appreciation
of the nature of spacetime. With his approach, it becomes clear how observers in
relative motion perceive each other’s length and time intervals differently. As well,
the compounding of velocities and the transformation between the space and time
coordinates of events as perceived by observers in relative motion are easily derived.
Of particular fascination is the so-called “twin paradox”, the seeming contradiction
that arises when one of two twins takes off on a lengthy journey at very high velocity
and returns to find the stay-at-home twin having aged more than the traveler-twin.
1 Introduction 3

One is inclined to ask why it could not be viewed just as well as the stay-at-home twin
having made the journey and hence the ageing effect reversed, hence the paradox.
This, as well as a missing ingredient in the Bondi approach, will be elucidated for
you, the reader.
Since paradoxes in Relativity have been a source of both wonderment and confu-
sion for readers, we have paid considerable attention to them in this book. In addition
to the twin paradox, we examine two other paradoxes, each useful in its own way to
bring out aspects of Relativity. First, we consider a stick that slides along the ice at
high velocity and gets smacked when it lies over a hole in the ice. The stick is viewed
as shortened relative to observers at rest on the ice. Thus these observers are seen as
readily capable of getting the stick to become submerged under the ice. However the
observers who are riding on the stick see the hole in the ice as contracted and hence
the hole seemingly too short to enable the stick to fit. Second, we consider a rod that
is being carried at high velocity through the door of a barn and being trapped inside
the barn with the door closed when the rod is entirely inside. However, relative to the
rest frame of the rod, the barn appears shortened and hence the paradox arises as to
how the rod could have become trapped in a barn that is too short to accommodate
it. For both of these cases, we extract the reality of what actually occurs and how the
phenomena appear for each set of observers. Einstein’s Relativity does not lead us
to logical contradictions.
We will explore the interesting nature of energy and momentum in Relativity, how
there emerges a base level of energy for any given amount of matter that exists and
how enormous that energy content really is. It is the matter of the famous equation
E = mc2 . With this equation, we appreciate the actual unification of the concepts of
mass and energy in Relativity.
At that point, we introduce the crowning achievement of Einstein’s vision, his
theory of General Relativity, the incorporation of gravity into Relativity. Here we
emphasize that true gravity, produced by matter and fields such as electromagnetism,
is not a field as we have come to know other fields in the past but rather it is a manifes-
tation of the curvature of spacetime, a property of a different kind of geometry. This
is in line with J. L. Synge who took issue with the Equivalence Principle, the notion
that gravity is physically equivalent to an accelerated reference frame. In actuality,
it is only an approximate equivalence of effect and not a physical equivalence. This is
a frequent source of confusion and we take pains to clarify this important issue. We
then go on to develop the equation of motion of a free body in General Relativity, the
so-called geodesic equation. In this, we see the essential geometric character of the
theory of General Relativity, the linking of physical motion to the paths of extremal
distance in curved spacetime. We briefly develop the field equations for General Rel-
ativity, the equations to which we are guided by the fundamental physical laws of the
conservation of energy and momentum. Many have come to view these Relativity
equations as the most fundamental and the most beautiful in all of physics.
At this point, we turn to what is arguably the most important solution of the
Einstein equations, the Schwarzschild solution, which describes the gravity produced
by any spherically symmetric body. The solution is beautiful both in its simplicity
and in the interesting physics that stems from it. We discuss the important tests of
4 1 Introduction

General Relativity that arise from the application of the Schwarzschild solution. The
solution also leads us to the concept of singularities and the ever-popular black hole
that has seared itself into every-day parlance. We discuss the ideas surrounding the
latter.
In Newtonian physics, gravity has the curious property of making any changes
in its value, its strength and direction of action, realized instantly everywhere in the
universe if there is any redistribution of mass. This is usually referred to as “action
at a distance”. Einstein’s Relativity removes this property by virtue of the character
of the field equations of General Relativity. Their so-called “hyperbolic” structure
allows for gravitational waves, a spreading-out through spacetime of the spacetime
curvature, i.e. the gravity, at a speed of at most the speed c of light in vacuum.
This in turn is in accord with the fundamental restriction imposed upon information
flow in Special Relativity. Intuitively, this is very satisfactory. After all, it would not
seem sensible, for example, to imagine that observers in distant galaxies, through
our induced subtle local variations in gravity, could be aware of our waving our
hands with absolutely zero time delay. The flow of a wave of gravity carrying the
evolving information at speed c at most, is surely more intuitively acceptable than
the awkward and implausible aspect of instantaneous effects demanded by Newton.
While gravity waves have been studied over the years with great interest, in this
book we cover only the very basics. We also focus on an aspect of gravity waves that
ties in with an issue that has been problematic in General Relativity from its early
days, the issue of its energy and its localization. While the location of energy in the
rest of physics is fairly straightforward, in General Relativity, the issue of its location
presents new complications which have been approached in a variety of ways by
various researchers over the course of nearly a century. Some of the most prominent
researchers have taken the view that energy simply cannot be localized in General
Relativity while others have proffered that energy, in principle, must be localizable.
In this book, we focus on a particularly simple solution which is embodied in our
energy localization hypothesis, that energy, including the contribution from gravity,
is localized in the non-vanishing regions of the energy-momentum tensor. The latter
is the very familiar and well-understood mathematical embodiment of energy and
momentum from Special Relativity so our hypothesis is simplicity incarnate. We
present reasons for our favouring this hypothesis and we discuss what may be viewed
as a troubling consequence of the localization hypothesis, that waves of gravity would
not convey energy in the course of their propagation through the vacuum.
In Chap. 7, we begin our exploration of the universe according to General Rela-
tivity. We build our picture of the basic sizes of astronomical bodies in the universe
by the method of scaling; we imagine reducing the size of the Sun to that of the head
of a pin and then asking on this basis, how far away the planets as well as the nearest
stars are located. The process gives us a feeling for how the matter is distributed in
the cosmos, something that we cannot achieve simply by having the actual incredibly
enormous numbers of size and distance presented to us. We have no intuition for such
numbers as they are outside of our sphere of experience.
In that vein, and following the positive experience of relying upon spacetime
diagrams to help us appreciate the otherwise-confusing aspects of Special Relativity,
1 Introduction 5

we introduce spacetime diagrams for the evolution of simplified universe models in

General Relativity. This is a subject with scope for much further development.
Having reached this point, we focus upon a subject of particular interest to the
authors, the motions of the stars in the galaxies. According to Newtonian gravity, the
velocities of the stars, as tracked with increasing distance from the axis of rotation
of a galaxy, should be falling off in speed. Instead, the speeds have been found to be
generally fairly constant. This has puzzled astronomers and physicists alike as they
have believed (and most do continue to believe) that Newtonian gravity is an ade-
quate tool for the studies of such stellar motions. After all, the motions are generally
very slow by Relativity standards and the gravitational fields are generally weak,
the criteria generally regarded as quite sufficient for the applicability of Newtonian
gravity. As a result, researchers had concluded that vast stores of extra matter, now
designated as “dark matter”, must be present in huge halos surrounding galaxies in
order to provide the added gravitational boost to the stellar velocities. This currently
envisaged dark matter is not of the normal non-luminous matter of our every-day
experience but is rather believed to interact only gravitationally and possibly also
by the so-called “weak-interaction” process. In spite of intensive searches, no direct
evidence for the existence of this elusive dark matter has ever materialized. Never-
theless, probably the majority of researchers regard the existence of dark matter as a
given.
The present authors took a different approach to the subject. We noted that the
motion of bodies that are moving solely under the influence of gravity, often referred
to as “free-fall”, is a more complicated subject in General Relativity than has been
generally realized. As a result, we undertook a study of the modeling of a galaxy in
the form of an axially symmetric steadily rotating distribution of dust. The choice of
dust derived from the need to incorporate the free-fall aspect, dust being a pressure-
free fluid. We developed the Einstein equations to the lowest significant order for
weak fields and we were surprised to find that these equations were non-linear in
contradiction to the prevailing wisdom. This was a spur to us for further investiga-
tion. We fit the known data of stellar velocities for three galaxies and our own, the
Milky Way, to satisfy the field equations. This enabled us to determine the density
distributions and hence the masses for the four cases. As well as fitting the so-called
“rotation curves”, the plots of stellar velocities as a function of their distances from
the galactic rotation axis, we found that in each case, the galaxy masses were lower
than those predicted on the basis of Newtonian gravity. There was no need to invoke
dark matter to realize the motions. Very recently, we extended our approach to three
additional galaxies with the same result. In conjunction with this study, we describe
our serendipitous discovery, how Einstein’s General Relativity predicts the loca-
tion of the visible edge of the galaxies, now with seven cases, from the input of
those mysterious rotation curves, an achievement that is unattainable with Newton’s
gravity. This, together with the other remarkable successes flowing from Relativity,
underlines our conviction that Einstein’s Relativity is truly the ultimate key to the
cosmos.
6 1 Introduction

In addition to developing this study, we also discuss some of the more common
critical aspects that have been directed against our work. We then consider further
desirable follow-up work.
In Chap. 9, we turn to the earliest analysis suggesting the need for dark mat-
ter, the anomalously high velocity of entire galaxies, taken as units comprising an
ensemble, galaxies which were observed in the Coma Cluster of galaxies. Again, we
confronted the possibility that General Relativity might provide a means to describe
the phenomenon without dark matter. The actual physical source consists of a chaotic
distribution of galaxy velocities in a cluster. Since we do not yet have a means of
dealing with such chaotic systems within General Relativity, we turned to what we
do have available at present, namely an idealized system of smoothly continuous
matter falling radially with perfect spherical symmetry. As before, to incorporate the
free-fall aspect, we consider the collapsing matter to consist of dust. In doing so, we
incorporate two new aspects: exactness of solution and explicit time-dependence.
Again we find that General Relativity produces results that differ considerably from
those on the basis of Newtonian gravity. Most significantly, we are led to the conclu-
sion that actual Coma Cluster data for galaxy velocities can fit this idealized model
without resort to additional dark matter.
Apart from dark matter, a second different form of mysterious substance has
come to be conjectured as the overwhelmingly dominant part of the universe, the
so-called “dark energy”. The evident need for its existence derives from the current
conclusion that while the universe has undergone an expansion with the naturally
expected deceleration for approximately half of its present lifetime, the latter half of
the expansion has proceeded with accelerated expansion. The properly designed dark
energy can be invoked to ensure that this process is followed; yet as is the case with
dark matter, there is no realization of its form as an element of particle physics. As
well, when we consider its connection with quantum field theory where the vacuum
assumes a very active life of its own, we find an absurd requirement forced upon the
theory to accord with the observations. This is discussed in Chap. 11.
In Chap. 12, we wind up our journey through Einstein’s Relativity with con-
sideration of what might be properly called “fantasy-land”. Over the years, some
researchers have seen in Relativity the potential for the realization of time-travel.
A person takes a journey through spacetime in such a manner as to return to his
or her past: Science-fiction thrillingly becoming science-fact. In this, we emphasize
that the process has entailed mathematical choice of the identification of points in
spacetime rather than actual physical necessity. In other words, one can make time-
travel happen in the free-wheeling world of mathematics but the physical world, the
world of reality, chooses it not to happen. It does so to preserve the cause-effect
relationship between real events, a constraint which mathematics need not honour.
A recent popular subject is the “mutiverse”, the idea that there exist universes
beyond our own (even though the word “universe” has always referred to all that
exists!) In our view, this is not an element of real science and however entertaining,
should not be seen as anything other than fantasy. We will explore this idea and some
of its byways.
1 Introduction 7

We also discuss briefly the idea of spacetime worm-holes and the speculations
that these have engendered.
Finally, in Chap. 13, we consider what may lie ahead as we work our way towards
further understanding of our remarkable universe.
The reader will discover that there are a fair number of equations (and a few
more-advanced equations) interspersed through the book. These have been placed
for those readers who have more of a mathematical training than the average citizen.
However the equations (as well as the appendices and the section on localized energy
in Chap. 6) can be safely ignored by the more general reader who, hopefully, will
still be able to achieve a solid understanding of the basic ideas without them.
Chapter 2
Preliminary Aspects of Classical Physics

2.1 Newton and His Laws of Motion

Before we get to Einstein’s Relativity, we should start with Isaac Newton, who was
another genius and also, in his own way, a relativist. Many, including ourselves, feel
that Newton and Einstein were the greatest physics geniuses of all time. We constantly
contemplate with awe what these two giants of insight and imagination have achieved,
how dominant have been their respective contributions to the evolution of physics.
Our focus will be on Einstein but it was Newton who most significantly set the
stage for what was to come. No physics treatise can logically skip over him. Besides
developing, in parallel with Leibniz, the indispensible mathematical discipline of
calculus, Newton formulated three important Laws of Motion and actually, he had
his own Principle of Relativity. These motion laws are very simple:
1. A body at rest remains at rest and a body in motion, continues in motion in a
straight line with constant speed unless it is subjected to an unbalanced force.
In other words, left undisturbed, bodies keep moving uniformly in the way they
have been moving. In fact, they will do so even if there are forces (pushes and pulls)
on them as long as any given force is compensated for by an equal and opposite
force, restoring the balance. But if there is a force on the body that does not have a
balancing partner, Law 2 comes into play:
2. If an unbalanced force acts upon a body, it is accelerated, i.e. it changes its
speed and/or its direction of motion. The extent of the acceleration is proportional
to the strength of the unbalanced force and the direction of the acceleration is in the
direction of the unbalanced force.
In other words, if forces on a body do not balance out, the body will not continue
to move as it had been moving prior to being subjected to the unbalanced force.
Rather, it will be accelerated, i.e. it will speed up or slow down and/or change its
direction of motion, depending on which way it had been moving beforehand and
which way the unbalanced force was directed. Physics students remember this law
in equation form,

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 9

DOI: 10.1007/978-3-642-30385-2_2, © Springer-Verlag Berlin Heidelberg 2012
10 2 Preliminary Aspects of Classical Physics

F = ma (2.1)

with F the unbalanced force, a the acceleration and the proportionality factor m is
the mass of the body. Notice that for a given F, the bigger the m, the smaller the a.
This is right in line with our everyday experience: the more massive a body, the less
it will get going for a given push.
You might get confused about these two laws if you do not realize that friction is
another force and it’s hard to get rid of friction. But imagine a hockey puck sliding
on a sheet of very smooth ice, approximating the elimination of friction. The puck
just keeps going at a (nearly) steady speed in the same direction; the smoother the
ice sheet, the further it goes. However, if you were to give the puck a solid smack
with the hockey stick, it would speed up dramatically, and the harder the smack, the
more it would speed up. Of course it does so in the direction at which it had been
smacked. These laws are part of our everyday experience and make excellent sense.
The Third Law is a bit more subtle:
3. For every action, there is an equal and opposite reaction.
If you push against the wall, the wall pushes directly back against you: action, the
push, reaction, the push back. It might sound strange to think that a wall can push
you but it must. The friction force on your shoes is pushing you toward the wall
because you are pushing backwards on the floor (again a case of action and reaction)
and since you are not moving at all, the wall must be pushing you in the opposite
direction to keep the forces balanced (Law 1). So it all makes sense.
These laws plus Newton’s law of “universal gravitation”, that every body in the
universe attracts every other body with a force F that is proportional to the product
of their masses m and M and inversely proportional to the square of the distance r
between them, in equation form

F = Gm M/r 2 (2.2)

(G is the constant of universal gravitation) are the bases of the important part of
physics called “classical mechanics”. It governs much of our everyday physical life-
experience.

2.2 Frames of Reference

Of course the application of Newton’s laws requires measurements. The measure-

ments to which Newton’s laws apply occur within frames of reference called “iner-
tial”. A frame of reference is a grid to which we can ascribe the positions of bodies at
any given times.1 In other words, it is Newton’s laws that pick out inertial reference
frames. A reference frame is not inertial if it is accelerated, for example a frame of

1Imagine a set of three mutually perpendicular sticks with markings at one centimeter intervals.
This constitutes a reference frame. A particle’s position is specified by its three-number set of
position coordinates relative to this frame.
2.2 Frames of Reference 11

reference that is anchored to a rocket-ship while its rockets are propelling it to higher
and higher speeds. Newton’s laws do not hold in such a reference frame. After all,
if you were to gently release a coin in such a frame, it would not stay put but rather
it would move with accelerated motion in the direction of the rockets. Since there
would not be any force acting on this released coin, its accelerated motion relative
to the rocket-ship frame of reference would be in violation of Newton’s First Law.
Newton’s Principle of Relativity states that the dynamical laws of physics are the
same relative to all inertial reference frames. The word “all” leads us to ask how
we can generate more inertial reference frames, once we have determined a single
inertial reference frame. This is easy: from one, we can create as many as we wish by
simply moving with respect to the first with a constant velocity2 (motion at a constant
speed in a straight line of specified direction). The laws hold relative to these frames
as well because bodies that are not subjected to unbalanced forces will still continue
to move at constant velocities relative to these frames; all that happens is that each
original constant velocity will just change to a new constant velocity as a result of
the transformation.
In addition to the grid, we have to bring into the discussion a set of clocks to
be able to determine where any given body is located as well as when it is there.
For example, from a given starting point (let us say your position), the body could
be 3 m to the west along the east–west line, 5 m to the north along the north–south
line and 14 m up along the vertical line. Suppose it is there at a time 10, the time
you set your clock to start. Those four numbers really pin it down. We would write
the body’s coordinates in time (the first number) and space (the second, third and
fourth numbers) as a bracketed unit (10, 3, 5, 14). Physicists call this set of four
numbers describing the body’s position in space and time an “event” even though, as
events go, this one is not really all that dramatic. It is an event in the combination of
space and time we refer to as “spacetime”. In our experience, it is four-dimensional
spacetime, three dimensions of space and one dimension of time; we do not perceive
any dimensions beyond these.3 It is worth pondering for a moment why this is the
case. Why do we not live in a universe with one spatial dimension? With two spatial
dimensions? With four, five, six, etc. spatial dimensions? Some have even considered
the notion of more than one time dimension.
If the body just stayed there relative to our grid, the first number 10 would change
(ignoring smaller uncrements) to 11, 12, 13, etc. as the clock ticks away while the
other numbers would remain the same. However, if the body were to be in motion
relative to the grid, all of the numbers could be changing continuously as we trace
out the body’s history in spacetime. This grid, populated by clocks, constitutes an
inertial frame of reference if Newton’s laws of motion hold true relative to this frame.

2 In physics, the word “velocity” takes on a very specific meaning; it has two aspects, speed that
only tells you how fast the object is moving and second, direction, which way it is headed at that
instant. It is indicated in physics diagrams by an arrow pointing in the motion direction with length
of arrow indicating the degree of speed.
3 Some researchers have dealt with higher-dimensional spacetimes but primarily with having the

extra dimensions curled up into such an incredibly small size that they are beyond our ability to
detect them.
12 2 Preliminary Aspects of Classical Physics

It is worth repeating: it would not be inertial if the grid were attached to a rocket
ship while the engines are blasting away. As well, it would not be inertial if it were
rotating. Such frames are accelerated. Accelerated reference frames are called “non-
inertial” and Newton’s laws do not hold true relative to such frames. Also, it is worth
repeating the example for emphasis: relative to the rocket ship frame, if you were
to release a coin from rest inside the rocket ship, it would not stay put as Newton’s
First Law would have it but rather, it would move with acceleration in the direction
to which the gases are being propelled away from the ship, just as if it had been
dropped from rest as you stand on earth.4

2.3 Let There Be Light

We now have some of the basics that we need. The next step is to bring into the
discussion the very mysterious and wondrous thing we call “light”. Wondrous indeed!
Imagine a universe without light. How fortunate we are to have eyes that permit us
to partake of all the beauty that surrounds us and how fortunate we are for the
protection that light provides for us from the many potential hazards that we face.
It is even more spectacular when we probe into its essential nature. Newton had
regarded light as a stream of “corpuscules”, much like a hail of miniscule pellets,
while Huygens claimed that light was really a wave. Curiously, in modern physics,
we regard both Newton and Huygens as being right! Light has a dual nature, a
particle aspect and a wave aspect. In modern parlance, Newton’s corpuscules are
called “photons”. The wave aspect of light is most apparent in the many experiments
that have been performed to probe its character.5 But what do we know of waves?
There are sound waves, waves that propagate in air and through water, generated
by chains of molecules set into vibration. There are water waves in the seas and the
oceans, there are waves traveling on a vibrating string. Common to all the familiar
waves is the presence of a medium, some form of matter to “wave” the propagating
disturbance. Physicists had reasoned that since we see the light reaching us from the
distant stars and since the atmosphere above the earth trails off rapidly in the near
emptiness of interstellar space, there must be some very rarified residual medium
present to support the “waving” of light. This supporting medium must be so rarified
that we cannot detect its presence in the obvious ways that we detect the media for
the common waves that we mentioned. This medium was given a name: The “ether”,
and much of the then world of science was just as confident of its existence as it was
of the existence of the traditional media, molecules, bulk water and string material
respectively in the previous discussion. It is well to remember the ether when we
convince ourselves that some form of matter must exist to satisfy some cherished

4In fact this phenomenon will serve as part of our lead-in to an understanding of General Relativity.
5 Other experiments bring out the dual particle character of light. See the photo-electric effect
discussed later.
2.3 Let There Be Light 13

theory. The theory, however seemingly inevitable, necessary and wonderful, might
nevertheless be wrong. It is nature that decides.
In this regard, we must remind ourselves constantly that an essential element of
any science is experimental verification. Michelson and Morley set out to detect this
ether in an ingenious experiment. They reasoned that as the earth moves through the
ether, an ether wind, however subtle, must be created. With the earth moving at speed
v through the ether, light would have a retarded upstream speed c − v as it plows its
way against the ether flow and a boosted speed c + v if it is directed downstream
from the ether flow. This is just like a person swimming against and with a current in
a river: watching the swimmer’s progress from the bank, we would see the swimmer
advancing more speedily in the second case than in the first.6
However, if light were directed perpendicular to the ether flow, it would not be
retarded or boosted but just partake of the flow. Michelson and Morley arranged
to have light travel in both the perpendicular and ether flow directions for equal
distances and be reflected back to their origin, thus allowing the light to interfere in
recombination.
Their experiment was not designed to measure the speed of light but rather to
measure differences in light speed along different directions or paths. Under the
assumption that the net speeds would differ in the different directions of travel through
the ether, the light beams would constructively and destructively interfere with each
other.7 Try as they could, they were never able to detect the expected effects of the
interference that would be present, had the light been traveling through an ether
wind. Following this effort, Lorentz and Fitzgerald had hypothesized that bodies
somehow underwent a physical contraction in the direction of motion when they
traveled through the ether and hence the path length would be altered in just the right
amount to remove what would have been the effects of the interference. However,
many viewed this as a contrived solution to the puzzle and as we view this idea today,
the Lorentz-Fitzgerald rationale appears even more naïve. In more recent times, the
equivalent of the Michelson-Morley experiment has been performed using lasers,
enabling far greater accuracy and the null result continued to hold.
It took the genius of Einstein to see the apparent contradiction in a totally different
light (if the pun may be excused). He reasoned that there is no ether and that light
has the speed c whether one measures it in any direction relative to the motion
of the earth in its orbit around the Sun. Einstein’s bold hypothesis: the speed of

6 As an interesting extreme example, if the swimmer’s speed relative to the river were the same as
the speed of the river relative to the bank, for the observers on the bank the swimmer would appear
to undergo no advancement at all when he swims upstream. His arms and legs would be flaying but
he would be going nowhere relative to the observers on the river-bank!
7 To get a simple picture of constructive and destructive interference, consider how one “interferes”

with the motion of a child on a swing in the playground. To provide (maximum) “constructive”
interference, that is to make the swing go higher and higher, one provides a push on the swing at
the point in each cycle when the swing is at its highest point, in phase with the motion. By contrast,
one renders “destructive” interference when one provides a push when the swing is still in motion
towards the position of the pusher, that is out of phase with the motion. The latter has the effect of
damping the swing motion.
14 2 Preliminary Aspects of Classical Physics

light in vacuum is always c. The combination of Newton’s Principle of Relativity,

that the dynamical laws of physics are the same in all inertial frames of reference
plus Einstein’s hypothesis of the universality of the speed of light and extending
the equivalence of inertial frames to all physical laws is generally referred to as
Einstein’s Principle of Relativity. The additional constancy of c hypothesis forces
dramatic changes in physical theory. While it becomes somewhat more palatable to
accept his hypothesis in conjunction with the idea that light can propagate in the
absence of any medium, it nevertheless remains counterintuitive. After all, suppose
we return to the corpuscular or quantum picture of light: little balls of energy moving
at speed c. If these “quanta” (recall that they are also called “photons” when referred
to light) were instead a stream of baseballs moving at speed c, then even without a
medium, if we were to run toward the balls with speed v, we would measure their
speed as c + v relative to us and if we were to run in the opposite direction, we would
measure a speed c − v. Relative to our catching the balls when we stood still, there
would be a greater sting in our baseball glove when running towards the direction
of motion of the balls and a reduced sting in the opposite direction. It is a reflection
of the different energies of motion. There is a corresponding effect with light that
shows itself as a shift in frequency rather than in speed of the light quanta (photons).
The truly remarkable, totally counterintuitive character of light is that it always
has the same speed c in vacuum, regardless of the motion of its source or the motion
of the observer. We do not see it as “speeded up” by running towards its source as
we do with baseballs. Now since it is an experimental fact that science has never
measured anything other than light8 moving even at the speed c, let alone faster
than c, and since we cannot boost the speed by running toward it when the speed
in question is c, we conclude that according to Einstein, there is a “speed limit” for
bodies and waves in nature, namely the speed c. The existence of this speed limit can
also be appreciated from the fact that it would require an infinite amount of energy
to bring a body of mass m to the speed of light, as we will see in Chap. 3.9
In turn, since all inertial reference frames are physically equivalent, that speed
limit had better be the same in all the frames. It is understandable that many resisted
Einstein’s bold hypothesis when it was first proposed. Some of the consequences
that derive from his hypothesis are even more astounding, as we shall explore in the
coming chapters.

2.4 The Whole Electromagnetic Spectrum

We have discussed an essential feature of light but there is a much bigger picture
to discuss. Electromagnetism is of great importance to physics in general and to
Relativity as well as to our lives, of course. In what follows, we will take you through
some of the basics, a very brief guided tour of the electromagnetic spectrum.

8 More generally, the family of electromagnetic waves, of which light is a member.

9 Whether this limit also applies to the speed at which information can be propagated becomes more
involved in the strange world of quantum mechanics, as we will discuss later.
2.4 The Whole Electromagnetic Spectrum 15

Light is just a small slice of the very broad electromagnetic spectrum with elements
that you have heard about many times: radio waves, microwaves, infrared rays,
ultraviolet rays, X-rays, and gamma-rays. The waves that allow us to see with our
eyes, the light rays, fit into this list between the infrared and the ultraviolet rays.
All electromagnetic waves are in the same family, composed in the same manner
of oscillating electric and magnetic fields. They all travel at speed c in vacuum.
The members are distinguished from each other by their wavelengths with the radio
waves having the longest wavelengths (and therefore the lowest frequencies),10 and
gamma-rays having the shortest wavelengths (and the highest frequencies).
While electromagnetic waves reach us from the Sun, the stars, the distant galax-
ies, the quasars, the pulsars and from the deepest recesses of space, we can (and
do) produce them by our own actions and our modern technologies. At the sim-
plest level of our everyday experience, we produce them when we strike a match.
At the more fundamental level, we produce an electromagnetic wave when we wiggle
(i.e. accelerate) a charge. Accelerated charges produce electromagnetic waves. Imag-
ine the wave propagating outward from the accelerating charge at the speed of light.
As we wiggle it faster and faster, we observe some interesting results.
Suppose for example, we were able to wiggle the charge at a frequency of 530–
1700 kHz,11 the wavelength would be up to 600 m long and we would produce noise
on the AM radio. A typical AM radio has the following numbers on the dial: 530–
1700 kHz. Now, if we were to shake the charge with a greater amplitude (while
maintaining the same frequency), we would produce a louder sound on the radio.
This process is called Amplitude Modulation or AM. We do not have any mechanical
device to wiggle a charge that quickly. Instead, radio waves are typically produced
by sending charges up and down a very tall antenna (100–200 m tall).
At a rate of oscillation of 88–108 MHz, we would produce noise on the FM radio.
In FM broadcasting, the loudness of the sound is not controlled by the amplitude but
by the frequency. For example, if we increase our “wiggle frequency” from 99.900
to 99.915 MHz, we would progress from a quiet sound to a loud sound. This varying
of frequency is called frequency modulation or FM.
At 900 and 2400 MHz we enter the spectrum of cordless phones and “wifi”
internet connection respectively. Cellphone signals reside at 800,850 and 1900 MHz.
VHF(very high frequency) signals have frequencies in the range 30–300 MHz and
UHF (ultra-high frequency) signals span the range 300–3000 MHz for television
broadcasting.
Microwaves reside at 3.1011 −1013 Hz or 300–30000 GHz. Producing microwaves
by sending electrons up and down an antenna at such incredibly high frequencies
is not physically possible. Instead, microwaves are generated by sending electrons
around a warped circle within a vacuum tube called a magnetron. One can think of

10 Frequency, or cycles per second, with symbol ν is inversely proportional to wavelength with
symbol λ and the product of the two is c, or in equation form, λν = c. It can be seen as follows: λ
is the advance in distance per cycle and ν is the cycle advance per second, hence the cycles cancel
in the product leaving distance advance per second or velocity which is c.
11 A kilohertz, kHz, is a thousand cycles (or wiggles) per second. A megahertz, MHz, is a million

cycles per second. A gigahertz, GHz, is a billion cycles per second.

16 2 Preliminary Aspects of Classical Physics

it as a particle accelerator. As the electrons weave through cavity, their oscillations

emit microwaves.
Proceeding to higher frequencies, at the bottom end of the optical spectrum, in
the region of 4.1014 Hz, infrared and light of red colour are next to each other. At the
higher end of the optical spectrum in the region of 8.1014 Hz reside the colours blue,
violet and the invisible ultraviolet rays. The energy quanta of each ultraviolet photon
is much, much greater than, for example, the microwave photons. One can see the
dramatic difference of effects produced by different types of photons in the “photo-
electric” effect: if we shine red light (or microwaves) onto a metallic surface in a
vacuum, no electrons would be ejected, no matter how many lamps we use. However,
if we were to shine ultraviolet light onto the surface, no matter how dim, electrons
would be ejected. This is a prime example of a quantum phenomenon.12 At still higher
frequencies, 1017 −1020 Hz, we encounter X-rays. Beyond X-rays in the spectrum,
we have gamma-rays at 1020 −1024 Hz. We can only produce them using particle
accelerators or observe them directly through nuclear decay in radioactive materials
or as cosmic rays arriving on Earth from outer space.
There may be yet more powerful radiation but the most energetic phenomena that
we have detected in the universe are gamma-ray bursts. These were discovered by
military satellites in 1967 during the Cold War in the course of detecting nuclear
weapons testing. Gamma-ray bursts are so powerful that if one were to occur within
our galaxy, it could cause mass extinction on Earth. Fortunately, it is estimated that
their frequency of occurrence within our galaxy is as low as one per million years
and even if one were to occur within our galaxy, it is unlikely to be pointing directly
at the Earth. Therefore we would not advise you to ask your insurance agent for
gamma-ray-burst insurance.

12
Einstein was awarded his one and only Nobel Prize for his explanation of the photoelectric effect.
Many have argued that he should have been awarded at least three Nobel Prizes.
Chapter 3
A Simple Approach to Einstein’s Special
Relativity

3.1 Navigating in Spacetime

Your everyday experience is governed by the “where” and the “when”. If you arrange
to have lunch with your friend, you might tell him, “Meet me at Joe’s Diner at
noon tomorrow.” In your mind, you know the location of Joe’s Diner in the two-
dimensional grid of your street map, let us say the corner of 53rd Street and 6th
Avenue and if the diner is on the third floor, you know the height that you have to
climb in the vertical direction. He knows it and you know it. And your watch tells you
when to be there—the same for your friend. We are back to those four numbers again;
we have the event of meeting the friend, where and when to do so. A presupposition
is that there is no ambiguity about clocks, that there is a universal time, an absolute
time (assuming that we all use good watches). We will see that Relativity changes
all that.
The first author was fortunate to have spent a year in Dublin at the Institute
for Advanced Studies under the directorship of J. L. Synge. Synge’s heritage was
remarkable. His uncle, J. M. Synge, was one of Ireland’s greatest writers and nephew
J. L., following in the family tradition of distinguished achievement, was a man of
both letters and science. Of particular interest for us here is Synge’s emphasis on
the use of spacetime diagrams. By means of a graph, a picture emerges in the mind
that words alone or banks of data alone cannot equal. We know this from our own
experience. If we were to present lengthy tables of data giving the price of General
Motors stock over a fifty year period, it would not have the same impact as the
simple graph that dramatically illustrates in a single glance, the rise and fall of
this corporation, as shown in Fig. 3.1. Here we have one dimension, the stock price,
plotted against the other dimension, time.1 It is even more helpful in Relativity where
the mathematics can easily obscure the essentials.
Bondi [2] was influenced by Synge and he used spacetime diagrams to illustrate
how one connects the observations of various “inertial observers”, observers who are

1A continuation of this plot through to 2012 would reveal at a glance the significant recent recovery
of General Motors.

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 17

DOI: 10.1007/978-3-642-30385-2_3, © Springer-Verlag Berlin Heidelberg 2012
18 3 A Simple Approach to Einstein’s Special Relativity

Fig. 3.1 The rise and fall of GM Stock Prices

GM stocks from 1970 to 2008 100
90
80
70
60

Price
50
40
30
20
10
0 1975 1980 1985 1990 1995 2000 2005
Year

not in a state of acceleration. We will use Bondi’s method and complete an element
missing in Bondi’s work for the resolution of the famous “twin” or “clock paradox”.
Our world is one of three spatial dimensions and time, which for mathematical
purposes is another dimension. Ideally, to illustrate our four-dimensional world, three
space dimensions plus one time dimension, we would draw four-dimensional graphs
and if we lived in five dimensions, we could do so. However, since we inhabit four-
dimensional spacetime with only three (macroscopic) spatial dimensions, at best we
can only draw in three dimensions and even this is difficult to portray suggestively on
a two-dimensional sheet.2 Therefore, we restrict ourselves to plots in two dimensions
with only one space dimension that we call x and, analogous to how we plotted the
General Motors stock price, we use the second (space) dimension on the page to
represent the one time dimension t. In this manner, we capture the essentials of the
spacetime view. The other two space dimensions y and z are suppressed. It is a
one-dimensional spatial world view, not the complete real world but useful for our
purposes. Thus, observers move back in forth on a single common x line and we will
imagine that they are beaming light to each other along this line.
We now get back to Ms Bright, whose first name is Beatrice. We first met her in the
fantasy-rhyme in the Introduction. Actually Beatrice (“B”) is one of a set of triplets
and has sisters Alicia (“A”) and Camela (“C”). All three Bright women are inertial
observers. We draw our plot from the viewpoint of Alicia who is always at rest at
position x = 0. The vertical t axis marks off the time as read by Alicia. Therefore the
path that she follows in spacetime is the vertical line, labeled A. Suppose Beatrice,
with path labeled B, moves at constant speed v relative to Alicia and kisses Alicia
“hello” and “goodbye” while both Alicia and Beatrice have their clocks read 0,
i.e. when they cross paths for an instant. This is at the origin O with spacetime
coordinates (t, x) = (0, 0) in the two-dimensional spacetime plot, Fig. 3.2. Since
Beatrice is moving at a steady speed, she advances in x proportionally to the time

2 As an example of how we do so, see Fig. 5.2. The guidelines below the Sun-planet picture are
used to give a perception of depth.
3.1 Navigating in Spacetime 19

Fig. 3.2 The paths of Alicia A B

and Beatrice in spacetime.
The picture is drawn from
the perspective of Alicia who
does not advance in position x.
She only advances in time,
the vertical axis. Beatrice
advances in both x and t from
Alicia’s perspective

O x

that elapses. Thus Beatrice’s trajectory in spacetime is the straight line at a slope
relative to the vertical that is Alicia’s path in spacetime. If Beatrice were to move
faster, her trajectory would be a line tilted further away from the t axis, i.e. greater
advance in x for a given amount of time t.
Now we bring light into the picture. Since light with speed c has the highest
possible speed of propagation, its trajectory plot has the largest possible inclination
with respect to the vertical t axis.3 Suppose Alicia turns on her flashlight beaming
it at Beatrice for a time period T . Clearly Beatrice receives the light for an amount
of time that is proportional to the amount of time that Alicia has beamed at her;
the longer Alicia sends, the longer Beatrice receives the light. At this point, we do
not know how the proportionality factor relates to the speed with which Beatrice is
pulling away from Alicia so, for now we will just give it a symbol k. It is called the
“relativistic Doppler factor”.4
So Alicia transmits for a period T , Beatrice receives for a proportional period kT .
Now there is nothing special about Alicia as compared to Beatrice; they are both
inertial observers. The physics is identical for both of them. Therefore, were Beatrice

3 To determine the degree of inclination, we note that in a time t, light advances a distance x equal
to the speed times the time, i.e. x = ct. Now relativists like to use units in which c is taken to be
1. It simplifies calculations and c is easily restored at the end if so-desired. Thus, with x = ct, we
have x/ct = 1 or x/t = 1 using c = 1, the advance in x equals the advance in t for light. As a
consequence, for our plots, all of the light rays are seen to be at 45◦ with respect to the t axis.
4 The Doppler effect, the change in measured frequency as a result of source or observer motion,

is familiar in our every-day life experience. We hear the steeply rising pitch from the whistle of a
train and the siren of an emergency vehicle as they approach closer towards our position followed
by the marked drop in pitch as they move away from us. The pitch is dependent upon the velocity
of the source.
20 3 A Simple Approach to Einstein’s Special Relativity

Fig. 3.3 Alicia shines light A B

to Beatrice for a period T
and Beatrice receives for a
period kT and vice-versa.
Note that the slopes of the
kT
light trajectories are more
inclined away from the vertical
axis than that of Beatrice’s
trajectory because light always
travels faster than humans. T
Note that in this, as in the
figures to follow, to avoid
clutter, we will draw only the kT
first and last rays of a beam
transmitted over a period of T t
time
O x

Fig. 3.4 Alicia beams light A B

to Beatrice for a period T
and receives light back from
Beatrice for a period k 2 T .
Beatrice receives light for a
period kT and sends light
back to Alicia for the same P
period kT T*

k(kT)

kT

to shine her flashlight to Alicia for a period T according to her (Beatrice’s) clock,
Alicia must also receive the light for a period kT (Fig. 3.3).
Now let us do an experiment that allows Alicia to determine Beatrice’s speed
relative to herself. As soon as they cross paths at O, let Alicia begin beaming her
flashlight towards Beatrice and shut it off when she reads the time T . Let Beatrice
also beam her flashlight back towards Alicia beginning at O and continue to do so
for as long as she receives light from Alicia. Thus, Alicia beams for a period T by
her clock and Beatrice receives the light for a period kT by her (Beatrice’s) clock.
Beatrice is beaming back to Alicia for a period kT so Alicia receives the return beam
from Beatrice for a period k times the emission period or k(kT ) = k 2 T (Fig. 3.4).
3.1 Navigating in Spacetime 21

Now Alicia is very clever and studious, reading physics books all day, so she
comes up with a plan to determine that unknown k factor with the knowledge of
how fast Beatrice is traveling relative to herself. She reasons that since the motion of
Beatrice is uniform, Beatrice must have been at the position P in Fig. 3.4 when she
was sending her last ray of light to Alicia at a time T ∗ . This is in the Fig. 3.4, half
way between the time that Alicia sent her last ray to Beatrice and the time the last
ray came back to herself (Alicia), in other words, at the time (k 2 T + T )/2. This is
how long it took for Beatrice to reach that distance by Alicia’s reckoning.5
If Alicia could now deduce how far away Beatrice was when she reached that point
when the last ray was sent, she (Alicia) could deduce an expression for Beatrice’s
speed relative to her in terms of k because speed = distance/time. Easily. Alicia
knows that the light from herself to Beatrice travels at speed c and most importantly
from Einstein, she knows that Beatrice’s light beamed back to herself also travels at
speed c. The ray that Alicia sent to Beatrice that reached Beatrice when she was at
P was sent off at time T and the return beam arrived at Alicia at time k 2 T . Thus, the
roundtrip time of the last light ray travel is k 2 T − T and the roundtrip distance is c
times this time, or c(k 2 T −T ). One half of this is the distance that Alicia had to know.
Alicia now takes one-half of this roundtrip distance and divides it by (k 2 T + T )/2,
the time it took to achieve that separation, to get the speed v in terms of k. The factors
of 1/2 cancel and we have for v/c:

(k 2 T − T ) (k 2 − 1)
v/c = = (3.1)
(k 2 T + T ) (k 2 + 1)

with T cancelling to yield the final expression. If the reader is up-to-date with his or
her algebra, he or she could readily solve for k in terms of v.6 The result is

1 + v/c
k= . (3.2)
1 − v/c

This k value determines how time periods get shifted when a source emits to a receiver
in motion.7
Now if we are dealing with a wave phenomenon such as light, the period is
the number of seconds per cycle of the wave and if we multiply this by the speed
of the wave, namely c, we get the length of a cycle, i.e. the wavelength of the

5 To see this in a step-by-step manner, note that the elapsed time O T ∗ in Fig. 3.4 is equal to Tplus half
the time between Alicia’s last emitted ray and the arrival of the last ray from Beatrice as it is received
by Alicia, i.e. O T ∗ = T + (k(kT ) − T )/2 = T + k 2 T /2 − T /2 = T /2 + k 2 T /2 = (k 2 T + T )/2.
6 If you would rather leave the algebra to others but still wish to gain confidence that (3.2) is the

correct result following from (3.1), substitute k = 3 into (3.1). You will find that v/c = 4/5. Now
substitute v/c = 4/5 into (3.2) and you will find that k = 3. Consistency! You might wish to try
this with different numerical choices of k. While this process is not a proof, it certainly will give
you confidence that (3.2) is the correct result.
7 If the speed v is much less than c, we could expand the expression in (3.2) to get the much simpler

approximate value k = v/c.

22 3 A Simple Approach to Einstein’s Special Relativity

received light relative to that of the emitted light. Notice that if v is positive as in the
Fig. 3.4 with Beatrice pulling away from Alicia, (3.2) shows that k is always greater
than 1 (with v a positive value),8 and therefore the period and hence the observed
wavelength is always greater than the emitted wavelength. This is the well-known
and very important phenomenon of “red-shift”, the shift of the wavelength of the
waves towards the red end of the spectrum of colours in the rainbow.9 As well, from
our cosmological observations of redshift, we know that the universe is in a state of
expansion with the distant galaxies rushing away from us. It is also noteworthy that
(3.2) is an exact expression, holding true even if Beatrice’s speed relative to Alicia is
nearly as great as the speed of light, c. Of course it can never reach c, not for material
objects like people! We will soon see why.
Instead of having Beatrice move away from Alicia with speed v, we could have
arranged it for her to be moving towards Alicia, i.e. with speed −v. In that case,
the derivation would have gone on as before and the net result would have been the
same as (3.2) but with v replaced by −v. With this replacement, the new expression
shows that k is less than 1, the period kT observed by Beatrice is now shorter than the
emitted period and the light is seen as blue-shifted rather than red-shifted, i.e. shifted
towards the shorter wavelength blue end of the visible spectrum. If astronomers had
seen the distant galaxies blue-shifted, they would have deduced that the universe is
in a state of contraction with the galaxies rushing towards each other and that we
were possibly heading towards the“Big crunch”.
Notice that when we replaced v by −v everywhere in (3.2), the result is the same
as if we had taken (3.2) and flipped it upside down. In other words, letting v go into
−v leads to k going into 1/k. This will be useful in what follows later.
Finally, Camela is getting into the action. Suppose she also travels relative to
Alicia but with a different constant speed. Let v AC be the speed of Camela relative
to Alicia, let v AB be the speed of Beatrice relative to Alicia that we had before and
let v BC be the speed of Camela relative to Beatrice. You trust from your everyday
experience that if you are watching a train go by at 100 km/h and people in the train
are moving toward the front of the train at 5 km/h, they appear as if they are moving
at 105 km/h relative to you, watching from the ground. In other words, the speeds
appear to simply add up. In the case of the triplets, we would have simply

v AC = v AB + v BC . (3.3)

Surprisingly, this is not quite correct in Relativity but you would not notice the
mistake unless the speeds were much higher. To arrive at the correct expression,
we have to distinguish between the k factors that relate the triplets’ observations.

8 Note that if v = 0, k = 1 and the periods are the same. This makes sense because in this case,

Beatrice and Alicia are at rest relative to each other.

9 The colours of the rainbow can be remembered by their first letters in order as “roygbiv”, red,

orange, yellow, green, blue, indigo and violet. (If you pronounce it a few times as if it were a
word, it sticks in your mind forever!) This is the visible spectrum with the red colour having the
longest wavelength and the violet colour having the shortest wavelength in the visible part of the
electromagnetic spectrum.
3.1 Navigating in Spacetime 23

Fig. 3.5 Alicia beams light A B C

to Beatrice who in turn beams
light to Camela for the same
period that she receives light.
Because of the invariance
of the speed of light, this
two-step process follows the k AC T=k BC (k AB T)
same lines as the one step of
Alicia beaming light directly
to Camela kABT
T

Let k AB be the k factor relating Beatrice to Alicia that we had before, let k AC be
the k factor relating Camela to Alicia and let k BC be the k factor relating Camela
to Beatrice. Let Alicia shine light to Beatrice for a period T . Beatrice receives the
light for a period k AB T and as long as she receives this light, she shines light to
Camela. Camela receives Beatrice’s light for a period equal to Camela’s k factor
relative to Beatrice times Beatrice’s emission period, in other words k BC [k AB T ] as
in the Fig. 3.5.
The very interesting thing is this: Thanks to Einstein, we could have just as well
seen this two-step process as the single-step emission of light from Alicia to Camela.
The lines would have been identical because the speed of light is always the same c
relative to any pairs of participants in the experiment. Viewed as the Alicia-to-Camela
emission, Alicia sends for a period T and Camela receives for a period k AC T . Now
the reception period for Camela is the same, whether she received the light in the
two-step process or directly from Alicia so we can equate the two values. Canceling
the common factor T , we have

k AC = k AB k BC . (3.4)

If the triplets’ mother Delilah (labeled D) decided that she wished to participate in
the same way, the compounding of the k factors would become k AD = k AB k BC kC D
and so on if we wish to include cousins and other relatives. We see that the com-
pounding of k factors in Einstein’s Relativity is just as simple as the compounding of
velocities in classical non-relativistic physics. It is simply a matter of multiplication
of k factors here replacing addition of velocities in classical pre-Relativity physics.
This raises the question: what is the relativistic replacement for the classical
compounding of velocities (3.3)? For this, we apply (3.1) and (3.4):
24 3 A Simple Approach to Einstein’s Special Relativity

k 2AC − 1 (k 2AB k 2BC ) − 1

v AC = c =c (3.5)
k 2AC + 1 k 2AB k 2BC + 1

and we use (3.2) in (3.5) to change the expression on the right hand side from one
that is in terms of k AB and k BC to the final expression in terms of v AB and v BC . After
a little bit of simple algebra, this gives the relativistic law for the compounding of
velocities,
v AB + v BC
v AC = . (3.6)
1 + v AB v BC /c2

Complexity is now evident relative to the simplicity of compounding in classical

physics. Clearly if the relative velocities are small compared to c, the correct rela-
tivistic expression will differ only slightly in comparison to the classical expression
of (3.3). However, if the velocities are appreciably close to c, the results of com-
pounding will be dramatically different. As a concrete example, suppose v AB and
v BC are both 7c/8. The classical compounding law (3.3) would give a value of 7c/4
for v AC , comfortably reaching well beyond c. However, you can easily verify that
the correct result for v AC using (3.6) is 112c/113, very close to c but not quite c! You
might wish to try to exceed c with v AB and v BC still closer to c but that denominator
will never let you reach c with compounding, let alone pass it. Nature’s speed limit
cannot be breached!10

3.2 The Relativity of “Now”

In terms of the spacetime diagrams, let us now see how different observers view pairs
of events differently. Let us return to Alicia and Beatrice and consider two explosion
events L and R (see Fig. 3.6) that Alicia deduces to have occurred equidistant from her
at position −x and x respectively. She knows that the explosion sites were equidistant
from her because at time t − x/c, she sends out flashes to her left and to her right to
arrive at the events L and R and they return to her simultaneously at time t + x/c.
She also deduces from these observations that the events L and R must have both
occurred simultaneously at the half-way time between the emission time and the
reception time of the returned flash, namely at the time t.
Beatrice meets Alicia at the time t. She also goes through the same steps as Alicia.
∗
She sends out her flash at time tRbef to reach the event R and she receives the return
∗ ∗
flash at time tRaft as shown in Fig. 3.6. Similarly, she send a flash at time tLbef to arrive
∗
at the event L and she receives the return flash at time tLaft . Like Alicia, Beatrice
also deduces that the flashes occurred equidistant from her because she notes that
the time intervals for the round trips of her flashes to R and L, namely tRaft ∗ − t∗
Rbef
∗ ∗
and tLaft − tLbef respectively, are precisely the same. Actually we can see this clearly
from the symmetrical positioning in Fig. 3.6.

10 But see later discussion in Chap. 13.

3.2 The Relativity of “Now” 25

Fig. 3.6 Alicia and Beatrice A B

each send beams of light that
arrive at sites of explosions to
their left and right. For Alicia, t*L
the explosions occur simulta- aft

neously but for Beatrice, the (0,t+x/c)

explosion to her right occurs t*R
aft
prior to the explosion to her
left

L R
(-x,t) x (x,t)

t*L
bef

(0,t-x/c)
t*R
bef

But the interesting thing is this: Beatrice sent out her flash to R before she sent
out her flash to L and she received the return flash from R before she received
the return flash from L. Beatrice deduces that event R happened before event L.11
The important deduction: Events simultaneous in one frame are not simultaneous in
another frame.
It is important to note that the blending of the observations for the two observers
would not proceed in this simple manner were it not for the essential fact upon which
the edifice of Relativity is constructed—that light rays from the observers who are in
relative motion actually move in unison over the indicated segments. As Bondi has
described it, “There is no overtaking of light by light”. This would not be the case
for a Newtonian observer.
What we learn from this picture is that the “now” of one observer will differ
from the “now” of another observer moving relative to the first in Relativity. This
phenomenon has led some researchers, such as the renowned mathematician K.
Gödel, to view time itself in an entirely new light. Some have come to question
whether time is eternal or whether time had a beginning and will have an eventual end
with the birth and death of the universe respectively (presupposing that the universe
is of finite lifetime). What are we to make of these speculations? Do they have any
tangible consequences for physics or are they only of relevance to philosophers?
Does it really make any sense to speak about time (or space, for that matter) if the
universe does not even exist?

11 Note that Beatrice would have deduced that R happened after L if she had been moving in the −x
direction rather than the +x direction relative to Alicia.
26 3 A Simple Approach to Einstein’s Special Relativity

3.3 Relativity Transformations for Space and Time

Let us return to our grid and our set of clocks telling us where something is present
or something happens, and when it is there or when it happens. It is important to be
able to connect the “where” and the “when” as seen by one observer, say Alicia, in
terms of how another observer, say Beatrice, designates the “where” and the “when”,
i.e. the transformation of the coordinates in spacetime of an event. We can derive the
essentials by keeping with our picture of Beatrice moving with speed v relative to
Alicia along their common x axes.
Let Alicia refer to distance along this axis as x and let Beatrice refer to distance
along this axis relative to her as x ∗ . In restricting ourselves to this picture, we are
setting the other two spatial dimensions relative to Alicia (call them y and z) and
Beatrice (call them y ∗ and z ∗ ) to have the same corresponding values for all time.
If Alicia and Beatrice set their clocks to 0 when they cross at O as in Fig. 3.7, Beatrice
will be at position x = vt at time t relative to Alicia and Alicia will be at position
x ∗ = −vt ∗ at time t ∗ relative to Beatrice. For a Newtonian, time is absolute and
there is no distinction between the clock readings t and t ∗ ; for Alicia and Beatrice,
according to Newton, t = t ∗ . However, we will see that in Einstein’s Relativity, time
is no longer an absolute.
We return to our spacetime plot for Alicia and Beatrice and now consider an event
E that occurs at position x at the time t as reckoned by Alicia and at position x ∗ at the
time t ∗ as reckoned by Beatrice.12 Alicia and Beatrice collaborate in an experiment
by having a ray of light go out from each of them in such a manner as to arrive
precisely at the position and time of the event E, and then be reflected back to each of
them. Referring to Fig. 3.7: let Q and Q ∗ be the corresponding points of the return
ray. The ray must leave Alicia at time t − x/c by her reckoning since the light travels
a distance ct in the time t. In this manner, the ray arrives at event E precisely at time
t in Alicia’s frame. Now Beatrice must do likewise (but in terms of her coordinate
system), have a ray leave her at a time t ∗ − x ∗ /c which insures that her ray will also
arrive at the event E at position x ∗ relative to her at the time t ∗ . Now by Einstein, the
speed of the ray for Alicia is the same as the speed for Beatrice; it is c for each of
them. As a result, the ray from Beatrice lies coincident with the ray from Alicia and
we can apply the same relations between the intervals that we used before. We only
use one k factor so let us define k = k AB . From before, we have the interval O P ∗
related to the interval O P as

(t ∗ − x ∗ /c) = k(t − x/c). (3.7)

The ray “illuminates” the event E and we let it reflect back to Beatrice first and then
to Alicia as shown. The reflected ray reaches Beatrice at the time of reflection plus
the time to return the distance x ∗ , i.e. at the time t ∗ + x ∗ /c. Similarly, it gets back
to Alicia at time t + x/c relative to her clock. To find the relationship between these

12 We do not prejudice the issue and let Beatrice have a clock rate of her own. If it should work out
to be the same as that of Alicia’s clock, so be it. We shall see!
3.3 Relativity Transformations for Space and Time 27

intervals, we note that here, it is Beatrice emitting to Alicia rather than vice versa.
Since the speed of now Alicia relative to Beatrice is to be taken into account, the
velocity to be used is −v. We recall that when v goes into −v, k goes into 1/k. As a
result, the O Q reception interval is related to the O Q ∗ emission interval as

1
(t ∗ + x ∗ /c) = (t + x/c). (3.8)
k
We leave it as an exercise to use (3.7) and (3.8) and to eliminate k with (3.2) to get

(x − vt) (ct − vx/c)

x∗ =  , ct ∗ =  . (3.9)
1 − v 2 /c2 1 − v 2 /c2

Note that (3.9) gives the values of x ∗ and ct ∗ in terms of x and ct. However, suppose
we wish to know the values of x and ct in terms of x ∗ and ct ∗ . If you are nimble
with your algebra, you can manipulate the equations in (3.9) to get

(x ∗ + vt ∗ ) (ct ∗ + vx ∗ /c)
x= , ct =  . (3.10)
1 − v 2 /c2 1 − v 2 /c2

However there is an easy way to arrive at this inverse result: since Beatrice has
velocity v relative to Alicia, Alicia has velocity −v relative to Beatrice. Also, for
the inversion of (3.9), we need Alicia to assume the role of Beatrice and Beatrice
to assume the role of Alicia. This is achieved by having the un-starred coordinates
of Alicia go into the starred coordinates of Beatrice and vice-versa everywhere in
(3.9). By the simple actions of these velocity and coordinate switches in (3.9), we
see that (3.10) is the result (note that (−v)2 = v 2 ). A little bit of thinking eliminated
the algebra work!
The factor 1/ 1 − v 2 /c2 appears very frequently in Special Relativity, so it is
helpful to define it as γ. It is a scaling factor which determines how much Relativity
stretches or contracts a dimension. At low velocity, say, v = 1 m/s, γ is very close
to 1 but at velocities comparable to the speed of light, say 99 % of the speed of light,
γ = 7.089. For a still higher value, at a velocity of 99.99 % of the speed of light,
γ = 70.712. You can see how dramatically the effect grows with increasing velocity.
Equation (3.9) is referred to as the “Lorentz transformation”. It finds its way into
many applications, some of which we will discuss in this book. The immediate point
to note from the second of (3.9) is that t and t ∗ are no longer the same. Time is no
longer an absolute as it was in pre-Relativity physics. This is one of the most amazing
and to many, one of the most unsettling aspects of Einstein’s Relativity. There is an
essentially new mixing between x and t in the Lorentz transformation. This mixing
of space and time will be discussed in the next section.
As we have seen before, the distinction becomes more and more essential for
relative velocities between Alicia and Beatrice getting closer and closer to that of
light. The difference fades into insignificance for velocities v much less than c which
is the case for our common daily life experience. That is why you were able to
28 3 A Simple Approach to Einstein’s Special Relativity

Fig. 3.7 The event E coin- A B

cides with the arrival of
the rays from both Alicia t + (x/c) Q
and Beatrice. The connec-
tion between intervals from
Bondi’s k-calculus is used
to build the transformation
between the space and time
coordinates used by Alicia Q* t* + (x*/c)
and Beatrice E: (t,x) re. A
(t*,x*) re. B

P*

P
t* - (x*/c)
t - (x/c)

make a reliable arrangement to meet your friend at Joe’s Diner at a time that you
both agreed upon without any problem, without invoking Einstein. However, suppose
you were “relativistic” observers, observers moving relative to each other with speeds
approaching c. Then the situation would become most interesting indeed, as we will
now witness.

3.4 The Twin or “Clock” Paradox

The twin paradox is one that has captured the imagination of many ever since it was
first discussed at the dawn of Relativity. We now delve into it with the help of our
delightful Bright sisters. Twins Alicia and Beatrice (Camela will have a role soon as
well) synchronize their clocks and Beatrice sets off on a long journey in her rocket
ship at a speed nearing the speed of light relative to stay-at-home Alicia. At a certain
point, Beatrice decides to come home, reverses direction and eventually returns to
the home of her twin Alicia. She discovers that there are strangers living in the
house, wearing clothing that she had never seen before and using electronic devices
that she could never have imagined possible. After discussions with the inhabitants,
Beatrice is shocked to learn that she is speaking to the great-great grand-children of
her beloved late sister.
The paradox consists in imagining that Beatrice had stayed put and it was really
Alicia who had made the roundtrip. Then one might have believed that it should
be Alicia who should be conversing with Beatrice’s great-great grand-children.
3.4 The Twin or “Clock” Paradox 29

Fig. 3.8 Beatrice and Camela A

use their synchronized clock
readings to simulate a twin
Q
who makes a return voyage
from home. The traveling twin B
returns, aged less than the T/k
stay-at-home twin T

kT

C
O

“After all”, one might say, “motion is relative”. Not really in this case. While inertial
reference frames are totally equivalent physically (when Alicia and Beatrice had a
constant velocity with respect to each other in our earlier examples), both Alicia
and Beatrice could not have been inertial at every stage in this case. Beatrice is the
one who makes the return voyage in an absolute sense in that she was the one who
underwent a period of deceleration followed by acceleration. She had to employ her
rocket thrusters and during the periods of deceleration and acceleration, even if the
rockets were very silent and her space-ship acoustically sealed, she could feel the
effects of being in her now non-inertial reference frame during the crucial periods.
She could feel it in the forces on her body as she sits strapped into her seat. There
is an essential asymmetry between Alicia’s and Beatrice’s journeys and it is this
asymmetry which leads to Beatrice rather than Alicia who returns to the world of
her sister’s great-great grand-children rather than vice-versa. We will now see why
it is Beatrice, the rocket ship adventurer, who is the one to live to witness the strange
experience.
Returning to Alicia and Beatrice, let Beatrice and Alicia synchronize their clocks
at the intersection point O and let Beatrice beam light to Alicia immediately after
their meeting at O for a period T (Fig. 3.8). Precisely when Beatrice’s clock strikes T,
Beatrice meets with sister Camela, who this time happens to be traveling towards
Alicia with the same speed that Beatrice is moving away from Alicia. Camela, like
Alicia and Beatrice, has an excellent clock and Camela sets her clock to time T at
the instant she meets Beatrice (point P in the diagram).
Camela begins to beam light to Alicia at that point and she continues to do so
until she meets Alicia at point Q in the diagram. Clearly from the symmetry of equal
30 3 A Simple Approach to Einstein’s Special Relativity

and opposite velocity, Camela’s beaming period is T just as it was for Beatrice. Since
Camela’s time was T when she met Beatrice at P, it is T + T = 2T when she meets
Alicia at Q. The question is: what time does Alicia read at Q?
To determine this, we note that since Beatrice is beaming to Alicia for a period T ,
Alicia receives the light for a period kT where k is the relativistic Doppler factor
between Beatrice and Alicia. We also recall that with v changing into −v as is the
case for Camela’s velocity relative to Alicia, the relativistic Doppler factor between
them becomes 1/k. Therefore, with Camela beaming to Alicia for a period T, Alicia
receives Camela’s beam for a period T /k. We deduce that the time elapsed for
Alicia between O, her meeting with Beatrice, and Q, her meeting with Camela, is
(kT + T /k) = (k + 1/k)T . While Camela says it is 2T o’clock at Q, Alicia says it
is (k + 1/k)T o’clock.13 You can easily satisfy yourself that k + 1/k is greater than
2 for all positive values of k (except for k = 1 which is of no interest since for this
special case there is no motion at all), and so we conclude that Alicia has aged more
than her triplet sisters. Notice that it is the asymmetry that determines which way the
relative aging goes: the single inertial observer Alicia ages more than the combined
inertial observers Beatrice and Camela. There was one observer in the first case, two
observers in the second case.
Our natural inclination would be to think of the combination of Beatrice and
Camela as being equivalent to the earlier described case of Beatrice alone, blasting
off in a rocket ship and returning younger than Alicia when she returns home. This is
because we could imagine the Beatrice-Camela combination being almost like a Beat-
rice voyage alone, with the only difference being a short deceleration/acceleration
period turnaround for Beatrice. However, the following objection is immediately
raised by various critics: while the different results hold for the readings of the clocks
when all the observers are inertial as described in Fig. 3.8, it does not follow that this
will be equivalent to replacing the two inertial observers Beatrice and Camela with
the single noninertial observer Beatrice. In fact various authors as well as pundits
have argued that the clocks of Alicia and Beatrice will get back into synchronization
during the deceleration/acceleration period, and Alicia and Beatrice will reunite with
the same time reading at Q.
In this regard the first author recalls the wonderful lectures that Bondi gave at the
Brandeis Summer Institute in Theoretical Physics in 1964 when he raised this very
issue. In his discussion, Bondi suggested that with the deceleration and acceleration
applied very gradually, there should be no significant effect on the clock, but he
admitted that the issue remained unresolved. Through the years, some have argued
that smooth as this may be, nevertheless the accumulated effect must be one of a
re-adjustment of the clocks to have their readings coincide at Q.
The counter-argument that we have advanced is as follows [1]: consider a variety
of journeys of different durations by Beatrice with the same constant velocity but
with all of the journeys having identical deceleration/acceleration turnaround peri-
ods. We have illustrated three of such journeys in Fig. 3.9. The essential point is this:
the different journeys result in time differences for Alicia and Beatrice that vary from

13 It is a very simple algebra exercise to show that k + 1/k = 2γ.

3.4 The Twin or “Clock” Paradox 31

Fig. 3.9 Three different A

return voyages for Beatrice
with each having identical
turnaround characteristics.
Since they are identical, they
could at most re-adjust Beat-
rice’s clock by exactly the
same amount in the three
cases. However the three voy- B
ages would require three dif-
ferent adjustments to get back B
into synchronism with Ali-
cia’s clock. Thus, we see that
the deceleration/acceleration B
effects could not bring the
clocks of Alicia and Beatrice
into synchronization

journey to journey as a result of the different values of T. Therefore, the decelera-

tion/acceleration periods would have to induce different degrees of make-up in time
for the different journeys to get the times to agree for the various cases. However,
assuming a turnaround does force a make-up in time, it must do so by precisely the
same amount in all the cases, because the turnaround is identical and there has been
no change in the state of the underlying spacetime for all of the journeys.14 Thus we
see that the argument that the deceleration/acceleration phases provide the required
adjustment periods is inconsistent with the varied make-up requirements and the
conclusion of asymmetric ageing holds.
The result is certainly counter-intuitive, so it is worth considering this very inter-
esting phenomenon in different ways. The essential aspect that Relativity brings to
bear here concerns the nature of time, that it no longer enjoys the absolute status that
it had in classical physics. Various treatments of Relativity refer to time as being just
another coordinate, on the same basis as the coordinates of three-dimensional space.
However this is very misleading. It is true of course that Relativity dethrones time
as an absolute, in contrast to position, which, even in classical physics, is dependent
upon the observer.15 But time’s new status in Relativity is not one of equivalence
to space; rather it bears a reciprocal relationship to space. Its essential distinction
from that of space remains in Relativity but it combines with space in a very special

14 We will have more to say about the nature of evolving spacetimes in General Relativity. Here
we are dealing with Special Relativity where the spacetime is always the same.
15 Example: if A is moving relative to B, A is always at the position O, the origin of A’s coordinate

system anchored on her body whereas B records a continuously different position of A as time
advances.
32 3 A Simple Approach to Einstein’s Special Relativity

manner that we will discuss in what follows. Einstein’s Relativity leads us to deal
with this wonderful marriage of concepts that we call “spacetime” in a manner that
essentially changes our view of physics.

3.5 Perceptions of Length and Time

Let Alicia hold a “rigid” meter rod firmly in her grasp. As Beatrice moves with
velocity v relative to Alicia along the x axis, Beatrice wishes to determine the length
of the rod as she sees it in her frame of reference. Suppose the left end is at position
x1 and the right end is at x2 in Alicia’s reference frame. Since it is at rest in Alicia’s
frame, the ends have these values for all times t of Alicia’s clock. The 1 m of the
rod is the difference between the right-end coordinate and the left-end coordinate,
i.e. 1 m = x2 − x1 . In physics, we call the length that is read in the rest frame of an
object, its “proper length”. Thus Alicia’s length of 1 m that she measures is the rod’s
proper length.
We wish to find the length of the rod that Beatrice perceives. We use (3.10), first
for the right end position x2 and then for the left end position x1 :

(x ∗ + vt2∗ ) (x ∗ + vt1∗ )
x2 =  2 , x1 =  1 . (3.11)
1 − v 2 /c2 1 − v 2 /c2

Now we have to consider an issue that we might easily take for granted. It is well to
ask: which times t2∗ and t1∗ are we to choose? Actually it does not matter as long as
they are the same. This is the only logical way to define the length as Beatrice sees
it: it is the difference of the end coordinates x2∗ − x1∗ with each end coordinate taken
at a common time t ∗ = t2∗ = t1∗ . For Beatrice, her length is this “snapshot” of the
end-point differences.
Thus with t2∗ = t1∗ , we subtract the second equation from the first in (3.11), the t2∗
and t1∗ terms cancel each other, and we find

(x ∗ − x1∗ )
x2 − x1 =  2 = γ(x2∗ − x1∗ ) (3.12)
1 − v 2 /c2

or 
l = l ∗ / 1 − v 2 /c2 (3.13)

where the lengths for Alicia and Beatrice are respectively

l = x2 − x1 , l ∗ = x ∗2 −x1∗ . (3.14)

Note from (3.13) that since 1 − v 2 /c2 = 1/γ is always less than 1 for velocities v
that are not 0, we have the result that l ∗ is less than l. Beatrice perceives Alicia’s meter
3.5 Perceptions of Length and Time 33

rod to be less than 1 m in length. Clearly Alicia’s perception of length, the proper
length, is the maximum length that can be measured for the rod. Measurement relative
to a frame having any motion relative to the rod brings in the 1/γ factor which is less
than 1 for any v different from 0. Proper length is maximal.
At this point, you might suspect that we are trying to sell you the Brooklyn Bridge.
You might wish to raise the objection that objects moving relative to you do not have
the appearance of having shrunk. However, your experience with motions of objects
are such that their velocities are much smaller than the speed of light, c. Suppose you
were to witness an object moving by at the incredible speed of 10 km/s. You can easily
calculate that for this velocity, 1/γ = 0.999999995! Even with this extraordinary
velocity, the effect of Relativity is so very minute. However in particle accelerators,
with particles approaching the speed of light, the relativistic effect is appreciable and
it cannot be neglected. This Relativity effect is witnessed on a daily basis in labs
throughout the world.
We now return to the issue of the perception of time. We considered this in the
previous section but now we will examine it using the Lorentz transformation. Let
Beatrice carry her clock which is at her position x B∗ . It is always at this position
relative to her coordinate system because she is carrying it. Let Beatrice record two
ticks of the clock, tick 1 at time t1∗ and tick 2 at time t2∗ . Consider the times to which
these ticks correspond in Alicia’s reference frame. We use (3.10) for the two ticks in
turn:
(ct ∗ + vx B∗ /c) (ct ∗ + vx B∗ /c)
ct1 = 1 , ct2 = 2 . (3.15)
1 − v 2 /c2 1 − v 2 /c2

We subtract the first equation from the second, the x B∗ terms cancel and we are left
with
c(t ∗ − t1∗ )
c(t2 − t1 ) =  2 . (3.16)
1 − v 2 /c2

Thus we see that the time interval T = t2 − t1 as read in Alicia’s frame is related to
the interval T ∗ = t2∗ − t1∗ as read by Beatrice’s clock as

T ∗ = T /γ. (3.17)

The relativistic γ factor appears as before but with a different significance. The
time T ∗ read in the rest frame of the clock (Beatrice’s frame) is called the “proper
time”. Note from (3.17) that Alicia’s time T is necessarily longer than Beatrice’s
proper time T ∗ . Proper time is minimal; proper length is maximal. This reciprocal
relationship between the proper quantities again underlines the essential difference
between space and time in Relativity. Time is not just another coordinate like space.
34 3 A Simple Approach to Einstein’s Special Relativity

Fig. 3.10 The plank, seen as shortened by Alicia and her friends, approaches the channel. They
prepare to strike the plank simultaneously by their reckoning

3.6 More Paradoxes in Special Relativity

Earlier, we dealt with the interesting twin paradox and now we turn our attention to
two paradoxes involving lengths. Suppose a channel that is 10 m long and 1 m wide is
carved out of the ice and Alicia and a group of her friends are standing with hammers
along its length. Beatrice invites a group of her friends to stand with her along her 10 m
“rigid” plank and they go hurtling along the ice at a relativistic velocity (Fig. 3.10).
By Relativity, Alicia and company view Beatrice’s plank as somewhat shortened and
when it is in a position over the channel, Alicia and her friends simultaneously take
solid smacks at the plank with their hammers (taking care to strike only in the spaces
between the plank riders!) (Fig. 3.11). The plank descends through the ice channel
and into the icy water below (Fig. 3.12).
However, examining the situation from the vantage point of Beatrice and her
friends, they see the channel that has shrunk. After all, relative to them, it is the
channel that is in relativistic motion while their plank is at rest and it is always the
moving object that appears shorter. The paradox that presents itself is clear: How
can this now-shortened channel accommodate the longer plank from the viewpoint
of Beatrice and company?
To resolve the paradox, two important aspects come into consideration: simultane-
ity and rigidity. While the smacks are simultaneous for the members of the Alicia
troupe, a careful use of (3.10) reveals that they are not simultaneous for Beatrice
and her friends who are standing on the plank. Beatrice et al see the leading end
being hit first and the following points along the plank hit in a succession of blows
along its length. The leading end enters the water first and succeeding plank portions
3.6 More Paradoxes in Special Relativity 35

Fig. 3.11 Alicia and her friends smack the plank simultaneously while it is totally over the channel

Fig. 3.12 The plank is now submerged under the ice and continues its forward motion

follow in turn as the plank seems to slither into the water below as if it were a noodle
(Fig. 3.13)!
However the plank was supposedly rigid, yet here we are speaking of a flexible
noodle. This demands an explanation and the one that we face might seem somewhat
astounding: truly rigid bodies cannot exist in Relativity! It is easy to see why this
is so. Suppose the plank in question were truly perfectly rigid. If that were true,
36 3 A Simple Approach to Einstein’s Special Relativity

Fig. 3.13 Seen from the vantage point of Beatrice and her friends, it is the channel that is shortened.
They see the Alicia crew smack the leading end first followed by successive smacks along the length
of the plank and the plank enters the ice in a noodle-like shape. After all the elements of the plank
have submerged, the straightened plank continues its forward motion under the ice

all of its elements would always have to move in unison as there could never be any
buckling. However, suppose the plank were to be smacked at one end. As soon as
it was smacked, that end would move. Since it is supposedly rigid, every portion
of the plank would have to move simultaneously with the smacked end. To do so,
the pulse from the smack would have to be transmitted along the plank with infinite
speed, but we know that effects can propagate with at most speed c. Thus the plank
must buckle, however minutely. The moral of the story: simultaneity is relative and
absolute rigidity does not exist.
Some years ago, we posed this paradox on a test to a group of students, asking
them to present the resolution. One answer stood out from all the rest. The resourceful
student, let’s call him Willy Wiseguy, reasoned as follows: because of Relativity,
the results are different for the Alicia and Beatrice observers. The Alicia observers
discover that the Beatrice group, along with the plank, end up submerged in the water
below whereas the Beatrice observers, seeing a shortened channel, never actually pass
through the ice and into the water but simply continue on their way along the ice!
Hmm. Very interesting, Willy. But your grade is an “F” nevertheless. At the end of
the operation, the plank is either above the ice or below, within the water. It is either
wet or dry. It cannot be both.
Another interesting paradox concerns the rod and the barn doors. Alicia has her
two friends stand guard at the front and rear doors facing each other 20 m apart in
the family barn. The front door is open. Beatrice holds a rod whose proper length in
its natural uncompressed state (recall: this is the length that is measured in the rest
reference frame of the rod) is slightly longer than 20 m. Alicia is determined to have
the rod fit into the barn so she directs Beatrice to run so fast that Alicia sees the rod
appear to have shrunk to 20 m. By this means, the rod can fit into the barn (Fig. 3.14).
Alicia has directed her friend to snap the front door shut when the rod ends are
precisely at the front and rear doors. The rod is in the barn (Fig. 3.15). The next step
follows: the rod is now trapped but in a compressed state. It is compressed because
it has a shorter length than it had when it was measured in its rest frame before being
trapped. As it decompresses and expands, it blasts the rear and front doors open as
the rod expands back to its original uncompressed proper length (Fig. 3.16).
3.6 More Paradoxes in Special Relativity 37

Fig. 3.14 In the figure, the

rod is seen approaching the
barn from the viewpoint of
the rest-frame of the barn.
From this viewpoint, the rod
which is longer in the rod rest-
frame, is seen, in the barn rest
frame, to have a sufficiently
shortened length so that it can
fit completely into the barn

Fig. 3.15 As seen from the

viewpoint of the barn rest-
frame, the front door of the
barn is shut when the rod is
totally within the barn

Now let us consider the situation from Beatrice’s viewpoint. For her, the barn
appears to have shrunk so that the distance between the front and rear doors is even
less than 20 m yet Beatrice is running with her longer-than-20-m rod, as she perceives
its proper length (Fig. 3.17). One would have to decide: either the rod did manage
to get inside the barn or it did not get inside the barn. We can well imagine Willy’s
response: “it’s relative. The rod gets into the barn as far as Alicia is concerned but it
never gets in according to Beatrice.”
Again, poor Willy falls short. However it must be said that this example is some-
what more complicated since we are now dealing with changes in velocity in the
direction of initial motion.
For Beatrice, the barn appears to be even shorter than 20 m. According to her view,
the leading end of the rod has stopped at the rear door while the rest of the elements
of the rod are still in motion with a portion of the rod still outside the front door
(Fig. 3.17). As time progresses for Beatrice, the elements of the rod make their way
through the front door until the rod’s trailing end reaches the front door (Fig. 3.18).
At that point, the front door is shut, an event later than the event of the front end of
the rod reaching the rear door. Then, the stresses build up to the point where the doors
38 3 A Simple Approach to Einstein’s Special Relativity

Fig. 3.16 From the viewpoint of the barn rest-frame, the trapped rod blasts the rear and front doors
open as the rod expands back to its proper length

Fig. 3.17 The rod is pictured as it approaches the barn from the viewpoint of the rest-frame of the
rod. In this frame, the barn appears to be even shorter than 20 m

cannot constrain the rod and it pops out the doors from Beatrice’s vantage point as
well (Fig. 3.19). For Beatrice, the stresses build in a different sequence from the one
experienced by Alicia.
The common theme in these as in other paradoxes revolves around the issue of
the Relativity of simultaneity. We can imagine how difficult it must have been even
for scientists in the days of Einstein’s first formulation of Special Relativity, as it is
not comfortably accepted by many, even to this day.
3.7 Exploring Spacetime 39

Fig. 3.18 From the viewpoint of the rest-frame of the rod, the leading end gets stopped at the rear
door while the rod elements are compressing to the point where the trailing end of the rod enters
the front door

Fig. 3.19 From the viewpoint of the rest-frame of the rod, the rear door gets blasted open slightly
before the front door gets blasted open

3.7 Exploring Spacetime

We are familiar with our feelings about the relativity of time in a psychological sense.
For example, we might say to a friend “Doesn’t it feel just like yesterday that we
were on that hiking trek in the mountains?” However we know that in actual fact,
it was three months ago. In earlier sections, we began to come to terms with the
powerful new idea that in Relativity, time is relative in a very real sense. It is not just
in our minds. We are now ready to probe further and familiarize ourselves with the
important concept of “spacetime”, Relativity’s amalgam of space and time. To really
appreciate how this comes together, we will have to use some mathematics, but it
will be kept simple.
40 3 A Simple Approach to Einstein’s Special Relativity

We start with the beautiful theorem of Pythagoras that we learned in grade school.
If we have a right-angle triangle with side lengths x and y, the hypotenuse length l
is related to the side lengths as in Fig. 3.20,

l 2 = x 2 + y2. (3.18)

The simple extension of this theorem to a rectangular box of sides x, y and z with
distance l between opposite corners is (Fig. 3.21)16

l 2 = x 2 + y2 + z2. (3.19)

We now wish to make the leap from space to spacetime. Alicia and Beatrice shift
into 3-dimensional space and Alicia, whose reference frame coordinates are (x, y, z),
arranges to have a flash of light shine from an emission time t1 at position (x1 , y1 , z 1 )
and be absorbed at time t2 at position (x2 , y2 , z 2 ) in her frame of reference. She could
station observer 1 at (x1 , y1 , z 1 ) and observer 2 at (x2 , y2 , z 2 ), each with clocks and
each at rest relative to Alicia. Their life experience is the same as that of Alicia; they
are part of her frame of reference. In terms of Alicia’s coordinates, the distance l12
that the flash of light travels, using (3.19) is given (as a square) by
2
l12 = x12
2
+ y12
2
+ z 12
2
(3.20)

where x12 = x2 − x1 , y12 = y2 − y1 , z 12 = z 2 − z 1 . However, there is another way

to express the distance that the flash has traveled. After all, distance equals velocity-
times-time and since the velocity is c and the time interval is t12 = t2 − t1 , we could
just as well write the distance as
l12 = ct12 . (3.21)

We now can equate the square of the distance expression in (3.21) to its value as
expressed in (3.20) and if we take the latter to the other side of the equation, we are
left with 0:
c2 t12 2 − x12 2 − y12 2 − z 12 2 = 0. (3.22)

Now Beatrice, who is moving with velocity v relative to Alicia, first considers the
event of the emitted flash. For her, it has for the spacetime coordinates, the time17
that it was emitted and the place from where it was emitted in her starred frame of
reference, (ct1∗ , x1∗ , y1∗ , z 1∗ ). The event of flash absorption according to Beatrice is
(ct2∗ , x2∗ , y2∗ , z 2∗ ). In the same manner as we had for the events in the Alicia reference
frame, we have for Beatrice’s reference frame

16 We could continue on in this vein with boxes of 4, 5, 6 etc. spatial dimensions but we stop at

3 because our (at least macroscopic) spatial world is 3-dimensional. Note that we cannot draw the
pictures for 4 and higher spatial dimensions.
17 It is handier to multiply the time by c.
3.7 Exploring Spacetime 41

Fig. 3.20 The sides x, y

of a right-angled triangle
are related in length to the
hypotenuse l as l 2 = x 2 + y 2

l
y

Fig. 3.21 The sides x, y, z

of a rectangular box are
related to the diagonal l as
l 2 = x 2 + y 2 + z 2 . The dotted
line indicates the hypotenuse
of Fig. 3.20
l

z
y

∗2 ∗2 ∗2 ∗2
c2 t12 − x12 − y12 − z 12 = 0, (3.23)

in other words everything that we had for Alicia, only now everything is starred
except c. The reason that c is unstarred is simple and important: c = c∗ , the invariance
of the speed of light.
The expressions on the left hand sides of (3.22) and (3.23) are of particular signifi-
cance in Relativity. They embody the square of what we call the “spacetime interval”
∗ relative to Beatrice) between the two events 1 and 2, in this case, the
s12 (and s12
events on the path of a light ray:
2
s12 = c2 t12
2
− x12
2
− y12
2
− z 12
2
. (3.24)
42 3 A Simple Approach to Einstein’s Special Relativity

and
∗2 ∗2 ∗2 ∗2 ∗2
s12 = c2 t12 − x12 − y12 − z 12 (3.25)

in Beatrice’s starred reference frame. From (3.22) and (3.23), we see that for the
events on the path of a light ray, both s12 and s12 ∗ are zero. Now suppose those

events 1 and 2 were very very close together. In calculus, we write the infinitesimally
small difference between the spacetime coordinate values and the spacetime interval
between the two events as

d x = x2 − x1 , dy = y2 − y1 , dz = z 2 − z 1 , cdt = c(t2 − t1 ), ds = s2 − s1
(3.26)
which allows us to express (3.24) as

ds 2 = c2 dt 2 − d x 2 − dy 2 − dz 2 (3.27)

and (3.25) as
ds ∗ 2 = c2 dt ∗ 2 − d x ∗ 2 − dy ∗ 2 − dz ∗ 2 . (3.28)

Earlier we saw that when the finite spacetime interval was zero for Alicia, it was
also zero for Beatrice. If it is so for finite quantities, it is also true for the infinitesimal
quantities ds and ds ∗ . Now suppose that we are dealing with a pair of events that do
not lie on the path of a light ray. Such pairs of events cannot be separated in such
a manner as to produce a zero spacetime interval because whatever the relationship
between them, it is not mediated by a signal at the speed of light c. So we now
consider the connection between ds and ds ∗ in general. In Appendix A, we provide
a simple proof that
ds = ds ∗ (3.29)

in general, even when they are not zero.

The important result that emerges is that the spacetime interval between any two
given events is an invariant.
It tells us how the special combination of space and time is something upon which
everyone must agree. Its structure is also of interest: notice that it is the positive
time interval squared and the negative of the sum of the space intervals squared that
combine to give us the invariant. This again points to the essential distinction between
time and space. Time is not just another coordinate like the space coordinates. The
difference in the sign in front of the quantities spells this out clearly.

3.8 Types of Intervals and the Light Cone

We have already encountered the special type of interval connecting two events that
lie on the path of a light ray, the emission of a flash and its subsequent absorption. This
type of interval is called “light-like” or “null”; its length is zero, s12 = 0. The reason
3.8 Types of Intervals and the Light Cone 43

Fig. 3.22 Alicia screams and some distance from her position, a goblet shatters

that we can achieve zero length with the necessarily positive values of the squares of
the distances and the square of the time interval is because the spacetime intervals
have pluses and minuses in front of them, plus in front of the time interval squared
and minuses in front of the space intervals squared. In the light-like intervals, the
pluses and minuses perfectly balance each other to give a net value of zero. Light-
like intervals represent the limiting boundary of what we term “causally connected
events”, those pairs of events, 1 and 2, for which event 1 can cause event 2. In the
flash example, the event of absorption was dependent on there being the event of
emission. There was a cause-and-effect linkage.
Let us consider an example of a different type of event pair: suppose Alicia at her
position x = 0 at the instant t = 0 emits a high-frequency scream. Some distance x
from her position, a fine crystal goblet sits on the table and at the slightly later time t,
the goblet shatters into many pieces (Fig. 3.22).
The spacetime interval s squared between the events of the scream (Event (0, 0))
and the shattering (Event (ct, x)) is18

s 2 = c2 (t − 0)2 − (x − 0)2 = c2 t 2 − x 2 . (3.30)

We now consider the sign of s 2 in (3.30). To determine the sign, we consider the phys-
ical connection between the events. The goblet shatters because Alicia has induced
high-frequency oscillations of the air molecules in her mouth and these oscillations,
the sound, has traveled through the air at the speed of sound, vs , until they reached
the goblet and the induced vibrations sufficed to shatter it. There was the cause, the
scream, and the effect, the shattering. The sound traveled a distance x in the time

18 y = 0 = z for both Alicia and the goblet.

44 3 A Simple Approach to Einstein’s Special Relativity

t so its speed was vs = x/t. Thus vs2 t 2 = x 2 . If we substitute this value for x 2 in
(3.30), we have
s 2 = c2 t 2 − vs2 t 2 = (c2 − vs2 )t 2 . (3.31)

Now the speed of sound is less than the speed of light so we see that s 2 is positive and
therefore s is a real number. Events separated by a real number spacetime interval
are said to be “timelike separated” events. Note also that the shattering of the goblet
occurred after Alicia’s scream. Now we know that Relativity can produce some
surprising effects so it is well to ask whether Beatrice, moving relative to Alicia,
could ever see the events occur in reverse order. In the footnote below, we show that
this can never happen.19 It is very well that this can never happen. It would be very
disconcerting for us if it were possible for Beatrice to travel with sufficient speed
that she should see the shattering before the scream! It would play havoc with our
common sense because the scream caused the shattering. Without Alicia’s scream,
that goblet would have continued its existence intact in its original state. To sum up,
the causally related events occur in a definite order in time for all observers.
It is useful to show this in a spacetime diagram bringing into our lexicon for the
first time, the “light cone” (Fig. 3.23). As before, we draw the spacetime picture from
Alicia’s viewpoint. She is at the origin O of her (t, x) coordinate system and at time
t = 0, she screams. The goblet shatters at a later time ts at some distance xs away
from her. We also have rays of light flash by her from left to right and from right to
left just as she screams. We draw these rays with just the right slopes so that x/t = c
for the left-to-right ray and x/t = −c for the right-to-left ray.20 Note the position
of the shattering goblet event: it occurs at a spacetime point that lies between the
vertical time axis and the left-to-right propagating light flash. It lies there because
we can trace the sound wave that came from Alicia’s mouth to the goblet and the
speed vs of the sound wave, traveling slower than c, follows the path that lies closer
to the t axis than the light ray path.
The idea of the preservation of order in cause-effect relationships is very important.
Later we will discuss how some researchers have taken seriously the notion that this

19 We use the second equation of (3.9). Here v is the velocity of Beatrice relative to Alicia. For the
scream event for Alicia, x = 0, t = 0 so we see from the equation that it occurs at time t ∗ = 0 as
well for Beatrice. The goblet-shattering event is at time t ∗ given by the second of the equations (3.9)
and therefore, for Beatrice, the separation in time (the later time minus the earlier time) between the
events is t ∗ − 0 = t ∗ . If we substitute vs t for x, we have t ∗ = γct (1 − vv
c2
s
). Clearly this is always
positive since both v and vs are always less than c. Thus t ∗ is necessarily positive and Beatrice sees
the shattering after the screaming just as Alicia did.
20 In the figure, we have gone beyond the use of only one spatial dimension x. We have imagined

that light rays flashed to Alicia from every direction in the x–y plane, arriving at her position just as
she screamed. Then, instead of two single rays as in our earlier spacetime diagrams, we have two
cones of rays, one cone of rays going in, the backward or past light cone, and one cone going out,
the forward or future light cone. Ideally, we would like to draw a three-dimensional light cone since
we inhabit three-dimensions of space. However, we require one dimension of the plot to mark off
the passage of time and we have no dimension left to do so after using the three spatial dimensions
for the illustration. We have to give up the aid of the spacetime diagram in this case and work only
with the mathematics.
3.8 Types of Intervals and the Light Cone 45

Fig. 3.23 In this figure, we

see the event of the shattering
of the goblet at a time later
than the event of the scream.
The shattering event lies
within the future light-cone of
the scream event

scream

ordering can be violated in physics. The subject is connected with what are technically
called “closed timelike curves” or “time machines” in the popular parlance. We will
also show how we have approached the issue, concluding that they are a mathematical
artifact rather than an element of physical reality.
Now let us consider a different pair of events. The first event is again Alicia’s
scream but now the second event is the landing of an astronaut on Mars (Fig. 3.24).
Let us suppose that the landing event occurred 5 s after the scream in Alicia’s reference
frame but at precisely the same instant in Beatrice’s reference frame. We are not at
all surprised that this could be so; after all, Alicia’s scream certainly has nothing to
do with the Mars landing in the sense of cause and effect.21 While we are aware that
they followed in that particular order for Alicia, we would naturally say “so what?”.
We would also say “so what? ” if Camela happened to be travelling at the right velocity
to witness Alicia’s scream after the Martian landings. The earlier transformations
can be used to show these possibilities and we are not at all surprised. Without a
cause-effect relationship, it does not really affect our sense of logic. However, it is
important to note that in the old Newtonian way of looking at reality, there was an
absolute time and there was an absolute ordering of events, whether they were of the
first variety, the scream and the shattering goblet, or the second variety, the scream
and the Mars landing. One of the events followed the other in that particular order

21 This is so even if there were air between the Earth and Mars to carry the sound wave!
46 3 A Simple Approach to Einstein’s Special Relativity

scream

Fig. 3.24 In this case, Alicia screams and in her reference frame, 5 s later, an astronaut lands on
Mars. However Beatrice determines that these two events occurred simultaneously in her reference
frame. There is no contradiction as the Mars landing event occurs outside of the light cone of the
scream event

for everybody. For Newton, time was absolute. For Einstein, time takes on a whole
new meaning.
Let us probe further into the Mars landing on the spacetime plot. It occurred
5 s after Alicia’s scream so if we were to draw a line to indicate a ray that connected
the scream to the landing, it would have to be a fictitious ray, i.e. unphysical because
it would have to be traveling at a speed greater than that of light.22 After all, even
light cannot travel from Earth to Mars in 5 s! On average, it would take approximately
12 min for light to travel from Earth to Mars. We say that the Mars landing occurs
outside of the light cone from Alicia’s scream.
The spacetime interval squared between the scream and the Mars landing,
s 2 = c2 t 2 − x 2 is a negative number because x 2 /t 2 is greater than c2 . Thus, the
interval itself, the square root of a negative number, is an imaginary number. Events
separated by an imaginary spacetime interval are said to be “spacelike separated”
because they occur at spatially different locations for all observers. How do we know
this? Easy! For suppose there existed a reference frame of someone like Beatrice
for whom the events occurred at the same place in space. If that were so, then the
interval squared would be s 2 = c2 t 2 − 0 which is a positive number. But then, the s 2
value, which, we recall is an invariant, being positive negates the fact that it must be
negative for all observers including Beatrice. So we cannot have these events occur
at the same spatial position for any observer; they are absolutely separated spatially.
So we have three kinds of regions relative to the scream, the O = (0, 0) event
in our spacetime plot: the region within the upper light cone is the “absolute future”

22 See however, the discussion in Chap. 13 where it was claimed that the recent neutrino experiment
reveals a transport of particles with speed exceeding c.
3.8 Types of Intervals and the Light Cone 47

Fig. 3.25 The opening-up of the light cones as the speed c is taken to increase. For infinite c, the
light cone is replaced by a plane representing the Newtonian “now”. At that stage, the “before”
zone is the entire region below the plane and the “after” zone is the entire region above the plane

relative to O; the region within the past light cone, with an analogous argument, is
the “absolute past” relative to O and the entire region elsewhere we designate as the
“absolutely separated” region relative to O. The remarkable difference in Newtonian
physics is that the absolutely separated region does not exist! For Newtonian physics,
which embodies interactions at infinite speed of propagation, the upper and lower
cones open up to the extent that they are both coincident with the (x, y) plane. The
cone structure is no longer present and we are left with simply the absolute past and
the absolute future. At that stage, the (x, y) plane represents the “now” where all of
the events are coincident in time with the scream event at time t = 0 (Fig. 3.25).
We will find it very useful to have this understanding about events in spacetime
when we come to the discussion of gravitational collapse and black holes.

3.9 Energy-Momentum—the Relativity Modifications

Let us return to our earlier discussion about the important mechanical laws of Newton.
Recall from the Second Law that when an unbalanced force F acts on a body,
it gets accelerated and the acceleration a is proportional to the unbalanced force. The
proportionality factor, the property of the body involved, is what we call its mass, m,
the measure of its inertia, if you like, the extent to which it resists any change in its
existing state of rest or motion. Expressed in equation form,

F = ma. (3.32)

This is one of the most important equations in classical physics. It also appeals to our
common experience: for a given force F, the bigger the mass m that is being pushed,
the smaller will be its acceleration. Actually the more general way to express this
law, to account for the fact that the mass might be varying, by for example shedding
48 3 A Simple Approach to Einstein’s Special Relativity

fuel exhaust, is by means of the quantity that we call momentum, p:

p = mv. (3.33)

The improved version of (3.32) is

dp
F= . (3.34)
dt
Here, dp is the little increment of momentum that occurs in the course of the little bit
of time dt in which it makes this momentum increment. From (3.34), the standard
development that you might wish to read about in the mechanics text books, leads to
an expression for “work” which is the force acting through a distance and how this
work gives the change in the energy, E of the body. The energy assumes the form

mv 2
E= . (3.35)
2
In Relativity, we follow a more general law of physics which is astonishingly suc-
cessful and beautiful. It is called the “Principle of Least Action”. The law is that
for any physical system, there exists an invariant quantity S called the “action” such
that the system, in going from one state to another state, follows the route which
minimizes this action. Many physicists regard this law as the most fundamental in all
of physics. An interesting simple account of this law is to be found in [4] and a more
sophisticated technical approach meant for more advanced students of physics in [3].
By applying the least-action law, we find that the proper expression for momentum
is no longer mv but rather
p = γmv (3.36)

mv 2
and the energy is no longer 2 but rather

E = γmc2 . (3.37)

It is easy to see that for small velocities compared to the velocity of light c, the rela-
tivistic momentum expression (3.36) merges with the classical mechanics expression
(3.33) since γ approaches 1 in that case. However the relativistic energy expression
takes on an important change. As we see from the above equations, while the clas-
sical energy and the classical and relativistic momentum go to zero when v goes to
zero, the relativistic energy goes to

E = mc2 . (3.38)

Remarkably, it tells us that at its base level, matter at rest, every bit of matter, regard-
less of how seemingly insignificant, is endowed with an intrinsic energy by virtue of
its having mass. If we put in the numbers, we can appreciate just how astounding an
3.9 Energy-Momentum—the Relativity Modifications 49

amount of intrinsic energy is contained in even a tiny amount of mass. As an example,

consider a mere wisp of matter, let us say one gram of sugar, which hardly shows
up in a teaspoon, and imagine it being converted entirely to energy. By (3.38), this
tiny amount would produce 9.1013 Joules of energy, enough to run a thousand-watt
heater for three thousand years!
Another important point to note is that γ in (3.37) would become infinite if the
body could ever attain the velocity c of light. This again points to the speed limit c
in nature.
Chapter 4
Introducing Einstein’s General Relativity

4.1 What is Gravity?

In the preceding chapters, we witnessed the new aspects of physics brought about by
Einstein’s Relativity. All of this arose because: (a) there is a speed limit in nature, c,
the maximum speed at which influences can be propagated and (b) because all inertial
observers are physically equivalent, they must all agree on this speed limit. How-
ever, not a word was mentioned about gravity and it is with the inclusion of gravity
that Einstein’s Relativity takes on a whole new and exciting complexion. Einstein’s
Relativity without gravity is called “Special Relativity” to distinguish it from Rela-
tivity with gravity which is called “General Relativity”. In brief, General Relativity
is Einstein’s theory of gravity. In what follows, we will delve into General Relativity,
showing how it is the curving of spacetime in the general theory that replaces the old
Newtonian idea of gravity being just another force.
The essence of gravity was appreciated by the cave-men and cave-women: If you
were a cave-dweller and you let go of your club, you found that it fell to the ground.
If you were to step off the edge of a cliff, you discovered that you would fall, and that
your speed would increase more and more as you fell further and further. The further
the fall, the more it would hurt you so you learned from childhood to take gravity
very seriously. If you were a cave-dweller, the power of gravity was an essential part
of your existence, as it is for us to this day. It was natural to think of gravity, that
agency that makes objects fall, as a force like the others.
Delving further, you would notice that a bird feather fell more slowly to the
ground than did a rock so you would naturally build into your mind-set the idea that
lighter objects accelerate less than heavier objects in falling under gravity. However,
you would be misguided because the air-resistance had been playing a crucial role in
retarding the fall of the feather more than of the rock. If you were a more scientifically
inclined cave-woman, you might have picked up a pebble and a rock and dropped
them together from a substantial height after positioning your significant-other cave-
man to act as observer. You would probably have been very surprised to learn from
him that the pebble and the rock hit the ground simultaneously. In this experiment,

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 51

DOI: 10.1007/978-3-642-30385-2_4, © Springer-Verlag Berlin Heidelberg 2012
52 4 Introducing Einstein’s General Relativity

the air-resistance played a much less significant role. In fact, a dramatic illustration of
the fact that bodies of different masses fall with the same acceleration under gravity
is achieved by re-doing the feather-and-rock experiment in a vacuum tube where
almost all of the air has been removed. The feather and the rock would be seen to hit
the bottom simultaneously.
Newtonian physicists of the 17th century had understood this well. After all, if we
were to combine (2.1) and (5.1), we have

ma = Gm M/r 2 (4.1)

where we now let m be the mass of the pebble or the rock and let M be the mass of
the Earth that is gravitationally attracting them. The mass m cancels on both sides
of the equation and we see that the acceleration a is

a = G M/r 2 (4.2)

which is independent of the mass m, the mass of the pebble or the rock. Thus, pebbles
and rocks fall the same way.
Soon we will discuss how Einstein used this result to launch his new theory of
gravity. However, we will first consider why Einstein was confronted with the neces-
sity to formulate his new theory. It is well to ask because Newtonian gravity, in
conjunction with Newton’s laws of motion had served physics well for centuries in
describing how the Moon revolved around the Earth, the planets revolved around the
Sun and in more modern times, how rocket trajectories were determined. The imme-
diate problem for Einstein was that issue of the speed limit in nature. In Newton’s
gravity, when the distribution of mass changes, for example if you were to stretch out
your arms, the gravitational influence of your doing so would be felt instantaneously
throughout the universe!
For the benefit of those with some background in calculus,1 the form of Newton’s
law of gravity in differential form is

∇ 2 φ = 4πGρ (4.3)

∂ 2 ∂ 2∂ 2
where ∇ 2 = ∂x 2 + ∂ y 2 + ∂z 2 , φ is the gravitational potential whose gradient deter-
mines the local acceleration and ρ is the mass density. This differential equation is
mathematically of the “elliptic” type for which changes propagate instantaneously.
Thus, a change in ρ in the example described above, as the removal of mass density
from one location (along the sides of your body) and relocating it to a new position
(perpendicular to your body), would instantaneously alter φ throughout the universe
and the acceleration experienced at all points in the universe would change without
any delay whatsoever. This would contradict Special Relativity, the theory which has
been remarkably successful in experiment after experiment.

1 This paragraph can be skipped by non-calculus readers.

4.1 What is Gravity? 53

Fig. 4.1 Standing in an

elevator on the surface of
the Earth, Einstein releases
a pebble and a rock from
rest. They fall along (nearly)
parallel lines at the same rate
of acceleration

It is worth repeating: The very successful theory of Special Relativity pivots on

the concept of a finite invariant speed limit in nature. So Einstein set himself on the
path of discovering a new theory of gravity, one which would meld properly with
Special Relativity.
Various physicists including the first author’s mentor, N. Rosen, who worked
closely with Einstein, have remarked upon a key aspect of Einstein’s approach:
simplicity. Einstein treasured simplicity and had the faith that the essential truths in
physics would be simple. His lead-in to a new theory of gravity bears witness to this
credo. Going back to those pebbles and rocks, Einstein imagined some experiments2
in an elevator. In Fig. 4.1, we have Einstein in a stationary elevator holding a pebble
and a rock. Released from rest, the pebble and the rock fall at the same rate to the
floor, as we discussed before.
Now Einstein imagined this elevator imbedded in a rocket ship far from gravitating
matter. The rockets are blasting, accelerating the elevator upwards in the Fig. 4.2 at
a rate of 9.8m/s 2 , the acceleration that objects experience at the surface of the
Earth. Now with Einstein releasing the pebble and the rock, he experiences the same
phenomenon that he did on Earth: He witnesses the pebble and the rock “fall” to the
floor with an acceleration of 9.8m/s 2 , just as before on Earth. Einstein saw the two
phenomena as being physically equivalent. However, relative to observers outside
the rocket ship, the released pebble and rock appear to hang suspended (Newton’s
First Law) while it is Einstein who appears to be accelerating upwards at a rate of

2 These are frequently referred to as gedanken experiments from the German, meaning “thought”
experiments, experiments not actually performed but rather imagined to have been performed.
Einstein was particularly fond of such exercises. He employed them in his intellectual battles with
N. Bohr over the issue of the probabilistic interpretation in quantum mechanics. See [5] for an
interesting account of these exchanges.
54 4 Introducing Einstein’s General Relativity

Fig. 4.2 A pebble and a

rock are released from rest
in an accelerating rocketship.
Relative to the rocket ship as
reference frame, they “fall” as
if they had been released at
the surface of the Earth

9.8m/s 2 . This equivalence is commonly encapsulated into a principle, the Principle

of Equivalence which states that a gravitational field is locally equivalent to an
accelerated reference frame. The gravitational field in this case is the field produced
by the mass of the Earth and the accelerated reference frame is that of the accelerating
rocket ship. Notice that the word “locally” has been slipped in to the statement of the
Equivalence Principle. This is because the equivalence is only an approximation: In
the first case, the pebble and the rock do not fall on precisely parallel lines but rather
on converging lines that point to the center of the earth (Fig. 4.3). Thus, they land
closer to each other than they were when they were released. On the other hand, in
the rocket-ship experiment, they fall on strictly parallel lines.
Now let us consider another pair of Einstein thought experiments. Einstein is
in that same elevator holding a rock in each hand at the same level and his wife
M. Maric facing him, is holding two rocks, one above the level of Einstein’s two
horizontally-positioned rocks and one below this level so that the four rocks form a
square. The elevator, at rest on the fourth floor, suddenly has its support cable severed
and the elevator goes into free-fall. Just as it does so, all four rocks are released. What
happens to the rocks? Ignoring for the moment the variation in the gravitational field
of the Earth, all four rocks as well as Einstein, Maric and the elevator itself, fall
4.1 What is Gravity? 55

Fig. 4.3 Examining the

motions of the pebble and
the rock more precisely, we
see that they fall on lines that
converge towards the center
of the Earth. Thus, they land
closer together than they were
when released

towards the ground at the same rate of acceleration, 9.8m/s 2 .3 So relative to the
frame of the elevator, there is no relative motion among all of the aforementioned
bodies; they remain together in a state of suspension.
Now we switch over to the elevator imbedded in the rocket ship. Einstein and Maric
repeat the arrangement. While the rocket engines are shut, they release the rocks as
before. Clearly by Newton’s First Law, the result is as before in the freely-falling
elevator: All the bodies remain suspended in their original positions. The Equivalence
Principle has applied again. We have ignored the variation of the gravitational field
so the effects of reference frame acceleration and the gravitational field have been
realized: There has been a local equivalence.
Let us now look at the situation more precisely: When the cable is severed,
Einstein’s rocks, falling on lines that converge to the center of the Earth, come
closer together with time. As for Maric’s rocks, the lower rock, being closer to
the center of the Earth than the higher rock, experiences a greater acceleration by
Newton’s law (5.1) (Fig. 4.4). (This effect manifests itself daily on Earth in creating
the tides in the seas.) Thus, her rocks actually separate with the advance of time.
The combination of both effects has been described by some as a kind of squeezing
process.
We will return to the Equivalence Principle later but for now, let us advance
towards the goal of formulating Einstein’s theory of gravity, General Relativity. For

3 We also ignore the very tiny interactions between the rocks themselves.
56 4 Introducing Einstein’s General Relativity

Fig. 4.4 Four rocks are

released from rest within
an elevator which is falling
freely near the surface of
the Earth. Examining the
motion very accurately rel-
ative to observers who are
also falling freely within the
elevator, we see that while
the horizontally-positioned
rocks move closer together,
the vertically-positioned rocks
separate in time. This is
because the lower rock expe-
riences a slightly greater
acceleration than the upper
rock

this, we will require a little bit of mathematics. We return to the spacetime interval
(3.27) of Special Relativity. When we moved with a constant velocity relative to
the original coordinate reference frame, the spacetime interval retained its original
structure, (3.28). However, it is easy to show that when we move with acceleration
relative to the original reference frame, the form of the spacetime interval changes
and depending upon the type of acceleration, often in a complicated manner.4 Various
combinations of the coordinates appear as multipliers of the d x 2 , dy 2 , etc. In fact,
in general, new products of the form d xdt, dydt, dzdt, d xd y etc. also appear. Let
us write this as

ds 2 = g00 dt 2 + g11 d x 2 + g22 dy 2 + g33 dz 2 + g10 d xdt + g01 dtd x + .. (4.4)

where we have used the symbol gik with i and k taking on the values 0, 1, 2, 3 to
represent these combinations of the spacetime coordinates that arise under the trans-
formation. Here we are writing 0 for t, 1 for x, 2 for y and 3 for z.5 Mathematicians
and physicists express this in generality in the elegant compact form

ds 2 = gik d x i d x k . (4.5)

4 See for example [3] for an explicit demonstration of the changes to the form of the spacetime

interval that are produced when the transformation to an accelerated reference frame is made.
5 Note that for simplicity of expression, we have dropped the stars on the quantities that arise under

the transformation to the new coordinate frame.

4.1 What is Gravity? 57

While our example has focused on Cartesian coordinates x, y, z, in General

Relativity, we apply this form to arbitrary systems of coordinates.6,7 The quan-
tity gik is called the “metric tensor”8 and it plays a vital role in Einstein’s new theory
of gravity. The reason is clear: The effect of acceleration is encoded in the functions
that make up gik . But by the Equivalence Principle, gravity is locally equivalent to
an accelerated reference system. Therefore, it is natural that this metric tensor plays
a key role for Einstein’s relativistic theory that includes gravity. Notice how it has
evolved from Special Relativity which focused on the invariant nature of ds 2 in 3.27,
so it has incorporated Special Relativity within itself properly as the limiting case.
It is all very logical.
However, we must keep in mind that while we can effect a transformation from a
situation with no gravity to a new accelerated frame in which there is the semblance
of gravity, a second transformation that is the opposite of the first brings us back
to the original situation of no gravity. You might be inclined to call this “hocus-
pocus”, and in a sense you would be justified. This is not “real” gravity; real gravity
is the field that is produced by real bodies rather than by coordinate transformations.
The interesting thing about real gravity, like the field around (and within) the planets,
the Sun and every bit of matter and radiation, is that no coordinate transformation
can remove it. In what follows, we will describe the essence of real gravity.

4.2 Towards a New Theory of Gravity

An important clue was already provided for us in our hunt for the new relativistic
theory of gravity. This was contained in the fact that with real gravity, those rocks in
the earlier example did not accelerate along parallel lines as they did for the simulated
gravity, but rather when there was real gravity, they moved along converging lines.
It leads us to ask: What is the difference between spacetimes that are the domain of
simulated gravity as opposed to the spacetimes with real gravity? As we will discuss
further, it will turn out that the spacetimes of simulated gravity are flat like the surface
of a pane of glass whereas spacetimes with real gravity are curved, like the surface
of a ball or the surface of a saddle. In fact General Relativity will bring us to the
realization that while all bodies and fields (apart from the gravitational field) live in

6 The repetition of an index means that it is to be summed over 0, 1, 2, 3. Here it is done for the
indices i and k. This is referred to as the “Einstein summation convention”. Note that the compact
expression expands out to the form (4.4).
7 Thus, x i could represent (r, θ, φ) spherical polar coordinates, (r, z, φ) cylindrical polar coordi-

nates, etc.
8 For mathematically inclined readers, the vector, which is the mathematical object of greater

familiarity, is actually a special case of a tensor. Like soldiers in the army, tensors are categorized
by their rank. Rank number is given by the number of indices attached to the tensor. A vector,
having the single index i, is a tensor of rank one. The metric tensor, having two indices, is a tensor
of rank two. Tensors are mathematical constructs defined by their transformation properties. A nice
development of the subject can be found in [7].
58 4 Introducing Einstein’s General Relativity

spacetime, gravity is spacetime, manifested by its curvature. Gravity is a “field” like

none other.
For the benefit of the more mathematically-inclined reader, we provide in Appen-
dix B a brief account of how the famous Einstein field equations of General Relativity
are derived. These Einstein field equations

8πG ik
G ik = T (4.6)
c4
have been described by many physicists as the most beautiful and profound in all
of physics. They connect the energies, momenta and stresses, the T ik , of all matter
and non-gravitational fields, to the metric tensor that is contained in the Einstein
tensor G ik in a very complicated way. And we recall that it is the metric tensor
which describes gravity. Einstein might have wished for simplicity but ironically
the Einstein tensor is anything but simple! A glance at Appendix B will reveal that
it actually has 10 separate components, G 00 , G 11 , G 22 , G 33 , G 01 ... etc. with each
one generating its own equation G 00 = 8πG c4
T 00 , G 11 = 8πG
c4
T 11 ... etc. These 10
equations are to be compared to the single gravitational field equation of Newtonian
physics that we wrote in (4.3). The difference is still more dramatic: The Newtonian
gravity equation has three simple terms on the left hand side whereas each of the 10
Einstein gravitational field equations has many terms with many that contain prod-
ucts of the components of the metric tensor gik .9 Such equations are referred to as
being “non-linear”. When you read the expression “non-linear partial differential
equation”, you could substitute in your mind the expression “king-sized headache”.
This is because they are most often very difficult to solve. Generally speaking, exact
solutions of useful non-linear differential equations have been found for the sim-
plest of these equations. When confronted by more complicated equations of this
type, approximation techniques are often employed to extract answers with hope-
fully accurate approximation to the exact solutions. Approximation procedures can
be rather complicated and they have generated quite a bit of controversy over the
years. While Einstein cherished simplicity, his visions embodying simple truths and
observations translated into immensely complicated mathematical structures. That
is the downside. However the upside is a wonderful new richness and a world of new
possibilities that have excited researchers, and in turn the general public, to this day.
In the succeeding chapters, we will explore various aspects spawned by Einstein’s
theory of gravity.

9 A note for the calculus-equipped reader: These are actually partial derivatives of the metric tensor
components.
4.3 Motion of Bodies in General Relativity 59

4.3 Motion of Bodies in General Relativity

Let us return to Newtonian physics and (2.1) and (2.2). Taken together, they tell us
that the acceleration a experienced by a mass m acted upon by the gravity of the
Earth of mass M, depends only on M and the distance r of m to the center of the
Earth. It’s value is
a = G M/r 2 . (4.7)

From elementary calculus, we can express this as

d
a=− (G M/r). (4.8)
dr
We refer to G M/r as the “gravitational potential” of the Earth and the equa-
tion (4.8) says, in words, that the acceleration of a body under the influence of the
Earth’s gravity is given by the negative of the gradient of the gravitational potential.
In Newtonian physics, the latter is the driver of gravitational acceleration. The “gra-
dient”, expressed by the symbol d/dr in (4.8), is just what the word implies: It is
the steepness with which the quantity next to it is changing, here the gravitational
potential, G M/r .
This is a particularly simple case. Generally, a distribution could well be more
complex than the simple picture given by (4.8) in which the Earth is assumed to
be a perfectly spherically symmetric distribution of mass. When the mass distrib-
ution is not spherically symmetric, the gravitational potential is accordingly more
complicated. In general, we call the gravitational potential, φ, and we express the
acceleration as
a = −grad(φ) (4.9)

where “grad” is the gradient. The simplest possible case would be one in which the
potential is totally uniform, totally without change from point to point. In this case,
the steepness with which it changes is zero, i.e. no steepness at all, the gradient of a
constant is zero. If you walk along a level path, there is nothing to climb, no gradient
whatsoever. This is equivalent to the situation in Newtonian physics where there is
no force at all, where Newton’s first law of motion comes into play and there is no
acceleration,
a = 0. (4.10)

Now with no force present and no gravity present, suppose we were to examine
the motion from the vantage point of an accelerated reference frame. Then, relative
to this new frame of reference, the equation of motion changes from (4.10) to what
is called the “geodesic equation”, where a new, general form of acceleration, aintrin ,
the “intrinsic acceleration”, arises and it is this new acceleration form that is zero for
force-free motion. The new equation of motion is10

10 See Appendix B for a more detailed description of this equation.

60 4 Introducing Einstein’s General Relativity

aintrin = a + uu = 0 (4.11)

where a here is the ordinary acceleration,  is the “Christoffel symbol” and u is the
“four-velocity”. The  is a geometrical quantity made up of the metric tensor gik and
its derivatives. Of course when we apply the special case of using a non-accelerated
reference frame in Cartesian coordinates, the  vanishes and (4.11) reverts to (4.10)
as it must.
The Newtonian approach to motion is based upon the traditional idea of force, in
this case gravitational force, inducing acceleration. The General Relativity approach
is quite different. In General Relativity, motion under gravity alone is “free” motion,
motion in which the new quantity, intrinsic acceleration, is zero. How does this arise?
Actually, it is all very logical. After all, by the Equivalence Principle, the effect of
gravity is locally like being in an accelerated reference system and mathematically,
the motion in an accelerated reference system is given by the geodesic equation
(4.11).11 In fact the geodesic equation is the geometrical expression of the wonderful
unifying principle in physics, the Principle of Least Action that we discussed briefly
earlier. It is another case of an extremal equation. As a simple example, in three-
dimensional geometry, the geodesic equation is the equation that guides one over the
shortest path between two points. On the surface of a sphere, it picks out the paths
along the great circles. For example, to fly from London to Vancouver, the pilots
minimize the flight time (and in general, the expenditure of fuel) by following the
shortest path, the route that cuts across the Arctic, tracing along the great circle. You
can draw this path by choosing the plane through London, Vancouver and the center
of the Earth and marking the plane’s intersection with the Earth (Fig. 4.5).
Now the interesting connection to the weak gravity that we experience for exam-
ple, on Earth, is this: The geometric quantity  approximates the gravitational poten-
tial gradient grad(φ). When it is simply shifted to the right hand side of (4.11), we
get back our Newtonian equation of motion (4.9). It all fits together as it must.
There is a dramatic change in our conception of gravity that has taken place in the
process just described. The Newtonian idea of the motion of a body under gravity
as being the effect of action as a result of a force, the gravity force, has been totally
replaced. In General Relativity, motion under gravity alone is force-free motion!
Instead of the Newtonian force-driven motion, it is now seen to be free motion along
these very special extremal paths in spacetime called geodesics.
J. L. Synge [6], a very brilliant mathematical physicist, was a very careful analyst
of the essentials in General Relativity. He decried with gusto, physicists’ fixation on
the Equivalence Principle, emphasizing correctly that the essence of gravity in Ein-
stein’s Relativity is embodied in spacetime curvature. The supposed essential equiv-
alence of gravity to the acceleration of one’s coordinate system that one often hears
about is, as Synge has emphasized, incorrect. Acceleration in a spacetime devoid of
gravity produces pseudo-gravity, not real gravity. Just as acceleration produces the

11In Appendix B, we discuss the mathematics behind the intrinsic acceleration and how motion
under gravity alone in General Relativity follows the geodesic equation.
4.3 Motion of Bodies in General Relativity 61

Fig. 4.5 The shortest path

between Vancouver and
London is the “geodesic”
curve, the arc-segment of a
great circle that is determined
from the intersection of the
plane containing Vancouver,
London and the center of the
Earth with the surface of the
Earth. The Earth illustration is
courtesy of NASA

semblance of gravity, the reverse of this acceleration removes it. Real gravity, i.e.
spacetime curvature, cannot be removed by changing one’s system of coordinates.
The distinction is essential and surprisingly, a fair number of researchers are either
unaware or are unappreciative of this fact. Nevertheless, the usefulness of the Equiv-
alence Principle as a guide should not be underestimated. It was this principle that
directed Einstein towards the development of General Relativity. We have seen how
it has led us to the metric tensor as the replacement vehicle for the Newtonian grav-
itational potential and how it has helped us to formulate the new General Relativity
law for the motion of bodies under gravity. It has provided the scaffolding in the
construction of the great edifice of General Relativity and once erected, there is no
need to retain it as if it were an essential part of the completed structure. In this
regard, Synge’s point is well-taken.
In a variety of writings, you will witness authors proclaiming that once accelera-
tion enters the picture, you are dealing with General Relativity, Einstein’s theory of
gravity. This is false. Bondi (see for example [2]) and others have produced excel-
lent examples of Special Relativity formulations (i.e. Relativity without gravity)
with accelerated reference frames. One must not equate acceleration to real gravity.
Spacetime curvature is to be equated to real gravity.
Chapter 5
Testing Einstein’s General Relativity

5.1 Motion of the Planets Around the Sun

It has been said that Newton was the first to state, with exemplary modesty, “If
I have seen further it is only by standing on the shoulders of giants.” While we
would place Newton and Einstein, as the ultimate giants at the very highest level
of achievement in the history of physics, there is much wisdom and justification in
Newton’s homage to earlier researchers. After all, the greatest advances were made
with the help of important advances by those who preceded Newton and Einstein.
Arguably the greatest of these was made by Copernicus who replaced the Earth with
the Sun as the central body in what we now recognize as the Solar System.
For our purposes in leading towards the role of General Relativity, a useful starting
point is the work of T. Brahe. He devoted much of his life to a careful study of the
motions of the planets around the Sun. His precise measurements enabled J. Kepler
to encapsulate the essential general truths that were buried in the vast stores of the
Brahe data. These “Kepler Laws” are:
(a) The planets move in elliptical orbits around the Sun with the Sun itself at a focus
of the ellipse.
(b) The planets traverse equal areas of their ellipses in equal amounts of time.
(c) The square of the period of the orbit of any given planet is proportional to the
cube of the semi-major axis of its elliptical orbit.
Note that the elliptical orbit as indicated by the First Law has been exaggerated
in the illustration Fig. 5.1. The planetary elliptic orbits are actually nearly circular.
The orbits are more like Fig. 5.2. Far more pronounced orbital ellipticity occurs in
the case of comets that are in bound orbits. However, by so drawing them, we are
able to illustrate more dramatically, the nature of the Second Law, namely that in
order to cover the same area at close approach as at the more distant orbital position,
this law tells us that a planet moves faster at close approach than at the more distant
position (Fig. 5.1). The Third Law tells us that the planets at greater distance from

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 63

DOI: 10.1007/978-3-642-30385-2_5, © Springer-Verlag Berlin Heidelberg 2012
64 5 Testing Einstein’s General Relativity

Fig. 5.1 To cover equal

subtended areas in equal
amounts of time, the planets
must move more quickly at
their closer approach to the
Sun
1 month A A 1 month

Fig. 5.2 The actual planetary

orbits are nearly circular

the Sun take longer to complete their orbits, that a Mercury “year”, for example, is
much shorter than a Neptune “year”.
While Kepler condensed the vast stores of Brahe data into three empirical laws,
Newton performed the ultimate further condensation. It came in the form of an
actual theory of gravitation, that every body attracted every other body with a force
proportional to the product of their masses and inversely proportional to the square
of the distance between them. This is Newton’s “Law of Universal Gravitation”, as
we saw in (2.2):
F = Gm M/r 2 (5.1)

with G, the constant of universal gravitation. It is quite remarkable to witness how

Kepler’s laws reveal themselves from Newton’s great generalization. There are a
number of excellent texts on classical mechanics to which the reader might wish to
refer, that demonstrate this very clearly.
Newton’s gravitational theory served physics very admirably for hundreds of
years. It was Einstein’s Relativity that created the necessity for its replacement.
However even prior to Einstein, in spite of its enormous success as a theory, a
lingering problem with Newton’s gravity could not be overcome. It arose as a result
of the detailed observations of the orbit of the planet Mercury, the innermost planet
of the Solar System. The history is interesting. While Newton’s gravity dictates that
a single planet circling a star will trace out a closed elliptic orbit, the presence of
additional circling planets in our Solar System must perturb the orbit of the planet
5.1 Motion of the Planets Around the Sun 65

Fig. 5.3 The planets beyond

Mercury’s orbit cause the
otherwise closed elliptic orbit
expected on the basis of
Newtonian gravity to precess
as shown. General Relativity
adds an extra amount of
precession

in question. As a result of the extra tugs from the other planets, the elliptic orbit of
any given planet will no longer be closed but rather will steadily shift or “precess”
as illustrated in Fig. 5.3. By precise calculations, the orbital precession in the case of
Mercury was supposed to be on the order of 500 s of arc per century. Surprisingly,
astronomers found that the observed value was greater than the Newtonian-based
calculated value by 43 s of arc per century. So convinced were astronomers of the
infallibility of Newtonian gravitational theory that they posited the existence of a yet-
unseen planet orbiting between Mercury and the Sun that would add an additional tug
to resolve the discrepancy. It was even blessed with a name, Vulcan, However, try as
they may, astronomers were unable to find any trace of Vulcan. Thus the matter was
left in a state of abeyance. Later, we will revisit Mercury’s anomalous precession.
We will witness the magical power of Einstein’s General Relativity to resolve the
problem. To do so, we will focus on what is arguably the most important solution of
the equations of General Relativity, that due to K. Schwarzschild.

5.2 The Schwarzschild Solution

Recall the old narrative of Newtonian physics: Bodies exert forces upon each other
in accordance with the inverse-square-law (5.1) and by virtue of the gravitational
force, they move in accordance with Newton’s Second Law of Motion, F = ma.
Einstein’s General Relativity replaces this narrative with a wholly different vision:
Matter curves spacetime according to the Einstein field equations and freely-moving
bodies in the resultant curved spacetime follow the extremal paths of the curved
spacetime, the geodesic paths. The first task is to determine the curved spacetime
that the source under study creates. In the case of a star such as our Sun, it is a
good approximation to consider it as a spherically symmetric distribution of matter.
Accordingly, we choose the coordinate system best suited to take advantage of this
66 5 Testing Einstein’s General Relativity

Fig. 5.4 Spherical polar z

coordinates r, θ, φ are the ideal
coordinates to use for systems
with spherical symmetry

θ
r

symmetry, namely spherical polar coordinates (r, θ, φ). The origin of the coordinate
system is taken at the center of symmetry, the center of the star, r is the distance from
the center of the star and θ and φ are the polar and azimuthal angles respectively that
determine the position of any given point. These are shown in Fig. 5.4.
Just as we were able to write the little increment of spacetime in terms of the
Cartesian coordinate increments of physical distance as d x, dy, dz in (3.27), so too
we can express it just as well in terms of spherical polar coordinates as

ds 2 = c2 dt 2 − dr 2 − r 2 (dθ2 + sin2 θdφ2 ). (5.2)

There is an important idea that we have to get used to here. It concerns the translation
from the mathematics to actual physical distance. When we express the little incre-
ments of distance in the x, y, z directions as d x, dy, dz respectively, we understand
these as real physical distances. Now in spherical polar coordinates, the dr of (5.2) is
the small increment in physical distance along the radial (r ) direction. However the
little increments dθ and dφ are not increments of physical distance. Rather, they are
increments of angle, the small shifts in polar angle and azimuthal angle respectively.
The small physical displacement in the θ direction is r dθ. You will recall from your
elementary geometry that arc length of a portion of circle circumference is equal
to radius times the angle that is subtended and if you wish to encompass the entire
circumference of a circle, you subtend the angle all the way around the circle, 360◦
or 2π radians of angle. Then all the r dθ increments add up to 2πr , the circumference
of the circle. Similarly, the physical increment of distance in the azimuthal φ direc-
tion is r sin θdφ. We must use the entire quantity, not just the little angle increment.
5.2 The Schwarzschild Solution 67

Fig. 5.5 In spherical polar z

coordinates, the small incre-
ments of distance in the
radial (dr ), polar (r dθ) and
azimuthal (r sin θdφ) direc-
tions are indicated. Multi-
plied together, they form
a small volume element
d V = (dr )(r dθ)(r sin θdφ) θ
dr

rd θ

x
rsin θ d ϕ

Of course the increment of distance traveled by light in the increment of time dt is

cdt, the velocity times the time increment. We will be referring to this idea in an
important manner in what follows for General Relativity.
We focus on determining the curved spacetime outside of the star where there is no
matter, where the matter tensor T ik = 0. The Einstein field equations (4.6) reduce to

G ik = 0. (5.3)

Within 2 months following the completion of General Relativity in late 1915, the
solution of the Eq. (5.3) in the case of spherical symmetry was found exactly by
K. Schwarzschild, and his solution had remained one of the key sources of further
research in General Relativity over the subsequent years. As non-linear differential
equations go, today we see these field equations for spherical symmetry as being
very simple. This is a reminder that it is unfair to disparage advances made by others
after the fact as being “obvious” or “trivial”, adjectives often bandied about by some
researchers who really wish that they had been the ones to have made the obvious
and trivial advances! Schwarzschild is to be fully commended for seeing what others
before him had failed to see. His name and work live on. Schwarzschild’s famous
metric is1

ds 2 = (1 − 2m/r )dt 2 − (1 − 2m/r )−1 dr 2 − r 2 (dθ2 + sin2 θdφ2 ) (5.4)

1 Readers with a more mathematical bent might wish to see the explicit form of the field equations
for spherical symmetry and how they are solved in [3].
68 5 Testing Einstein’s General Relativity

where we have adapted our coordinates to make G = 1 and c = 1. In the standard

coordinates, where 2m/r appears in (5.4), we would have 2G M/c2 r .2 Note that if we
set m to zero in (5.4), we recover (5.2), the spacetime interval of Special Relativity.
It is most interesting to examine the effect of the simple introduction of mass on the
spacetime interval. Recall that when we wished to express physical distance in the
azimuthal direction, for example, we had to use the entire quantity that is multiplied
by the azimuthal angle increment dφ in the spacetime interval, namely r sin θdφ. With
sin θdφ
gravity in General Relativity, it is even more involved, namely √r 1−2m/r . Similarly,
in the radial (r ) direction, the physical increment of distance is √ dr
1−2m/r
and in the
polar (θ) direction, it is √ r dθ .
1−2m/r
This brings us to what might seem a surprising next step but a step that neces-
sarily follows: What √was a small increment of time in Special Relativity, dt, is now,
according to (5.4), 1 − 2m/r dt. We are led to the fact that relative to observers at
different radial positions, periods of time actually change in the presence of gravity.
From the perspective of an observer at large r , the clocks at an r value near to a
star run slower than his clock and the increments of radial distance appear shorter.
By contrast, from the perspective of an observer at small r value, clocks at large r
value appear to run fast relative to his clock and radial increments of distance appear
lengthened. In all of the cases discussed, we see that for a given r , the larger the
mass m, the more dramatic the effect on the spacetime geometry, and the passage of
time, with time being an element of spacetime, is no exception.
Recall the approach of General Relativity to motion: Matter and fields such as
electromagnetic fields curve the spacetime according to the Einstein field equations.
In vacuum, these equations are (5.3). Bodies in vacuum outside of these sources
move in the curved spacetime according to the geodesic equations (4.11). For the
curved spacetime in vacuum under the assumption of spherical symmetry, the metric
is (5.4), the solution of (5.3).
When we substitute the solution (5.4) into the geodesic equations (4.11), we
find, after some substitutions and simplifications, the equation of the orbit.3 General
Relativity provides an extra piece to the motion equation which is the lowest-order
correction from Einstein’s theory. If this term is neglected, the result is Newton’s
equation which yields the closed elliptic orbit for a single planet. The Einstein term
provides the extra orbital precession.

2 Setting c = 1 means that we are setting 3.108 m/s = 1, a pure number. This implies that 3.108
m = 1s. What this means is that with this choice, we replace every second that appears by 3.108 m
in any subsequent point in the analysis. Similarly, setting G = 1 enables the replacement of every
kilogram that appears by an appropriate number of meters. Thus, all quantities that follow are
expressed in numbers of meters. While this procedure, highly favoured by relativists, might seem
strange, it is really very sensible. Quite apart from simplicity of expression, it enables one to compare
quantities meaningfully in terms of size as they all appear in terms of the same units, meters. In the
end of the calculations, we can readily revert to conventional units if we wish.
3 For the benefit of the more mathematically-inclined reader, the equation of motion is d (1/r ) +
2
dφ2
1/r = M
l2
+ 3m
r2
where l is the angular momentum per unit mass of the planet.
5.2 The Schwarzschild Solution 69

Over the years, we have witnessed the thrill felt by many students of General
Relativity in seeing how the addition of the Einstein term in the equation of plan-
etary motion almost magically provides that extra 43 s of arc per century of orbital
precession for the planet Mercury. There are no free parameters involved in making
the fit of theory to observation. Either it works or it does not work. And it works
beautifully! It is hard not to become a firm believer in General Relativity when one
witnesses this wonderful result.
It should be noted that a more dramatic demonstration of the precession is provided
by the motion of the components of the binary pulsar P S R1913 + 16, a system
consisting of a neutron star, the source of the observed very regular almost precisely
periodically recurring pulses of radio-wave emission, in orbit with another very
compact body, possibly a White Dwarf star. The orbit is very tight and hence the
velocities are very high, yielding a more dramatic precession. Here the precession is
in degrees per year rather than seconds of arc per century! Again, Einstein’s General
Relativity stands triumphant.4

5.3 Other Classical Tests of General Relativity

The correlation between the General Relativity prediction and the observation of the
anomalous precession of Mercury’s orbit is classified as one of the three “classical”
tests of General Relativity. There are two other tests labeled classical. One is the
gravitational red-shift, the difference in frequency of the light from emitting atoms
in the Sun’s photosphere as compared to their frequency as observed on Earth from
like atoms. Earlier, we discussed the role of the observer’s velocity in determining
the period between the ticks of a clock in Special Relativity. As we saw above,
in General Relativity, periods of time vary according to the local strength of the
gravitational field: The stronger the field, the longer the period of time as judged
by a distant observer. In other words, clocks run more slowly relative to a distant
observer when they are in stronger gravitational fields in comparison to clocks that
are in weaker gravitational fields. Since larger periods mean smaller frequencies, the
light that is emitted from a given type of atom at the position of the Sun will have
a lower frequency (and hence a longer wavelength) than that of light emitted by the
same type of atom on Earth.
It is natural to ask how one would identify a given photon as having come from a
particular atomic transition in a given type of atom. This is easy: When excited, each
atomic species produces its own particular set of characteristic spectral lines, its own
DNA, if you like. When you see a particular pattern, you might, for example exclaim,
“Ah, it’s sodium!” There could be no mistaking it for any other atomic emission, so
particular are the “fingerprints” of any given atomic species. You know the pattern
from observing the emissions from heating sodium in the lab here on Earth. Now one
might think that by observing the emissions from the Sun, we would have an excellent

4 We will have more to say about the binary pulsar when we discuss gravity waves.
70 5 Testing Einstein’s General Relativity

means of testing General Relativity. Unfortunately, there are various complications.

First, the Sun is a very agitated body whose emitting atoms are in continual turmoil.
As a result, any shifts in frequency of the observed photons are due not only to the
effect of General Relativistic clock-slowing but also to the Doppler-shifting of light
coming from a moving source.
As an additional complication, there is a simple means of deducing the lowest-
order gravitational effect without even invoking General Relativity [8]. It works as
follows: Take a photon at the surface of the Sun with intrinsic energy hνs and effective
potential energy due to its presence in the gravitational field and have it lifted to the
Earth where it would have an intrinsic energy hνe and a new effective potential energy
due to its location in a different level of gravitational potential. This is like throwing a
ball up in the air: It starts with a certain speed that you give it in the throwing, a certain
kinetic energy, and a certain energy because of the gravitational field that is present at
the position at which you are throwing it. It rises in the air and as it does so, it slows
down. It loses kinetic energy but gains gravitational potential energy, the energy
that comes from being located at a higher point in the gravitational field. You can
appreciate that it is a gain in potential energy because it has the potential to do more
damage by dropping from higher up. So there is a trade-off between kinetic energy
and potential energy but the essential thing is that the sum of the two forms of energy
in each position is constant, total energy is conserved. This is one of the holy grails
of physics. Now if you apply this kind of simple reasoning to a photon instead of a
ball and you substitute its intrinsic energy hν for the ball’s kinetic energy (the photon
always has the same speed c so its intrinsic energy shows up in terms of its frequency
rather than its speed), and apply the principle of energy conservation, you get the
lowest order effect that General Relativity predicts for the photon’s frequency shift.
Admittedly it is an unusual “mishmash” of physical ideas but it works to this level
of accuracy. At this lowest order, there were very delicate experiments performed
by Pound and Rebka (see [8] for references) to test this “gravitational red shift”
effect. To bypass the uncertainties of source velocity of photons arriving from the
Sun, they performed their experiments in a tower on the Earth where the source and
absorber velocities could be carefully controlled and monitored. They relied upon
the difference in gravitational potential at the base of the tower relative to the top
of the tower. For their time, their experiments yielded a good level of accuracy in
verifying the effect. Vessot and Levine were able to improve on the level of accuracy
in a satellite experiment called Gravity Probe A.
The third “classical” test of General Relativity is the gravitational deflection of
light. Again, thinking of a photon as a little ball of energy and therefore having an
effective mass (not rest mass, a photon has zero rest mass) according to E = mc2 ,
when it passes by a massive body like the Sun, there will be a tug on it resulting in a
perceptible shift in its direction of motion. Again, an amalgam of the quantum idea
of a ball of energy with effective mass deduced from Special Relativity together with
Newton’s law of gravitational attraction would lead one to predict such an effect.
However General Relativity, through its geodesic equation of motion (now applied
to light) within the curved spacetime created by the massive body, predicts a greater
degree of deflection compared to that predicted by Newtonian gravity.
5.3 Other Classical Tests of General Relativity 71

Earth

Venus
Sun

Fig. 5.6 Pulsed radar signals are directed to Venus during superior conjunction. The echo timing
is used as a confirmation of the effect of General Relativity

In the early years after the advent of General Relativity, there was much interest in
having this deflection measured. Would the observations confirm General Relativity
or would Newton’s gravity theory hold sway? However, there was a major obstacle
for these observations. As they pass by the limb of the Sun, the very faint light from
distant stars would not be visible to observers as the intense light from the Sun would
dominate. (Try to see the stars in broad daylight!) However at a time of total solar
eclipse with the Moon covering the bright disc of the Sun, the daylight briefly turns
to near darkness and many stars are visible in the sky. By noting the positions of the
stars in the given region, one could record their new positions during solar eclipse
and so be able to note the resulting deflections.
Fortunately, during this period, a total solar eclipse was due to occur in Brazil
and Africa. The very prominent astronomer, A. S. Eddington organized expeditions
to the ideal viewing locations to record the deflected starlight positions. While there
was some controversy surrounding the interpretations of the data (see, e.g. [5]), the
results indicated in favour of General Relativity. At the time, there was much fanfare
in the world’s press regarding the results as a triumph for General Relativity. In more
recent times, more accurate results have been claimed by observations of radio-
wave deflections from quasars. These waves can be observed in daylight which
greatly facilitates the observations. Moreover, with radio astronomy, there is the
major benefit of much higher angular resolution, particularly with very long baseline
interferometry using arrays of radio telescopes.
In the 1960s I. I. Shapiro initiated a new test of General Relativity, the “Fourth
Test”. This consisted of sending pulsed radar signals from the Earth to the inner
planets Venus and Mercury during periods of “superior conjunction”, when these
planets are located on the side of the Sun which is opposite to the position of the
Earth, as shown in Fig. 5.6.
The signals bounced off the inner planets and the time delay of the echo was
recorded back on Earth. The delay predicted by General Relativity was consistent
with the observed time delay. Even greater accuracy was achieved using interplan-
etary space probes with transponders receiving the signal from the ground station,
amplifying it and re-transmitting back to Earth with a known phase lag via the direc-
tional antennae of the space probes.
72 5 Testing Einstein’s General Relativity

5.4 Singularities and Black Holes

Let us return to the spherically symmetric spacetime in vacuum described by the

metric (5.4). A spherical body like the Sun curves the spacetime around it and the
extent of the curving is determined by the mass and compactness of the body, m:
The bigger the mass for a given compactification, the greater the curving, i.e. the
stronger the gravity. There is nothing stranger about this than what we experienced
in Newtonian gravity which hinged, mathematically, on the Newtonian gravitational
potential, m/r : The larger the m for a given r , the larger the potential. With this
expression, we also see that the smaller the r , that is the closer we approach the
body, the larger the potential. This told us that the Newtonian gravitational force,
which derives from the gradient of the potential, grows with increasing mass and
decreasing separation (Newton’s inverse square law in a different guise).
What we had not discussed before is the issue of tracking r down to the value
zero while maintaining m to be some non-zero value. With m being different from
zero, we are not speaking about universal vacuum. Consider fixing m to be a very
very tiny 1 kg spherical ball for example, and approaching it closer and closer. As
we approach from 1 m to 1/2 m to 1/4 m to 1/8 m to 1/16 m to.. etc., the potential
climbs from 1 to 2 to 4 to 8 to 16 to.. etc. When we finally let r approach to zero,
we approach a number that we cannot even write down (even larger than a googol
which is 1 followed by 100 zeros). We give it a name, “infinity” and we assign to it a
symbol “∞”. Of course our being able to contemplate this closer and closer approach
assumes that we can shrink that ball to smaller and smaller radii. Generally speaking,
when we have situations in physics such as with the potential where r goes to zero
and the potential goes to infinity, we say that we have a “singularity”. The standard
view in such cases is that physics has figuratively hit a brick wall in understanding
what is really going on, physically-speaking and we have to “go back to the drawing
boards”. A new physical theory is called for.
Now insofar as the Newtonian potential example is concerned, we could dodge the
issue by saying that infinity looms only when we approach a presumed finite mass m
having zero size and no finite mass can have zero size. However, the issues become
much more involved when we deal with Einstein’s General Relativity. There, rather
than examining the Newtonian potential m/r , we actually must confront this very
same quantity in a new context, that of the Schwarzschild spacetime expressed by
the metric (5.4). We see m/r enter here in the g00 and grr components as (1 − 2m/r )
and −1/(1 − 2m/r ) respectively. The g00 component goes to minus infinity as
r approaches zero so our former issue of dealing with infinity at zero value of r
remains. However, in addition there is a totally new issue that arises. At the non-zero
value r = 2m (r = 2Gm/c2 in terms of the conventional coordinates), the g00 is
zero and the grr is minus infinity.
In the earlier days of General Relativity right up to as late as the 1960s, physicists
referred to the spherical surface of radius 2m as the “Schwarzschild singularity”.
In a very early paper, Einstein had considered a spherically symmetric swarm of
test particles. He showed that as the sphere contracted, the particles moved with
5.4 Singularities and Black Holes 73

increasing speed, finally approaching the speed of light as the sphere shrunk towards
the size r = 2m. Thus, Einstein saw the r = 2m surface as a barrier that could never
be breached. Years later, Synge [9] showed that such test particles do not actually lie
on the light cone as they cross r = 2m heading inwards to smaller distances from
the center. By not lying on the light cone, they are not violating the key tenet of
Relativity that regular material particles can never attain the speed of light.
Through the later years, Eddington followed by G. Szekeres, D. Finkelstein and
M. Kruskal, presented new systems of coordinates in which there was no appar-
ent singular aspect at r = 2m in terms of the new coordinates. This underlined
what Synge had shown from a different perspective. However, a new aspect was
introduced in the process. These new systems of coordinates showed the spacetime
as being time-dependent, yet there is nothing really changing at all when we are
at radial coordinate r bigger than 2m. There is no harm in this and it is perfectly
understandable. For example, in the system of coordinates displayed in [3] (see also
[1]), the coordinates used are those that are actually attached to observers who are
freely-falling. To help visualize such a “comoving” coordinate system, imagine, for
example, a grid with an observer stationed for all time at every corner of the grid.
Now imagine that this grid is being expanded uniformly. Each observer maintains
his grid coordinates even though the distance between every pair of observers grows
with time.
Thus, for a person who is falling freely inwards toward the attracting body, his
newly acquired r coordinate, let us call it r ∗ , always takes on the same value. The fact
that he is continually getting closer and closer to the body is displayed by his ever-
shrinking proper distance to the body.5 Thus, we can appreciate that the spacetime
metric as expressed in terms of this new r ∗ radial coordinate, is necessarily time-
dependent.
Now there are some very interesting, albeit somewhat bizarre phenomena that
emerge when we study this in mathematical detail (see for example [1]). First, while
we can switch from the r ∗ time-dependent representation back to the familiar r radial
coordinate description of the spacetime for r values greater than 2m (and thereby
assure ourselves that the spacetime is really intrinsically static), this is no longer
the case when r has the value 2m or less. For those values, the spacetime becomes
intrinsically dynamic! No coordinate system can ever be found for which this interior
region is time-independent.
What does this entail physically? Let us build a scenario to appreciate it. Suppose
a spherical star6 has reached a point of contracting whereby its radius has shrunk to
the value 2m. In that case, the standard picture is that the star will enter a phase of
complete “gravitational collapse”. All elements of the star will move radially inward
and nothing can stop this process. Nothing can emerge from the interior r less than 2m
values to the exterior r values equal or greater than 2m, not even light. A “black hole”

5This is a reminder to us of the importance of proper measures in Relativity.

6 For simplicity, we consider here the case of a non-rotating body. Rotation injects additional
complexity.
74 5 Testing Einstein’s General Relativity

is said to have been formed, “black” because no light can emerge.7 Now it is well
to ask what happens when the star’s elements reach the center. All would agree that
the r = 0 point is singular. The detailed analysis tells us that these elements cannot
simply stop there because to do so would entail the stopping of time itself! Some have
suggested a rescue for the situation by imagining that this material re-emerges as a
“white hole” in another universe. The white hole is the black hole with the direction
of time being reversed, i.e. running the movie backwards. In other words, instead of
continually and necessarily collapsing, the white hole is the situation of continually
and necessarily expanding. Others have theorized with a different conclusion: Prior
to reaching the very center of the star under gravitational collapse, when a sufficiently
high density of concentration is reached, it is no longer legitimate to treat the material
of the star as having its physical behaviour described by classical General Relativity.
The argument is that a new yet-unknown theory of quantum gravity will eventually
be found which will save the day, so to speak.
Let us consider the r = 2m surface from another vantage point, that of the observer
who is falling in towards this surface. When we examine the situation from his point
of view, we can gauge his options at each stage. The detailed analysis reveals that
the light cones at each point of his position on his path change their shape as he falls
inwards. Recall that no matter what he does, continue to fall freely or turn on the
retro-rockets of his space-ship to resist the fall, his spacetime trajectory must always
stay within the light cone. He can never even coincide with a light trajectory which
lies on the light cone, let alone proceed beyond the light cone. When he is outside
the r = 2m surface, as the Fig. 5.7 shows, he has options. He can continue to fall
ever closer to r = 2m or he can blast his way towards larger r values, i.e. he can
escape the awaiting doom. There exists a wedge of opportunity between the light cone
boundary that he cannot cross and the trajectories that head towards larger values of r .
However should he dally and decide to act when he reaches r = 2m, he is too late.
To even maintain his r = 2m value, his spacetime trajectory is one of lying right on
the light cone. However only light itself can do this, not our courageous explorer!
So at that point, he must fall in, crossing through the r = 2m surface and continuing
towards ever-smaller r values. In fact after crossing the r = 2m surface, even light
itself must proceed to smaller r values. Therefore our hapless explorer cannot even
beam outward to his family and friends back home that this is “goodbye” forever.
The r = 2m surface acts like a kind of horizon, like the one that we observe when
we look out at the sea, where boats fade from view. Now since we are dealing here
with events in spacetime, the r = 2m surface has come to be known as the “event
horizon”.
It is actually somewhat confusing to speak of a value of r when it is smaller than 2m
as being a “where” since r then becomes a time-like coordinate rather than a space-
like coordinate. To make sense of it, it is best to follow the process from the point of
view of the coordinate system attached to the falling observers. Because of the bizarre
circumstances that ensue for r = 2m and (inwards) beyond, Einstein and Rosen

7 S. Hawking has theorized that quantum-mechanically, black holes will actually evaporate particles
that emerge to cross out of the r = 2m sphere but the process is very slow for stellar-sized bodies.
5.4 Singularities and Black Holes 75

r=0

R
r<2m Escape

r=2m

r>2m

Fig. 5.7 The vertical line with constant value of R represents the path in spacetime of a freely-falling
observer and the r = constant 45◦ line represents the trajectory of an observer who maintains a
fixed distance relative to the central body. Light cones are drawn at some points along the path of
the freely-falling observer. The vertical axis is the time t axis and hence the history of an observer
in spacetime is one that proceeds upwards in the diagram, i.e. with time advancing. Unlike the
cases in Special Relativity, here using General Relativity, the light cones become thinner as r = 0 is
approached. Before r = 2m is reached, an escape hatch is available. This is indicated in the figure
as the zone between the r = constant 45◦ line and the line tangent to the light cone. Motion in
that zone is physically possible as it lies within the light cone and because it comprises directions
with slopes exceeding 45◦ , it entails motion away from the body, achieving increasing r values.
The observer could use rocket power to escape at that stage of his motion

developed a different kind of solution, one in which a copy of the Schwarzschild

solution was joined to the existing one precisely at r = 2m. In this manner, the
awkward region within r = 2m was entirely excised out of existence. This join
region or “throat” came to be known as the “Einstein-Rosen Bridge” (Fig. 5.8). These
authors did not waver in their opposition to the idea that black holes are elements
of physical reality, and their construction of the interior-free solution was an effort
towards building a picture in tune with their conception of reality. In their view,
r = 2m was a physical singularity while the prevailing view is that it is merely a
“coordinate singularity”, an artifact of an inferior choice of coordinate system. In
fact the prevailing view is that black holes are common occurrences of nature at the
centers of galaxies as well as remnants of the collapse of very massive stars. If it
should be the case that nature can present us with concentrations of mass for which
further compression cannot be resisted and an event horizon can be reached, then the
prevailing view has merit. However, it is conceivable that nature conspires to resist
such formations. For example, we witness the phenomena of supernova explosions
76 5 Testing Einstein’s General Relativity

Fig. 5.8 The figure depicts

the Einstein–Rosen bridge.
Two Schwarzschild space-
times are joined at the “throat”,
thus averting the troublesome
region interior to the r = 2m
surface

r=2m

wherein vast quantities of matter are blown away as certain massive stars collapse.
Could these be instances of nature resisting complete gravitational collapse?
While it would appear that most researchers have become comfortable with both
the theory and the reality of black holes, singularities are another matter. However, we
discussed the problem of complete gravitational collapse which seems to indicate
the inevitability of the eventual formation of an indisputable singularity, the kind
related to having r go to zero while m remains non-zero in the instances where the
expression m/r appears. So distasteful is the idea of such a singularity that the noted
researcher, R. Penrose, devised an ingenious solution, now referred to as the “Cosmic
Censorship Hypothesis”. He conjectured that in the process of complete gravitational
collapse, nature, acting as a censor, conspires to cast off any non-spherical elements
as the body collapses so that what remains is an event horizon at r = 2m. The casting
aside of such elements is necessary for his hypothesis, since it is only the spherically
symmetric Schwarzschild solution among static vacuum spacetimes that allows the
existence of a non-singular event horizon. For Penrose, the day is saved thereby since
no information can be leaked to outside observers about the inevitable singularity
that forms at r = 0. Outside observers are shielded from this embarrassment of a
breakdown in physics.
However, there are problems with Penrose’s solution to the problem. First, while
outside physicists are spared the anguish of dealing with this now-hidden or “clothed”
singularity, what about the physicist who falls in? Surely it would be wrong to dismiss
him merely because he ultimately gets crushed to death. During the period that he
is an observer, he is not shielded. Second, we have shown that even in spherical
collapse, if the matter has an exotic equation of state of the form P = kρ where
k is a constant lying between −1 and −1/3 (P is the pressure and ρ is the density),
a “naked” singularity will form [10]. A naked singularity, as the name implies,
5.4 Singularities and Black Holes 77

is one that is not hidden from view. Outside observers can witness this singularity.
While these types of equations of state for matter leading to the formation of naked
singularities are not for the kind of matter that, as Bondi used to say, “you can buy
in the shops”, nevertheless they have an attraction all of their own. They are of the
types that are believed to be behind the inflationary birth of the universe. Could there
be some deeper connection? We are always on the hunt for clues that guide us to new
insights.
Chapter 6
Gravitational Waves and Energy-Momentum

6.1 Introduction

In earlier chapters, we discussed sound waves, the actions of vibrating atoms and
molecules passing along their motions in a linear chain which finally elicit vibrations
in our ear drums. Our brains interpret these as sound, the information by which we
communicate with each other and the pleasure that we gather from great vocalists
and musical instruments that generate these waves in a skillful, blended manner.
We also discussed electromagnetic waves, the transversely propagating electromag-
netic fields which enter into every phase of our existence. At the most fundamental
level, we understand their origin as deriving from the acceleration of a charge. Wiggle
a charge back and forth and an electromagnetic wave is generated, flowing at speed
c in vacuum away from the source. Depending on the frequency of oscillation, these
waves create noise in our radios, warmed food in our microwave ovens, heat in our
skins, vision in our eyes, etc.
As charge is to electromagnetism, mass is to gravitation. Thus it was natural for
Einstein to conjecture the existence of gravitational waves in analogy with elec-
tromagnetic waves. While we will be discussing various similarities to which one
can point, it is well to consider their overall essential difference: an electromagnetic
wave is a distinct field disturbance propagating in spacetime. This field consists of
oscillating electric and magnetic components which are perpendicular to the direc-
tion of the wave propagation direction. A gravitational wave is also an oscillating
propagating disturbance but the disturbance is of spacetime itself! It is not an add-on
to spacetime. For gravitational waves, it was the acceleration of a mass rather than
the acceleration of a charge that would be the source of its production. Moreover,
there was a more fundamental reason calling for the existence of gravitational waves.
After all, whether in Newton’s gravity or in Einstein’s gravity, the form of the grav-
itational field is dependent upon the distribution of mass: changing this distribution,
even by the simple act of raising one’s hands in the air, must lead to changes in the
gravitational field throughout the universe. Moreover, to be consistent with Special
Relativity, this change must not be pasted onto the entire universe instantaneously,

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 79

DOI: 10.1007/978-3-642-30385-2_6, © Springer-Verlag Berlin Heidelberg 2012
80 6 Gravitational Waves and Energy-Momentum

for to do so would violate the basic tenet of a finite velocity of propagation of interac-
tions. This is where Newton and Einstein follow separate paths. It is at this juncture
that for Einstein, the concept of a wave of spreading gravity arises, making changes to
the distribution of matter realized in ever-increasing distances from the disturbance
as time passes.

6.2 Gravitational Waves in Einstein’s Theory

The beauty of Einstein’s theory of gravity is manifest in the manner in which these
fundamental ideas that we discussed above blend together so naturally. To set the
stage, we return to Newton’s gravity and its fundamental equation connecting source
of gravity, mass density ρ to gravitational potential φ, namely (4.3). From this equa-
tion, Newtonian gravity assures that any change in the distribution of mass density,
even the dropping of a pin, affects the gravitational potential instantaneously through-
out the universe. This is the nature of Newton’s gravity equation. It derives from a
very simple theory. The mathematics dictates this behaviour which runs counter to
the highly successful theory of Special Relativity.
By contrast to Newton, the much richer Einstein gravity theory replaces the
Newtonian gravitational potential by the metric tensor gik of curved spacetime, and
the mass density source of gravity by the full energy-momentum tensor Tik of all of
the matter, stresses and fields other than gravity which are present. The Einstein field
equation that replaces the Newton field equation is (4.6). Under the right circum-
stances, when the gravitational field is weak, and a particular choice of coordinates
is made, the Einstein field equation becomes very similar, in approximation, to the
Newtonian equation (4.3). This is as it must be because we must always remember
that Newtonian gravity theory works very well to a high degree of accuracy in many
situations. In the Einstein case, the metric tensor component g00 replaces φ and there
is a very important addition of a calculus time-rate-of-change operator. This operator
is responsible for making the changes to the matter distribution relay the information
to the field in the form of a propagating wave. The wave moves outwards from the
source at speed c in vacuum, in various ways analogous to the case in electromag-
netism. The idealized sources of gravity waves first discussed by Einstein, Eddington
and others nearly a century ago were simple mass oscillators (masses affixed to the
ends of a spring and oscillating in a line) and rotating rods.
At first glance, given the analogy with electromagnetism, and given that the obser-
vation of electromagnetic waves produced by analogous mechanisms with charge
sources are the routines of our experiences, one might be inclined to believe that
these gravity waves would likewise be readily observed. However, in the almost
century since their first consideration, there has never been a direct detection of a
gravitational wave produced by a source in any laboratory and their only indicated
physical existence derives from the very tiny changes in the periods of some care-
fully observed binary pulsar systems such as the first ever observed, PSR1913+16.
The generally-held view is that these period changes are directly connected to a
6.2 Gravitational Waves in Einstein’s Theory 81

loss of energy from the system, energy that is carried away by the gravitational
waves that are being generated. We will have more to say about this topic in what
follows.
Given that there is such a fundamental phenomenon in the gravity wave concept
and given the paucity of evidence for its very existence, it is quite understandable
that many papers and discussions have been devoted to gravity waves over the years.
A small number of researchers have even questioned their very existence. For reasons
discussed earlier, we feel the necessity for their existence. As to why they have never
yet been observed in an experimental set-up, a convincing case is made on the basis
of two essential differences from the electromagnetic wave detection case. First,
the gravitational interaction is inherently very weak. It is mediated by the constant
of universal gravitation, G, which is a very small number in conventional units.
As an example of what this entails, it is of interest to note that the electrostatic
force between two electrons is on the order 1045 times larger than their gravitational
interaction. That is 1 followed by 45 zeros! It is worth contemplating the enormity of
this number to help us appreciate the impact that this would naturally entail. Second,
with regard to the type of wave of strongest possible order, in the electromagnetic
case it is the “dipole” wave whereas in the gravitational case it is the “quadrupole”
wave.1 To elaborate on their distinction would take us beyond the scope of this
book. Suffice it to say that the latter are much weaker than the former. Thus, we
are prepared for a challenge in dealing with gravitational waves on an observational
basis.

6.3 The Energy Issue

The concept of energy in General Relativity has been problematic from the earliest
years. Energy and its conservation has been a vital anchor for physics in general.
When there arose a situation in which energy was seen not to be conserved, a new
form of energy would be identified as the missing link and energy conservation would
be restored. In classical physics we have energy of heat, of sound, of compression,
energy stored in the electromagnetic field, in the gravitational field, etc. In Special
Relativity, we discussed the inherent energy in a mass m given by mc2 . We have
become quite comfortable with energy as a concept, particularly when it is nicely
localized, as it is within the location of the mass in Special Relativity.
However, it is somewhat less attractive as a concept when we have to deal with
the energy in Newtonian gravity. For example when two bodies with masses m
and M are moving under their mutual gravitational attraction, a conservation law
for energy is extracted from the sum of two kinds of energies, the kinetic energies
of the bodies, 21 mv 2 and 21 M V 2 where v, V are the velocities of the bodies and a
potential energy—Gm M/r attributed to gravity. While we can feel comfortable with
the “where?” question for the kinetic energies (“they are right where the masses are”),

1 The mathematically inclined reader can read about these “multipole” aspects in [3].
82 6 Gravitational Waves and Energy-Momentum

the potential energy is not where the bodies are located but rather it is spread out all
over space. Repeated exposure to the idea tends to induce a certain level of comfort
(after all, those Newtonian mechanics homework problems work out so nicely) but
when we step back and contemplate the situation more critically, a sense of unease
can set in quite readily. Can we find a more intuitively satisfying approach?
We have hypothesized that energy in General Relativity, including the contribution
from gravity, is localized in the regions of the energy-momentum tensor Tik [11].
Recall that the energy-momentum tensor is the mathematical construct that embodies
all of the different forms of energy except the contribution from gravity itself. In other
words, by the hypothesis, energy is where the “stuff” is and the stuff includes not only
the matter that we are used to, like rocks and water but also non-gravitational fields
such as the electromagnetic field. Our hypothesis: the contribution from gravity is
there as well.
You might object to this hypothesis on the basis that electromagnetic fields are
spread out just as are gravitational fields, but there is the essential difference that
is worth repeating: in General Relativity, all matter and non-gravitational fields are
contained within spacetime but gravity is spacetime, manifested by its curvature.
All of the standard “stuff” is an add-on to spacetime but gravity is not an add-on to
spacetime-it is spacetime. Recognizing the difference makes it easier to accept the
idea of such a localization. Localizing all the energy where the energy-momentum
tensor is non-zero requires some fundamental re-thinking along the lines that the
energy aspect attributable to gravity could well play a different role. You might feel
it natural to propose that the location of gravitational energy is where the spacetime
curvature is present, but we will provide reasons as to why this is not really the
correct answer. The choice of localization in the regions of non-vanishing Tik calls
upon us to face the perhaps uncomfortable idea of a gravitational wave propagating
in vacuum while not carrying any energy with it. How did we arrive at the idea of
such a localization?
1. In both electromagnetism and gravitation, we considered the simplest possible
waves, plane waves, waves that have the same structure and magnitude over
a plane. Plane electromagnetic waves are described by the energy-momentum
tensor Tik and as a result, their energy density is always positive, regardless
of the coordinate system in which we choose to examine them. They convey
a force when they impinge upon a wall (“radiation pressure”). They enter our
eyes, causing an excitation in our brains and as a result, we see. Their reality
as an energy-carrying physical construct is indisputable. Compare this to the
simplest possible gravitational waves, plane gravitational waves. These waves
have their energy described by what is referred to as the “energy-momentum
pseudotensor” tik . As the name implies, a pseudotensor lacks the important tensor
property of “covariance”, that is having a well-defined structure in every system
of coordinates and which, if non-zero in any given frame, is non-zero in every
frame. By contrast, the pseudotensor tik of the gravitational field can be made to
be zero at any spacetime point that one might choose by an appropriate choice of
coordinate system. And with plane gravitational waves, it is even more dramatic:
6.3 The Energy Issue 83

for them, one can choose a coordinate system in which this pseudotensor is zero
everywhere in spacetime! [12] This fact makes it easier to accept the localization
hypothesis. After all, how can one be confident of an energy of gravity in vacuum
if the construct that is traditionally used to measure it can be totally wiped
out for plane gravity waves simply by making the correct choice of coordinate
system?
2. We considered the work of W.B. Bonnor [13] who analyzed General Relativity
solutions for dust clouds that collapse without spherical symmetry, solutions
found by P. Szekeres. Bonnor was able to match these Szekeres solutions to
the Schwarzschild metric asymptotically, implying that the region within has a
constant mass given by the m in the Schwarzschild solution. However, by the
traditional approach to energy in General Relativity, such a collapsing cloud,
moving without spherical symmetry and thus having a changing mass quadru-
pole moment, must be pumping out energy into the vacuum beyond the dust
cloud. Again, this is problematic for the standard picture of energy in General
Relativity.
3. While many researchers have been convinced that gravity waves carry energy
because of the observed period change of the binary pulsar, there is a more fun-
damental alternative explanation: over half a century ago, in a largely unknown
but important work, A. Papapetrou [14] showed that the field equations of
General Relativity do not allow for the existence of purely periodic solutions.
On this basis, the period-changing binary pulsar is simply manifesting its con-
formity with the mathematical demands of Einstein’s General Relativity rather
than the preconceptions regarding energy. We can view this as another indication
of Einstein’s Relativity being the ultimate key to the cosmos.
4. By our localization hypothesis, being connected as to location with the tensorial
matter energy-momentum Tik , gives the contribution from gravity to energy a
tensorial basis. We will have more to say in this regard in what follows.
It is well to step back and consider what objections could be raised to this local-
ization hypothesis:
1. The need for a graviton.
Quantum theorists, working at the sub-microscopic level, see the world in terms
of quanta of energy and angular momentum, tiny discrete units associated with
elementary particles and fields. At the level of dimensions 10−10 m and smaller,
nature is no longer seen as a continuum but rather one of lumpiness, and what was
certainty of position and velocity is replaced by a bizarre world of uncertainty
and probabilities. The quantum theory is wonderfully successful in practice, par-
ticularly as applied to electromagnetism, even though some of its most promi-
nent practitioners admit to being at a loss to fully comprehend its fundamental
underpinnings. The quanta of the electromagnetic field, called “photons”, have
discrete angular momentum  = h/2π, h being Planck’s constant, and energy
hν, ν being the frequency of the underlying electromagnetic wave. This wave
is constructed from a vector potential Ai . Because their angular momenta are a
84 6 Gravitational Waves and Energy-Momentum

single unit of , they are referred to as “spin 1” particles.2 While in some experi-
ments, electromagnetic waves behave as one would expect for wave phenomena,
in various other experiments, the photons that are the quanta of the underlying
waves behave like little balls with energy as we would visualize from classical
mechanics.
The tensor that underlies the gravitational field is the metric tensor gik and
because it has two indices i and k, quantum theorists ascribe a quantum to the
gravitational field of “spin 2”, i.e. an angular momentum of 2 and an energy
2hν where ν is the frequency of the underlying gravitational wave. This quantum
has been given the name “graviton”.
While this is a logical continuation from the quantum theorist’s natural view-
point, it neglects the essential distinguishing fact that while all of nature’s parti-
cles and fields are inhabitants of spacetime, gravity is spacetime. Thus, we should
not be led to necessarily assume that the gravitational field must be quantized,
as most particle theorists and others would have it. In fact there is no experiment
that shows such a necessity. This is in stark contrast to the situation in elec-
tromagnetism wherein, at the turn of the twentieth century, there were various
fundamental problems that were in contradiction with classical (non-quantized)
electromagnetism:
(a) The photoelectric effect, the phenomenon of electrons ejected from a metal
plate by electromagnetic radiation shone upon it.
Regardless of the intensity of the incident radiation, no electrons were ever
emitted unless a threshold frequency was used and even the weakest beam
with the right frequency would suffice to eject electrons. Such behaviour is
inconsistent with classical electromagnetism.
(b) The “ultraviolet catastrophe”.
The spectrum of radiation intensity versus wavelength from a heated black-
body, dropped precipitously at ultraviolet frequencies rather than increasing
steadily as one would expect on the basis of classical electromagnetism.
(c) The spectrum of quantized electromagnetic emissions from atoms.
Discrete frequencies rather than a continuum of frequencies, were observed,
again not as one would expect from classical electromagnetism.
From these phenomena, it was clear that classical electromagnetism was
left wanting from the ultimate physical arbiter: the experiment. Physicists
can invent the most beautiful of theories, theories built upon the most elegant
of mathematics, but if nature does not agree with its predictions, the theories
rather than the experiments must be discarded. Thus it came to pass that
classical electrodynamics was confined as a theory for the beyond-atomic
domain and had to be supplanted, ultimately with quantum electrodynamics,
in the atomic and sub-atomic domains.

2What we ordinarily think of fundamentally as particles, such as electrons which are also quantized
entities having “spin 1/2”, i.e. intrinsic angular momentum /2, actually can also be regarded as
waves. This wave-to-particle association is referred to as “wave-particle duality”, an insight tracing
back to L. de Broglie.
6.3 The Energy Issue 85

By contrast, when it comes to gravity, while some non-gravitational quan-

tum phenomena can be sufficiently influenced by the gravitational field to
show up in an experiment,3 there is no experiment that tells us that gravity
itself must be quantized. The latter would be a leap of faith driven by a
quantum theorist’s natural predilection rather than a logically-driven pro-
gression.
2. The Feynman thought experiment
R. P. Feynman devised a thought experiment to lend credence to the view that
gravitational waves carry energy. He considered a pair of rings threaded through
a bar, free to slide with friction and confronting an incoming gravitational wave
train. Noting that free particles have their relative distance periodically changed if
they are oriented perpendicularly to an incident gravity wave, Feynman reasoned
that the ring separation would also vary periodically. Thus, with the rings sliding
along the rod under friction as their interspacing varies, heat would be generated.
After the wave train passes, the net effect would be a bar at a higher temperature
than previously, confirming that a transfer of energy has ensued from the waves
to the bar.
At first glance this certainly sounds like a convincing argument, but what has
been overlooked is that the bar itself has been presumed to be unaffected by
the gravity waves, that the bar, in effect, is a perfectly rigid body. Our earlier
discussion laid such a notion to rest. Nothing escapes gravity because gravity is
spacetime and everything inhabits spacetime, including the bar. Thus we must
inquire as to how the bar is affected by the gravity waves. In spite of the bar
not being perfectly rigid, (you could think of it as an extremely taut spring with
a sufficiently strong pulling-apart, it will lengthen ever so slightly and it will
spring back upon release) it is well to ask if there would be relative motion
between the rings and the bar nevertheless. To test this, we replaced the bar by
an idealized elastic medium modeled electromagnetically as follows: consider
a section of the bar to be modeled as two parallel equal and oppositely charged
capacitor plates with a high frequency electromagnetic wave bounced between
them. The opposite charges provide Coulomb attraction while the bouncing
waves provide electromagnetic radiation pressure. With the correct choices of
charges and wave intensity, the forces balance and the system is in equilibrium.
Changing the strengths of the balancing elements changes the degree of rigidity
that is being modeled.
It is readily shown, moreover, that the system behaves like an oscillator. When
the plates are pushed together and released, the increase in radiation pressure
forces the plates back. They overshoot their equilibrium position and as the
pressure diminishes, the Coulomb attraction brings the plates back towards each
other [15]. In this manner, we have a simulation of an elastic medium that we can
analyze with the Einstein equations in conjunction with the Maxwell equations

3 This should not be seen as at all surprising: after all, quantum phenomena are aspects of the
ingredients of our universe and they would most logically be regarded as being affected by their
habitat, namely spacetime.
86 6 Gravitational Waves and Energy-Momentum

of electromagnetism. The result of the analysis reveals that the plates’ separation
follows the same distance variation as do two free masses at positions adjacent
to the plates. The result is independent of the choice of balancing elements.
Considering the free masses as the rings and the electromagnetic oscillator as
the section of the rod, we conclude that there is no relative motion, i.e. no rubbing,
hence no energy transfer.

6.4 Can Energy Be Localized?

Prior to the advent of General Relativity, the issue of whether or not energy can
be localized did not receive much attention. While we discussed earlier the spread
throughout space of the negative Newtonian gravitational potential energy, it was
generally viewed as non-problematic, perhaps because its gradient gave the correct
force exerted on a body within the framework of classical physics. Some interesting
energy issues were discussed by Feynman [4] and were dealt with admirably in his
characteristically engaging style, but no remaining issue outside of General Relativity
raised any alarm bells. The most enduring source of difficulty and controversy arose
in connection with energy-momentum in General Relativity.
Earlier we discussed the energy-momentum pseudotensor, the problem-plagued
construct that had been used for gravity in General Relativity to parallel the suc-
cessful energy-momentum tensor that had served and continues to serve for the
rest of physics. Bondi [16] introduced an energy flux construct that he named the
“News Function” as a supposed antidote to the pseudotensorial affliction that plagued
General Relativity but Madore [17] and we independently [18] for a particular
case, demonstrated the equivalence of the News Function to the energy-momentum
pseudotensor. Thus, the Bondi construct did not provide the solution that was hoped
for. Years later, Bondi [19] affirmed that energy in General Relativity must be local-
izable. In the years that followed, various authors arrived at a kind of intermediate
stance, declaring that energy in General Relativity is “quasi-localizable”. They devel-
oped rather complicated structures, structures plagued with problems of their own.
One might be inclined to regard the notion of quasi-localization of energy as akin
to quasi-pregnancy of a woman. Such is the nature of various solutions proposed at
times by physicists in difficult situations to set their minds at ease.
We kept the faith in our localization hypothesis but what remained lacking was a
tensorial construction that would serve us for General Relativity. Recently, M. Dupre
[20] proposed that the Ricci tensor Rik was the required tensor. This was particularly
attractive in that it is a tensor which we know, from the Einstein field equations,
is different from zero only where the energy-momentum tensor Tik is different from
zero, in accord with the localization hypothesis. In what follows, we will follow
a sequence of steps that led us to what we regard as the final forms for energy-
momentum in General Relativity [21]. It is instructive to do so in order to give you,
the reader, the nature of the thinking that goes on in physics research. However,
6.4 Can Energy Be Localized? 87

some more sophisticated mathematical concepts are involved and the reader who is
unprepared for such is advised to proceed directly to Chap. 7.
We first ask if there is some aspect that would propel us to pursue the Dupre
proposal. There is, in that it has an immediate appeal, knowing that Tolman [22]
had shown many years ago that a stationary or static system with energy-momentum
tensor Tik has total energy, including the contribution from gravity, of value

√
E= (T00 − T11 − T22 − T33 ) −gd xd ydz (6.1)

where −g is the positive value of the determinant of the complete four-dimensional

spacetime metric tensor (g itself being negative for spacetime). This is the standard
form in which the Tolman formula is given, but by the Einstein field equations (4.6),
this can be expressed in the simpler form

c4 √
E= R00 −gd xd ydz. (6.2)
4πG

Thus, with the appearance of the Ricci R00 component, we see that the Ricci tensor
is a natural candidate for energy-momentum in General Relativity.
Now in (6.2), we have only the energy and we have only one component of a
tensor, namely R00 . We seek tensor expressions and we want more than just energy;
we also wish to incorporate momentum. The minimal way to achieve this and still
maintain the Tolman energy expression for the special case of stationary or static
systems is to create a new tensor-like expression

c4 √
E ik = Rik −gd xd ydz. (6.3)
4πG

From (6.3), we see that the component E 00 is precisely the Tolman expression for
energy.
This is a good start but when we look further, we see an unusual mixture of ele-
ments. This concerns the aspects of volume. Earlier we spoke about proper measure
of distance, the truly physical measure. When we extend this idea to three dimensions,
we can express the truly physical element of spatial volume, the “proper volume”
d V p which can be written in terms of “coordinate volume” d V as
√
d Vp = 3 gd V (6.4)

where 3 g is the determinant of the 3-space part of the metric and where d V
is the “coordinate volume”, d xd ydz when we are using Cartesian coordinates,
r 2 sin θdr dθdφ when we are using spherical polar coordinates, etc. In curved space-
time, d V is not the volume that is occupied physically, rather it is d V p .
88 6 Gravitational Waves and Energy-Momentum

When we examine (6.3), we see the coordinate three-dimensional volume d V =

√
d xd ydz but it is multiplied by −g, i.e. the wrong metric determinant for three-
dimensional space. This led us to extend the concept of energy-momentum in General
Relativity, introducing a new construct, “spacetime energy-momentum” E ∗ ik ,

c4 √
E ∗ ik = Rik −gd 4 x (6.5)
4πG

where d 4 x is the four-dimensional spacetime coordinate-volume element. In Carte-

sian coordinates, d 4 x = d xd ydzd(ct), in spherical polar coordinates, d 4 x =
r 2 sin θdr dθdφd(ct), etc. It was the mathematics that was guiding us to the new
concept.
While at first glance it might appear to be a radical step, in fact it is rather benign
when considered more closely. For a stationary system, the component E ∗ 00 is simply
the Tolman energy multiplied by the time that it is being observed. However the new
concept assumes a more interesting role for dynamic systems which we will discuss
in what follows.
In E ∗ ik , we have an integral of a four-tensor, Rik over proper four-volume which
is as tensorial as one can get for an integral. Only tensorial elements are involved in
its construction. While it reduces to the Tolman energy times the time of observation
for stationary systems, it is well to ask whether it is merely the end result of a more
complicated expression in the general case of a dynamic system. The answer is “no”.
For suppose we were to have a more complicated tensorial form in general with
jk
Rik replaced by, for example Rik + Si j + · · · where the dots indicate other second-
rank add-on tensors. However, these add-on tensors must vanish for the particular
cases of static or stationary systems in order to agree with Tolman’s expression, and
hence they must be of the form of a partial derivative in time, ∂.. ∂t . While we can
certainly construct forms of such S-type tensors (or variations thereof), in the form
of covariant derivatives (to keep them tensorial) which will indeed have the required
partial time-derivatives, they necessarily have in addition, partial space derivatives
that do not vanish when particularized to non-dynamic systems. This contradicts the
Tolman form and hence the Ricci tensor Rik is the unique possibility [21].
While (6.5) is as far as we can go with an integral incorporating covariant ele-
ments,4 we can go one step further in expressing spacetime energy as a four-scalar.
We do so as a logical progression from Special Relativity in expressing the energy
of a body with four-momentum pi . Suppose that body is viewed by an observer
with four-velocity u i . Then the energy observed can be expressed in (four-scalar)
invariant form as
E = pi u i . (6.6)

This E has the familiar value mc2 / 1 − v 2 /c2 where v is the relative speed between
the observer and the body and m is its mass. Clearly, if the observer is chosen to be

4 Being an integral of a tensor makes it non-transformable as a second rank tensor.

6.4 Can Energy Be Localized? 89

comoving with the body, v = 0, and then we have the Einstein iconic expression
for E, namely E = mc2 .
Again, we follow the natural course for spacetime energy. We express it in the
invariant form 
c4 √
E∗ = Rik u i u k −gd 4 x (6.7)
4πG

where u i is the four-velocity of the observer as before only now it is with a field of
observers. In [21], we have developed this further with the example of the gravita-
tional collapse of a spherical ball of dust.
We have followed through with this new approach to energy-momentum in
General Relativity, partially for interest in its own right, but also as an example
for the reader of how we search for clues to advance research. The process is seldom
straightforward. The path is generally laden with obstacles and curious twists but the
reward, well-worth the effort, lies in the hoped-for advancement of understanding
the nature of the universe and its workings.
Chapter 7
The Universe According to Relativity

7.1 Getting Acquainted with the Elements of the Cosmos

In the previous chapters, we have already witnessed some of the greatest marvels
that Einstein’s Relativity has provided for us. In this chapter, we will describe one
marvel that many consider the greatest of all: the power of Relativity to describe the
overall structure, evolution and ultimate fate of the entire universe. Can one imagine
a more profound, weighty, intellectual achievement? To get us started, we will try to
give the reader a real feeling for the basic elements and incredible distances that we
deal with in cosmology.
We begin with the Earth, our home, the third planet from the Sun. It is a wondrous
beautiful place that may harbour the only life in our Solar System, the system com-
prising the Sun plus the family of bodies that circle the Sun. Many of us have taken
continent-size trips of approximately 4,000 km. Having a feel for this great distance,
it is not a difficult stretch to contemplate the actual size of the Earth. A trip around
the Earth is approximately 40,000 km in distance, about 10 times the continental-
sized trip. Around the Earth we have the circling Moon. It lies at a distance of about
400,000 km from the Earth, about 10 times the already great distance for circum-
navigating the Earth. Now let us contemplate that flaming hulk that is our Sun, the
behemoth that keeps us alive with its tremendous radiance. It is so big that the Earth
with our Moon in orbit about us could easily fit inside a sphere the size of the Sun!
To go beyond this, we use a technique that gives us a feeling for the next level of
distances and sizes, the technique of scaling. Imagine that we were so big that the
Sun would appear to us as a tiny burning pebble the size of the head of a pin. On that
basis, how big would the Solar System appear to us? The answer might be surprising:
at this pinhead-sized scale for the Sun, Pluto (now no longer regarded as a planet) at
the outer edge of the Solar System would appear as a microscopic-sized speck 20 m
from the pin-sized Sun. The other planets would appear as specks in-between with
the larger planets barely visible to the naked eye. If you ever wondered why we were
not undergoing collisions with the other planets, you can now appreciate why this is

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 91

DOI: 10.1007/978-3-642-30385-2_7, © Springer-Verlag Berlin Heidelberg 2012
92 7 The Universe According to Relativity

the case. There is a lot of space available in which those little specks can fly around
safely without colliding with each other.
The usefulness of the pin-head scale continues. The Sun is a rather unexceptional
star and the universe is teeming with stars. It is well to consider on the pin-head scale,
where the closest star is situated relative to our Sun. The answer might astound you:
the nearest pin-head sized star is about 50 km away! This gives you a picture of just
how great a challenge it would be to visit a solar system nearby to that of our own.
It also illustrates why we are not seeing stars smashing into each other as a common
occurrence: there is a lot of room out there in which these stars can move!
Staying with the pin-head Sun scale, we next consider the galaxy which is the
basic building block of the cosmos. Galaxies are conglomerations of literally billions
of stars, assuming a variety of shapes, often beautifully symmetric disc-like shapes.
At this scale, a typical galaxy would be about 300,000 km in diameter, roughly three-
quarters of the distance from the Earth to the Moon! We have learned to picture this
kind of distance, so we now have some feeling for the basic structure of a galaxy.
In order to go beyond the single galaxy, we must compress the scale even further for
visualization. We imagine that the galaxy that we have just described is incredibly
scaled down to the size of a dime, a coin roughly a centimeter across. A dime is
actually a good visual aid since like a dime, a typical galaxy has a radius about ten
times its thickness. We have frequently asked students to guess how far away the
typical neighbour to this galaxy is located on this dime scale. Answers have ranged
from 100 m to 400,000 km or more. They are invariably amazed to learn that on the
dime scale, the neighbour is typically a mere meter away! The universe is teeming
with galaxies in reasonable proximity, making it understandable that galaxy collisions
are not very rare occurrences. The observations presented to us by astronomers of
galaxies in collision are wondrous to behold.
Like the planets that avoid falling into the Sun by virtue of their circular motion
about the Sun, the galaxies generally support their structure by being in rotation
about an axis. With such an enormous structure, it should not be a surprise that a
single complete rotation takes a tremendous amount of time, typically 100 million
years. Galaxies tend to cluster in groups of widely varying numbers, from just a few
to thousands. We describe the overall distribution of galaxies in the universe as fairly
“isotropic”: looking out into all the possible directions from our vantage-point, we
see a more-or-less similar picture of averaged galaxy distribution. In what follows,
we will discuss how this fits in with current ideas about the origin of the universe.

7.2 Evolving Views on the Universe

It is well to begin with some perspective. Throughout the centuries, humankind has
entertained a variety of ideas about the nature of the universe that seem laughable to
us today. However, the ideas then current were embraced with often fairly solid belief,
even certitude. Very early-on, even some of the learned entertained the picture of a
flat Earth with oceanic water dripping off the edges. Then came the true picture of the
7.2 Evolving Views on the Universe 93

Earth as a sphere, but the then-truism was one of a central positioning and a central
importance of the Earth in the universe. The Sun and the stars were seen to revolve
around us, and the stars were considered points of light of no great consequence,
unlike the bright Sun. These ideas, that we now regard as exceptionally naïve, were
the result of very limited information. As information has increased, knowledge has
increased and many scientists regard the state of cosmological status-quo knowledge
today as very near ultimate truth and wisdom. It is well to step back and remind
ourselves that such was the attitude of many scientists at the various stages throughout
the centuries.
This brings us to a very basic question concerning the universe: is it static or
is it dynamic, evolving in time? The former view prevailed for many decades, in
fact into the first three decades of the previous century. What is surprising to us is
that even the great genius of Einstein embraced the static picture of the cosmos.
We have always found this to be amazing. After all, gravity makes bodies attract
each other and they will move unless they are held back. The basic building blocks,
the galaxies of the cosmos, are not held back, at least not by anything that is at
all apparent. Therefore it would seem eminently sensible to assume the view that
the universe should be dynamic. Elementary, it would seem. Could it have been
the desire for stability, the fear of the unknown catastrophes that one might have
imagined in the face of an evolving universe that had so many scientists fixated on
a static universe? Regardless, we are very aware of the tendency of many scientists
to cling with fierce tenacity to prevailing wisdom. That being said, Einstein can be
excused. His revolutions in other directions showed clearly that he was not one to
fear new ideas. This is well-documented in [5].
Almost 200 years ago, H. W. Olbers presented a paradox regarding the universe.
Even well before his time, the stars were recognized as suns and the universe was
regarded as a sea of these suns stretching to infinity. A simple calculation shows that
with this picture, the sky at night should be ablaze with light from the sea of starlight.
However the night sky is dark. Hence the paradox arises. In more recent times E. R.
Harrison and P. S. Wesson et al. showed that the darkness of the night sky is due
almost completely to the relatively short age of the universe counting from the time
of the Big Bang. This places a severe limit on the amount of light that the galaxies
could have produced. While the expansion and the red-shift of the light also come
into play for the dynamic universe, their effect, these authors have calculated, account
for no more than a factor of two and hence would not affect the result qualitatively.
However, red-shift is a key factor for another reason: it provides the essential
evidence that the universe is in a state of expansion. We discussed earlier how when
two observers are moving away from each other, light received by one observer nec-
essarily appears to be shifted in frequency towards the red end of the electromagnetic
spectrum in comparison to the frequency as seen by the other observer at rest relative
to the light emitter.
In 1929, E. Hubble announced that the light from the distant galaxies was red-
shifted and that the degree of red-shift that he observed was proportional to the
distance from us of the galaxy observed. With red-shift proportional to distance
and also red-shift proportional to velocity, it follows that velocity is proportional to
94 7 The Universe According to Relativity

distance: the further a distant galaxy is from us, the faster it is moving away from us.
This is codified into what is referred to as “Hubble’s Law”,

v = H D. (7.1)

In (7.1), v is the velocity of a distant galaxy relative to us and D is our distance

from that galaxy.1 The H in (7.1) is called the “Hubble constant” although strictly
speaking, it is not really constant as it evolves with the evolution of the universe.
When people refer to the value of H , they are really speaking of its present-day value.
It is more appropriate to call H the “Hubble parameter”.2
It was the observations of galactic red-shifts that overturned the notion of a static
universe. There was a widely distributed photograph of Einstein peering through a
telescope with Hubble looking on, no doubt with a great feeling of joy. The Hubble
observations convinced Einstein that he had been wrong about the static universe
idea, and he described as useless the work that he had done to produce a static model
universe. We will discuss this issue later.
We now turn to the problem of universe model-construction. To begin, we look
back in scientific history. The earliest picture was one of the Earth being the center
of the universe. Later when it was understood that the planets, including the Earth,
circle around the Sun as center, the picture shifted to one in which the Sun was taken
to be the center of the universe. Later, when it was appreciated that the myriad of
stars in the sky were also suns, the idea of our position being one of dominance in the
cosmos was completely overturned. If the ancient astronomers had observed the red-
shifts of the distant galaxies (that they would have taken to be little pin-pricks of light
rather than huge blazing conglomerations of suns) and importantly, that they were
seen as such in all directions from our position as center, and had they understood the
significance of red-shift, likely they would have used this knowledge to fortify their
claim that we are the center of the universe! However from our present vantage point
that our galaxy, the Milky Way, is in no way special, we now make the assumption
that on the average, every observer in the universe should see the same averaged
picture. This is codified into a principle, the “Cosmological Principle”: “At a given
time, the universe has the same averaged appearance for all observers”. Since we see
an essentially isotropic (the same in every direction) picture of the distant galaxies
moving away from us with increasing speed with distance, so too should this be the
case for all observers.
Recall in Chap. 4 how it is the geometry of spacetime that describes gravity in
Einstein’s General Relativity. As well, recall that it is the “metric” that determines
the spacetime geometry. We require the metric for a spacetime that satisfies the
Cosmological Principle, one that accords the same averaged picture of the universe
for all observers at any given time. It can be shown that this is most easily expressed

1 However it has been reported by H. Kragh that K. Wirtz and L. Silberstein had preceded Hubble

by several years in discussing the possibility of a Hubble-like law.

2 Because H is not really a constant, the law is not really a linear one. It is treated as a linear

relationship as an approximation that is reasonably good for not-too-distant galaxies.

7.2 Evolving Views on the Universe 95

Fig. 7.1 This figure depicts the two-dimensional analogue of the spatially-flat k = 0 universe. The
ratio of circumference to diameter of a circle drawn in this space has the value π

in spherical polar coordinates that we used in Chap. 5 in the form3

⎡ ⎤
⎢ du 2 + u 2 (dθ2 + sin2 θdφ2 ) ⎥
ds 2 = dt 2 − R(t)2 ⎣  2 ⎦. (7.2)
ku 2
1+ 4

We need not be concerned about how this was developed nor the details about it.
We have presented this equation to illustrate the power of General Relativity, how
this rather simple metric can describe the geometry in averaged terms of the entire
universe! The symbol k can assume the values 0, 1, or −1. For k = 0, the metric is
nearly of the form of Special Relativity, i.e. no gravity at all, that we saw in Chap. 3.
In fact when the R, which normally varies with time t 4 should be the special case
where there is no change with time, R now being a constant can be folded in with u.
We would then define Ru as the r that we had before and this gives precisely the
metric of Special Relativity. So we see that for this special k = 0 case, were it not
for the time evolution of the universe, with the R varying with t, we would have
a very simple universe with no gravity. It is the time evolution that saves it from
triviality! Many cosmologists believe that this evolving k = 0 case is what describes
our universe. It is infinite in size, it has no boundaries and any snapshot in time gives
the geometric appearance of flatness. Its geometry is the geometry of Euclid that we
studied in grade school. If we were to draw a circle in the space, the ratio of the
circumference to the diameter of the circle would measure π.
To get a better picture of this universe with time dependence, we resort to a two-
dimensional analogue version. Consider a flat sheet of infinite size on which spots
are painted equally spaced everywhere on the surface. We have homogeneity and
isotropy- sameness at all points and in all directions. Imagine this sheet being heated

3 Recall from Chap. 5 that ds is the tiny element of distance in spacetime, dt is the tiny increment
of time as a coordinate, and dθ and dφ are the tiny increments of angle in the system of polar
coordinates. Here du plays the role of the earlier dr .
4 And so was written R(t) to show the dependence.
96 7 The Universe According to Relativity

by the same amount everywhere. As it expands, the spots are seen to increase their
spacing between them. Clearly, any spot that we choose as center for observation
will present the same picture: the spots around it all recede from the center in the
same manner. The homogeneity and the isotropy are maintained.
The k = −1 and k = +1 cases are more interesting. The k = −1 case represents a
universe of what is called negative curvature. The analogous two-dimensional model
is that of the surface of a saddle. If you were to carve out a circle5 at the center of a
saddle and attempt to flatten it onto a plane, you would find that it would not do so
without folding some excess material. Clearly, if you were to measure the ratio of
circumference to diameter of this circle, you would find that it is greater than π, i.e.
the circumference in this case is larger than it was for the corresponding k = 0 case.
It turns out that this universe is also of infinite extent.
The analogue model for the k = +1 case is the two-dimensional surface of a
balloon. Again, if you were to paint spots everywhere equidistant from their neigh-
bour dots, you would see that for any point chosen as center, the same appearance is
presented. The model for the universe here is particularly appealing: the expanding
universe is represented by blowing up the balloon. As you do so, the spots separate
but they do so uniformly. Any chosen center continues to experience the same view
as any other chosen center. Again, the attempt to flatten the material cut out from a
circle shape here will fail. Clearly, the material would have to be slit open to allow
it to flatten onto a plane. There is less circumference for the given diameter of circle
than there was in the k = 0 case. Now the ratio of circumference to diameter is less
than π. We refer to this case as one of positive curvature. Again it is unbounded.
Any journey on the surface never encounters an edge. However, unlike the other two
cases, this universe is of finite size, the size of the surface of the sphere. Journeys in
this universe can go on and on but one retreads the same territories, unlike the other
two cases.
At this point, we ask the natural question: which of these three possibilities is
our universe? As well, while we have the evidence that our universe is in a state
of expansion at present (the observations of red-shift), we could imagine for any of
the three possibilities that the universe could reverse direction at some point or even
perform an unexpected “dance”, reversing its direction followed by resuming its
expansion, or even more complications in motion. The power of General Relativity
is that it provides definite answers, answers which depend upon the averaged density
of the matter in the universe. This is the point where the importance of the value of
the Hubble parameter H enters:

(a) If the present averaged density of the matter in the universe is less than 3H 2 /8π,
the universe would be of the negative spatial curvature k = −1 variety and
according to the Friedmann solution of the famous Einstein equation of General
Relativity, it would expand forever.

5The totality of points equidistant from the chosen center point is the formal definition of a circle.
We can apply this here as we did in the flat space.
7.2 Evolving Views on the Universe 97

C/D>π

Fig. 7.2 In this two-dimensional analogue of the negative-curvature k = −1 universe, that of the
surface of a saddle, the ratio of circumference to diameter of a circle drawn in this space is greater
than π. The saddle-material has to be folded in segments in order to collapse onto a plane

(b) If the present averaged density of the matter in the universe is equal to 3H 2 /8π,
the universe would be of the zero spatial curvature k = 0 variety and according
to the solution of the Einstein equations of General Relativity, it would also
expand forever. However in this case, it would approach a zero expansion rate
as time approaches infinity. In both this case and the former case, the infinite
universe would get ever less dense as time evolves.
(c) If the present averaged density of the matter in the universe is greater than
3H 2 /8π, the universe would be of the positive spatial curvature k = +1 variety
and according to the solution of the Einstein equations of General Relativity,
it would reach a maximum size at which point it would stop expanding and
reverse direction. In this later contracting phase, the light that we would begin to
98 7 The Universe According to Relativity

C/D<π

Fig. 7.3 In this two-dimensional analogue of the positive-curvature universe, that of the surface of
a sphere, the ratio of circumference to diameter of a circle drawn in this space is less than π. For
this model, in order to have the sphere material collapse onto a plane, it is necessary to make cuts
in the material. The result is a sequence of wedges of material interspersed with no material

Fig. 7.4 The Einstein field R

equations determine that
the open (k = −1) and
the critically open (k = 0)
universes expand forever open
whereas the closed (k =
+1) universe will reach a
maximum radius and then
re-contract as shown critical

closed

O
t

observe would be blue-shifted rather than red-shifted as it is now. And of course

the nature of the universe would be so very different from the previous two cases.
The k = +1 universe has a finite amount of matter and a finite size.
It is interesting to ask ourselves: which type of universe seems more “reason-
able”, an infinite universe with an infinite amount of matter or a finite universe
with a finite amount of matter? Probably the majority of people would find it
hard to entertain the idea that one could go on and on and on and find more
and more and more matter, matter without end. They would more likely side
with the more palatable idea that there is only so much matter out there and we
7.2 Evolving Views on the Universe 99

would sample it multiple times in an ever-continuing journey of exploration. At

the present time, the value of the Hubble parameter H would indicate a value
of the density to favour case (a) as described on page 140. However, there is far
more to this fascinating story that we will develop, as we proceed.
Chapter 8
Spacetime Diagrams for General Relativity

In the earlier chapters, we saw that much can be appreciated pictorially by suppressing
two of the three spatial dimensions in order to bring time into the illustration for
Special Relativity. We now do this with the universe itself where General Relativity
comes into play. Gravity is the primary mover in cosmology, and this justifies our
dependence upon General Relativity, our best theory of gravity, to describe the overall
dynamics of the universe. We consider the origin O of our two-dimensional plot as
the “big bang”, the event of the birth of the universe1 and instead of having the space
coordinate as used previously to map out distance along x, one of the triad x, y, z
of mutually perpendicular axes, we will use angles instead. The use of angles for
plotting position is familiar to us: We describe our position on the surface of the earth
using the angles of latitude and longitude. Where we sit, composing this book, our
latitude is near to the 49th parallel, the boundary between the USA and Canada and
our longitude is approximately 123◦ .
For the evolution of the universe plotted in a spacetime diagram, we require one
dimension to indicate time, so we have only one space coordinate at our disposal and
for this, we use the angle coordinate on a planar slice of the universe. Rays fan out
in all directions from the big bang O, from an arbitrarily set line as 0◦ all the way
around to 360◦ where we return to 0◦ . The matter of the universe bursts forth from
the big bang along the rays as indicated in Fig. 8.1:
The twist here is that distance along the rays is measuring time since this is a
spacetime diagram. In Fig. 8.1, we see the particles advance in time as we trace
along the rays outwards, and here, the universe is in a state of expansion forever,
the rays extending radially indefinitely. Cosmologists refer to this model as being
“spatially flat”. It is a universe that is infinite in size.
This spatially flat picture of the universe is one that is in favour with many, if not
most, cosmologists at present. However, as we discussed previously, there are two
other basic types of possible universe pictures that have been considered since the
early dawn of General Relativity and here, we will focus on one of these, simulating

1 We will have more to say about the universe in what follows.

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 101

DOI: 10.1007/978-3-642-30385-2_8, © Springer-Verlag Berlin Heidelberg 2012
102 8 Spacetime Diagrams for General Relativity

Fig. 8.1 The birth and ever-

lasting life of the “spatially
flat” universe expressed as a
spacetime diagram

Fig. 8.2 This figure repre-

sents a spacetime model for a
universe with simulated posi-
tive curvature. It is like a globe
with longitude curves, O at
the South Pole, C at the North
Pole. Arrow heads pointing up
along the curves indicate the
advance in time

the finite “positive curvature” universe. Instead of a sheet, we now draw a spherical
shell and place the big bang event O at the South Pole as in Fig. 8.2.
The curves of latitude that are drawn on globes of the Earth are now drawn with
a different meaning: The matter bursts forth from O at the birth of the universe with
the advancing latitude curve denoting the size of the universe that has been reached
to that point. The universe elements spread out in all directions, advancing along
the time lines. As they proceed in time, they pass the “equator” marker at which
point contraction begins. The evolution continues as we get closer and closer to what
we would ordinarily regard as the North Pole of the global Earth model. Finally,
all of the elements come together at C, the “big crunch”, which is the reverse of
the big bang. Some refer to this as the death of the universe but others see this as a
point at which rebirth occurs; a new big bang followed by cycles of expansion and
contraction, continuing indefinitely. This raises new issues which we will discuss in
the chapters ahead.
Chapter 9
The Motion of the Stars in the Galaxies

9.1 The Dark Matter Paradigm: An Overview

In the 1930s, F. Zwicky discussed the motions of the galaxies that comprised the
elements of the Coma Cluster of galaxies. He noted that their velocities were not
falling off with distance from the center of the cluster as would be expected on the
basis of Newton’s theory of gravity.1 Years later, V. Rubin observed that within an
individual galaxy, the velocities of the stars did not fall-off with distance from the
galactic axis of rotation, again unexpected on the basis of Newtonian gravity. Rather,
the stars essentially maintained the same velocity over the entire range from near the
galaxy core to the visible galaxy edges. It should be noted that Rubin faced great
opposition regarding her claims that the stars far from the galaxy centers were trav-
eling at such high speeds. It was thanks to the emerging radio telescope technology
in the 1970s that it became possible to measure with accuracy the velocities of the
stars in the galaxies. These observations led to the idea that there was a great deal
of matter that was not yet seen, which was providing the required extra gravitational
action to make these motions possible, again according to the demands of Newton.
Thus arose the theory that the universe contained a vast reservoir of “dark (exotic)
matter”, matter that did not have any electromagnetic interaction, unlike the nor-
mal “baryonic” luminous and non-luminous matter that constitutes the substance of
our every-day world. This exotic dark matter could never make itself “visible”, in
the usual sense of the word. At present, there are observational research projects in
progress in an attempt to find this elusive dark matter.2 There are the observers at
CERN in Geneva hoping to find new particles in their impressive ultra-high energy
experiments. There are observations being carried out in very deep mine shafts to
detect “WIMPS”, weakly interacting massive particles, as potential candidates for
dark matter. Thus-far, there has been no evidence for dark matter in these searches.

1 We will return to this issue in the following chapter.

2 In more recent years, the word “exotic” has been dropped in reference to the conjectured mysterious

material and it is generally referred to simply as “dark matter”. At times, we add the word “exotic”
as a reminder.

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 103

DOI: 10.1007/978-3-642-30385-2_9, © Springer-Verlag Berlin Heidelberg 2012
104 9 The Motion of the Stars in the Galaxies

In fact, in the Gran Sasso Xenon100 experiment, a deeply submerged and

very carefully isolated container of xenon has been observed for evidence of
decays resulting from supposed WIMP-xenon nuclei collisions. The dark matter
approach conjectures the existence of vast quantities of dark matter particles through-
out the galaxy and given our motion in the galaxy, collisions with dark matter parti-
cles should be the routine occurrence, should these particles be present. The problem
has always been one of distinguishing the dark matter effects from collisions with
ordinary matter and it is the superb isolation afforded by the Gran Sasso mine that
provides the required isolation. Recently it has been reported that the observations
have reached a very high level of sensitivity yet no evidence for the existence of
WIMPS, the leading dark matter candidate, has emerged.
Even prior to the yet-unsuccessful attempts to detect direct evidence for dark
matter particles, a number of investigators found this dark matter prospect too akin
to the ephemeral “ether” controversy of the previous century, a controversy ending
with the discarding of its previously presumed necessity of existence. In place of the
ether theory, there emerged the theory of Special Relativity and the ether vanished as
easily as it entered the discourse of physics. Beginning with A. Finzi and followed by
D. Milgrom, modifications of Newtonian gravity were proposed which were designed
to explain the large stellar and galaxy velocities without the requirement for any
dark matter. Milgrom’s approach was put into a relativistic context by J. Bekenstein.
J. Moffat and collaborators in turn produced new theory modifications in efforts to
do away with dark matter.
We took an entirely different and, we would argue, the most conservative view of
all, even more conservative than what many would regard today as the conservative
view, namely that our conventional baryonic matter is dwarfed by vast amounts of
exotic dark matter. Our reasoning had, at its base, the fact that General Relativity
is universally considered the premier theory of gravity, and justifiably so. It had
already demonstrated its stunning success in accounting for the subtle discrepancy
in the orbital motion of the planet Mercury. Most importantly, the theory was not
designed with the specific task of doing so; rather the number, the residual precession
of Mercury’s orbit, simply falls out from Einstein’s theory without any “tweaking”.
For us, and probably for the many objective readers, this represents a very powerful
endorsement, a sense that there is almost a kind of mystique attached to Einstein’s
great achievement.
The inevitable question arises: since Einstein’s General Relativity has been
appreciated for so long as our best theory of gravity, why was it not invoked before to
analyze the motions of the stars in the galaxies and the galaxies in their clusters? The
answer stems from the prevailing conviction that whenever the gravitational field is
weak, and when the motions of the bodies involved are not approaching the speed
of light, i.e. not “relativistic”, General Relativity will hardly make a difference, that
Newtonian gravity will provide the correct basic results. Since the motions in ques-
tion are not relativistic and since their gravity is not very strong, the idea of applying
General Relativity to these problems was never entertained previous to our work [23],
at least not in print to our knowledge. In what follows, we will outline how we have
applied General Relativity both to the issue of stellar velocities in galaxies as well as
9.1 The Dark Matter Paradigm: An Overview 105

to galaxy velocities in clusters [24]. We will show that General Relativity presents
many more possibilities than does the very simple Newtonian gravity, that its impor-
tance and richness in the galactic domain has been naïvely overlooked. We will show
that General Relativity can indeed account for what is now known about the high
velocities in the cosmos without vast quantities of exotic dark matter. Interestingly,
it can also describe these motions with much more matter than is seen. This is not
surprising because Einstein’s gravity theory is far richer than that of Newton. We
will develop this picture in what follows. We will devote a considerable amount of
discussion to this subject as it brings together so much of fundamental importance
to physics, astrophysics and cosmology.

9.2 Galactic Dynamics: Newtonian and General

Relativistic Approach

Since the unexpectedly large velocities of stars in the galaxies have been the key
factor in postulating the existence of dark matter, we will deal with this subject in
some detail. Moreover, in keeping with the chosen targeted readership, insofar as
reasonably possible, we will simplify any technicalities.
Recall from Newtonian gravity that the force between bodies of masses m and M
separated by a distance r is
F = Gm M/r 2 . (9.1)

Let m be the mass of a planet, treated as a test mass in the field of the very massive
Sun whose mass is M. Then, invoking Newton’s Second Law of dynamics, F = ma,
in conjunction with (9.1), we have

ma = Gm M/r 2 . (9.2)

For circular motion about the Sun, the acceleration a of the planet is related to the
velocity and the distance as

a = v 2 /r (9.3)

where v is the orbital velocity. Canceling m on both sides of (9.2) and substituting
v 2 /r for a from (9.3), we have

v 2 /r = G M/r 2 (9.4)

which simplifies to 
v= G M/r . (9.5)
106 9 The Motion of the Stars in the Galaxies

So we see that according to Newtonian

√ physics, the velocities of the planets fall off
with distance from the Sun as 1/ r . This would also apply to an extended spherical
Sun with large radius. Over the years, astronomers applied essentially this kind of
reasoning√in the case of a galaxy to extrapolate that the stellar velocities should fall
off as 1/ r as they viewed stars with increasing distance from the galaxy’s axis
of rotation, the axis which cuts symmetrically through the galaxy’s central core.
When Rubin announced her findings that the stellar velocities remained essentially
constant as tracked outwards along the galaxy disk with increasing r , her results were
regarded with √ great skepticism, so convinced was the scientific establishment of the
expected 1/ r velocity fall-off with distance. As the evidence mounted, they could
no longer ignore her work. For many years, her “flat galactic rotation curves”, the
plots of velocity as a function of radius r from the galactic rotation axis, have been
well-accepted and frequently corroborated by quite a number of researchers.
As we noted above, in fact much earlier than the work of Rubin, Zwicky in the
1930s had come to a similar unexpected conclusion regarding the velocities of entire
galaxies taken as individual unit constituents of the Coma Cluster of galaxies. He, as
did Rubin, felt obliged to conjecture the existence of vast quantities of unseen matter
that would act to drive these larger-than-expected velocities, matter now designated
as “dark matter”. The issues of the discovery of the particles that constitute dark
matter and its essential nature have been regarded as issues of paramount importance
in particle physics and astrophysics for quite a number of years and into the present
day. This is understandable as the dark matter is generally believed to constitute
approximately eight times the mass of the known conventional varieties of matter in
the universe.
What surprised us was that researchers had accepted unquestionably the view that
Newtonian gravitation was adequate for the analysis of both the motions of stars in
the galaxies and of galaxies within galaxy clusters. They had been led to this stance
because generally, for both types of phenomena, the motions are non-relativistic
and the gravitational fields are weak. It has long been the accepted mantra that
these conditions suffice for the legitimate application of Newtonian gravity theory.
General Relativity, it was declared, could only present minor corrections. While in
many physical situations, this is indeed the case as it was for the extra perihelion
precession in the case of Mercury’s orbit, there are reasons for the exercise of greater
caution. For example, suppose we were to consider an ordinary rotating rod moving
at non-relativistic speed or a normal linear mass oscillator, two masses affixed to the
ends of a spring. If we were to accept the standard dictum, we would miss a most
profound prediction of General Relativity that is totally absent in Newton’s theory,
the emission of gravitational waves. While the wave field is very weak at nearer
distances in these cases in comparison with the dominant Newtonian component, the
wave field gradient falls as 1/r with distance in comparison with the 1/r 2 Newtonian
component and hence the wave field ultimately dominates in the far-zone.
Another aspect concerns the non-linear structure of the Einstein equations and
its role in sources whose dynamics are driven by gravity itself, rather than the much
stronger forces that are actually ultimately electromagnetic in origin. Many years
ago, Eddington [26] referred to this aspect when he developed the sequence of
9.2 Galactic Dynamics: Newtonian and General Relativistic Approach 107

approximation equations for slowly moving weak-field sources. He demonstrated

the importance of non-linearities when gravity itself is the agency responsible for
the motion of the bodies.
These considerations led us to adopt an open mind when it came to the analysis
of stellar motions in a galaxy. While there are a variety of phenomena at play within
a galaxy, the essential dynamics are driven by gravity itself (what we frequently
refer to as “free-fall”, when gravity as the driver is dominant) and therefore we
were particularly motivated to investigate whether there had been any appropriate
theoretical development within General Relativity to model the dynamics of a system
that could be seen as an idealization of a spiral galaxy. Fortunately, there had been
indeed. Many years ago, van Stockum [27] developed the Einstein equations for a
stationary axially symmetric system of rotating dust and Bonnor [28] had developed
this work further. Their work was in the category of pure inquiry, the quest for probing
General Relativistic solutions with different sources. For us, it was to recognize their
idealized dust solutions as an excellent tool of transition to model the magnificent
spiral galaxies presented to us in nature.
Now it is certainly true that the galaxies are not made of dust but rather of stars,
planets, comets and a great variety of other debris. It is also true that the galaxies are
not the idealized structures of perfect axial symmetry as studied in the van Stockum-
Bonnor works. The galactic structures have a decided lumpiness with highly variable
alterations of mass density, mostly in the form of stars, separated by near vacuum
for the most part and they often are seen to have huge spiral arms, further negating
the assumption of axial symmetry. However, General Relativity, as we have seen,
presents a major mathematical challenge and hence while it would be most wonder-
ful indeed to reach the level of modeling to fully reflect these aspects, we recognize
that simplifications are necessary, at least at this stage in our development of the
theory. The reality, as we understand it, is that the galactic constituents are primarily
driven by gravity and the overall picture is one of a vast, roughly axially symmetric
distribution of matter in steady rotation about the axis of symmetry, like the top that
children pump into rotation and release. The primary gravity driver aspect justifies
the modeling with dust, unlike the material of the child’s top whose metallic ele-
ments are bound together by forces that are ultimately electromagnetic in origin. It is
important to understand that given the nature of our modeling, there is no basis for
expecting detailed concordance with astronomical observations. For example, some
astronomers [29] have criticized our work on the basis that their observed density
profile of matter at our location in the Milky Way is much greater than the density
profile at the same radius in our model. We have offered a number of reasons for
such a difference, quite apart from the essentially approximate nature of our model
in [23]. First, the error bars in the data of [29] are very large and our solar neigh-
bourhood could be a region having density considerably higher than average for our
galactic radial position. Second, we have used a mere ten-parameter fit to model an
average distribution. Third, without a knowledge of the densities at different regions
of the Milky Way at the same r value, one cannot logically arrive at their conclusions.
Fourth, quite apart from these factors, it must be remembered that the typical galaxy,
albeit in basically gravity-driven motion, is a very lumpy distribution of stars and our
108 9 The Motion of the Stars in the Galaxies

model is a smooth dust continuum. In spite of all of these factors, in what follows we
will present strong independent evidence for the essential concordance of our model
with the reality of the galactic dynamics as they are observed in nature.
Astronomers would like to see the mathematical model present a galaxy more
like the pancake-like appearance of the beautiful spiral galaxies that are observed.
While further refinements of our dust model might very well get closer to this shape,
it is quite possible that close matching is intrinsically impossible. The reason for
this, as we suggested above, is that our model consists of dust rather than the lumpy
distribution of reality. Thus, while our model captures the free-fall aspect, it fails to
provide the additional localized enhanced tugs from each individual lump in the form
of a star. Taking this lumpiness into account might very well bring the distribution
into pancake-like distribution. Intuitively, one would imagine that dust would not
tend to form pancake-like distributions but rather would tend to diffuse into more of
an extended shape.

9.3 General Relativity Applied to the Observed Galactic

Velocity Data

We will tread lightly on the mathematics of General Relativity to describe how the
solutions are obtained for the galactic motion problem. First, the spacetime metric for
our very idealized stationary axially-symmetric model is most naturally presented in
the cylindrical polar coordinate system (t, r, φ, z) as shown in Fig. 9.1.
Having a coordinate (r ) to measure distance from the galaxy’s rotation axis and
a coordinate (z) to measure distance above and below the galaxy’s symmetry plane
is very useful. Naturally as with the spherical polar coordinate system, we need
the azimuthal angle (φ) as the third spatial coordinate. Note that there will be no
φ-dependence in any of the metric functions because of the axial symmetry assump-
tion. It is invariably helpful in choosing coordinate systems, as we have done here,
to incorporate those coordinates that match the symmetries of the physical system
being studied (Figs. 9.2, 9.3, 9.4).
It can be shown that the general form for a stationary axially-symmetric metric
can be expressed as3

ds 2 = −eν−w (udz 2 + dr 2 ) − r 2 e−w dφ2 + ew (cdt − N dφ)2 (9.6)

where u, ν, w and N only depend upon r and z. It turns out that for weak fields,
the u function can be chosen to be constant which we take to be 1 for maximum
simplicity. The stationary aspect means no time-dependence and the axial-symmetry
aspect means no φ dependence.
In problems of this form, there is an additional choice to make. We could have
the coordinate system being non-rotating but it might perhaps appear surprising that
it is most useful to have the coordinate system “comoving” with the matter, in this

3 A fuller development of the mathematics that follows can be seen in [1, 23].
9.3 General Relativity Applied to the Observed Galactic Velocity Data 109

dz

dϕ

dr

Fig. 9.1 A cylindrical polar coordinate system with coordinates (r, z, φ, t) is illustrated in the
figure. The small increments of the spatial coordinates are indicated

case co-rotating. This choice, first made by van Stockum, actually simplifies the
mathematics. Comoving coordinates are also used extensively in cosmology.
It is straightforward to show that the approximate angular velocity of the matter
relative to non-rotating observers is simply related to the N function as

Nc
ω= (9.7)
r2
and the tangential velocity V of an element of the galaxy being studied is simply the
angular velocity times the radius r at which the element is located,

V = ωr = N c/r. (9.8)

After much mathematics, the essential Einstein field equations for the metric (9.6)
simplify to
Nr
Nrr + Nzz − =0 (9.9)
r

Nr2 + Nz2 8πGρ

2
= (9.10)
r c2
110 9 The Motion of the Stars in the Galaxies

Milky Way velocity curve-fit

250

200

150
velocity

100

50

0 10 20 30
r

Fig. 9.2 Velocity curve-fit for the Milky Way in units of m/s versus Kpc

Milky Way radial profile density Milky Way vertical profile density

1.5e–20 1.5e–20
density

density

1e–20 1e–20

5e–21 5e–21

0 10 20 30 0 10 20 30
r z

Fig. 9.3 Derived density profiles in units of kg/m3 for the Milky Way at (a) z = 0 and (b)
r = 0.001 Kpc

for the metric component N , and the ν function must satisfy the equations

2r νr = Nz 2 − Nr 2 , r νz = −Nr Nz . (9.11)

In the compact manner that we prefer to write our equations as we have above, you
might be lulled into thinking that these are quite simple. However, a subscript next
to N , for example Nr means, for the calculus-trained reader, a “partial derivative”
9.3 General Relativity Applied to the Observed Galactic Velocity Data 111

NGC5033 Velocity Curve-Fit

250

200
velocity (km/s)

150

100

50

0
10 20 30

r (kpc)

NGC5033 Density at z=0

2e±20
density (kg/m^3)

1.5e±20

1e±20

5e±21

0
10 20 30

r (kpc)

Fig. 9.4 Velocity curve-fit and derived density for NGC 5033

of N with respect to r , i.e. ∂∂rN in the language of calculus and two subscripts mean
two such operations. While this in itself is not generally a problem in “linear” equa-
tions such as (9.9), difficulties mount when such terms are squared as they are in
(9.10) and (9.11). It is this non-linear aspect of the General Relativity field equations
which make this theory so much more mathematically challenging as compared to
112 9 The Motion of the Stars in the Galaxies

the linear Newtonian theory of gravity. Quite apart from the fact that astronomers
have invested so heavily in Newton’s gravity, they are understandably reluctant to
embrace the far more difficult Einsteinian gravity theory. However it is incumbent
upon researchers to adopt the best that science can offer. Insofar as gravity is con-
cerned, the preferred theory is General Relativity and when the preferred theory
renders essentially different results, it must be chosen.
As it turns out, for the galactic problem, we are very fortunate that at least one of the
Einstein equations, namely (9.9), happens to be linear. Linear equations are not only
much simpler to solve than non-linear equations but also, they afford us the luxury of
what is termed “linear superposition” of solutions. This means that if S1 , S2 , S3 . . .
etc. are solutions, then the linear superposition C1 S1 + C2 S2 + C3 S3 + · · · where the
C  s are constants, is also a solution. It is this facility that we exploited in modeling
Einstein solutions that conform to the observed galactic rotation curves.
For each studied galaxy, we chose the superposition of N function solutions of
(9.9) to match the observed rotation curve velocity distribution according to (9.8).
With N found, (9.10) determines the density distribution ρ as a function of r and z,
i.e. throughout the galaxy. (From axial symmetry, it is the same for any azimuth angle
φ value.)
In our first applications, we studied our own galaxy, the Milky Way, and three
external galaxies NGC 3031, NGC 3198 and NGC 7331. These are discussed in
detail in [23]. Because it is so “close to home”, we present here the results from
the Milky Way solution, that of our own galactic residence. More recently [30], we
have analyzed three additional galaxies, NGC 2841, NGC 2903 and NGC 5033.
We include here the table of results for the new set of galaxies as well as the plots
for NGC 5033.
In the table of galactic data, the “critical radius” refers to the distance from the
galactic center at which the last visible light is observed. Data beyond that radius
is referred to as “HI” radiation. This data is in the radio-wave zone of the electro-
magnetic spectrum. The “critical density” refers to the density that is read from the
density plot at the critical radius. The “mass” values are those deduced from our Gen-
eral Relativistic model and “Kent’s value” refers to the Newtonian-gravity-derived
values listed in the Kent paper [31].
The first thing to note in our velocity curve fits is how well they agree with the
observation data points. The next thing to note is the steep fall-off in density as we
examine points further and further from the center, as we would expect for a good
galactic model. It should be noted that these plots were realized with a mere ten
parameters in the expansion of the solution, the C1 , C2 , C3 , . . . , C10 coefficients, as
we discussed above. As well, we note that the masses that we compute are all smaller
than those which have been deduced on the basis of Newtonian gravity. While we
have listed here the table of results for the most recent study, [30], the same situation
held as well for our previous study of galaxies [23].
Of particular interest is this: in our earlier analysis [23], we found that in the
galaxies studied, the critical density was very near to 10−21.75 kg/m3 and we see this
result again for the three new galaxies analyzed (see Table 9.1). This concordance
demonstrates the power of General Relativity to take from the data of observed stellar
9.3 General Relativity Applied to the Observed Galactic Velocity Data 113

Table 9.1 Galactic data

Galaxy Critical Critical density Mass Kent’s
radius (Kpc) (kg/m3 ) (M · 1010 ) value(M ·1010 )
NGC 2841 17.0 10−21.60 51.5 74.6
NGC 2903 12.1 10−21.59 8.9 19.4
NGC 5033 23.2 10−21.83 22.2 37.1

velocity distribution in the galaxies to extract a connection with the essence of the
galactic structure, a power that Newtonian gravity cannot provide. Moreover, it is
to be noted that this agreement occurs for galaxies greatly different in their masses
and at greatly different critical radii. Given the idealization of our model and given
the small number of fitting parameters used, our results are as well-matched to the
actual galaxies studied as one could possibly expect.
One can only marvel at the power of General Relativity to exhibit this concordance.
Moreover it has reached the level of a prediction: present a galaxy with a given
rotation curve and our theoretical structure will predict the approximate radius of
its optical edge. Prediction followed by experimental/observational confirmation is
what we seek in science to build confidence in the correctness of our theoretical
efforts.
Chapter 10
The Motion of Galaxies in Galaxy Clusters

10.1 Preliminary Notes

Zwicky’s observation of the unexpectedly high velocities of the individual galaxies

in the cluster of galaxies known as the Coma Cluster led to the first appearance of
the dark matter hypothesis. Unlike the organized motions in a nearly steady-state
of the individual stars in a single spiral galaxy, the galaxies within a cluster are
generally observed to move chaotically. Since we are not yet in a position to model
such intrinsically time-dependent chaotic systems in General Relativity, we took the
first step in the direction of dealing with time-dependence, again with an idealization.
In what follows, we will describe our idealized model that brings into play intrinsic
time-dependence. As a bonus, we are able to counter some of the criticism that had
been levied against our earlier work.
First, unlike our idealized study of the stationary motion of stars in a single
galaxy,1 we are able to observe General Relativity’s power to deal with systems that
actually evolve in time. Second, unlike the solutions to the approximate Einstein
equations that we employed previously, we are able to work with exact solutions.
The latter is particularly gratifying as there has been a plethora of criticism over
the years against a variety of research studies regarding the use of solutions to the
approximate equations. Working with exact solutions removes all doubt so we were
particularly gratified to be able to make the transition to exactitude.
While we leave the treatment of chaotic systems for the future, for now we study
a highly idealized system of spherical symmetry, a spherically symmetric conglom-
eration of galaxies falling freely with spherical symmetry under gravity, all directed

1 We remind the reader of the technical language that we use in Relativity: “Stationary” in General
Relativity does not necessarily mean “not moving”. It just means that there is no explicit time-
dependence in the description of the motion. For example a disk that is perfectly axially symmetric
and spinning at a constant rate about its symmetry axis, presents a totally time-independent picture
to the observer. The distribution is moving but in terms of its density as a function of position, it
is not varying in time. Such was the nature of our idealized model in the previous chapter. “Static”
is the special case of “stationary” where the meaning is not only time-independent but also “not
moving”.

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 115

DOI: 10.1007/978-3-642-30385-2_10, © Springer-Verlag Berlin Heidelberg 2012
116 10 The Motion of Galaxies in Galaxy Clusters

towards the center-point of the distribution. As in our study of stationary systems,

the system of individual concentrations of galactic matter is idealized as a continuum
of pressure-free matter, i.e. “dust”.

10.2 Perceptions of Velocity: Free-Fall in Vacuum

A key aspect of Relativity centers around the observer’s perceptions of distance and
time. In Special Relativity we focused on how observers in relative motion perceive
length intervals and time intervals differently. In General Relativity, the differences
were manifested as a result of observers being in regions of different gravitational
field intensities, or in the new vocabulary, in different regions of spacetime curvature.
With velocity being the distance covered divided by the time interval for it to be
covered, clearly these differences of perception carry over into velocity as well.
Before we get to the point of examining astronomers’ perceptions of the velocities of
galaxies in our idealized Coma Cluster, we turn to a very familiar subject to people
studying General Relativity, the tracking of a small mass falling radially in vacuum
towards the center of a spherically-symmetric mass. The detailed treatment of this
subject is admirably covered in [3]. Here, we will aim to capture the essence of the
phenomenon with the minimum of mathematics. This will provide us with the natural
lead-in to the galaxy cluster model at hand.
Recall that when we discussed the nature of proper physical measurement of space
and time, the metric tensor came into play. For example, the infinitesimal √ proper
radial distance for the Schwarzschild spacetime with metric (5.4) √ is dr/ 1 − 2m/r
at radial coordinate position r and the proper time interval is dt 1 − 2m/r . As a
result, for the infinitesimal coordinate increments dr, dt, this observer judges radial
velocity vr as the ratio of these quantities, which, after simplification, is

dr
vr = . (10.1)
(1 − 2m/r )dt

However an observer very distant from the mass, views spacetime geometry in
line with the metric (5.2). For him, the increment of radial distance is dr and the time
increment is dt. r is radial distance for him and t is time for him. As a consequence,
the radial velocity vr for him is
vr = dr/dt. (10.2)

When we solve the equations for free motion of a particle released from rest at a very
large r value, the solution is found to be
 
dr 2m 2m
=− 1− . (10.3)
dt r r
10.2 Perceptions of Velocity: Free-Fall in Vacuum 117

In line with (10.2), this is the velocity that the distant observer assigns
 to particles at
any given r value. We see that for r being very large, the factor 2m r in (10.3) makes
the perceived velocity value approach zero as a consistency check. Furthermore, as r
approaches 2m, the first factor in (10.3) makes the perceived velocity approach zero.
Thus from the viewpoint of the distant observer, the falling particle effectively stops
and never actually crosses the “event horizon” at r = 2m.
The picture is very different from the point of view of an observer who is stationed
adjacent to the falling particle. For him, the perceived velocity is given by (10.1) and
when we substitute the requisite dr/dt from (10.3) into (10.1), the result is

vr = − 2m/r . (10.4)

This is the complete opposite of the distant observer’s view: as r approaches 2m,
from (10.4), we see that the adjacent observer views the falling particle’s velocity
approach ever closer to −1. Since we have assigned the value 1 to the speed of light
for this coordinate system, we see that relative to the adjacent observer, the particle
is approaching the speed of light with inward direction of motion (the minus sign).2
The particle does not actually attain the speed of light since, as Synge had shown,
the falling particle’s crossing of the r = 2m surface occurs within the light cone.
All of this is well-known to most relativists. We include it here both for its own
sake as it is likely that the average reader is seeing it explained in this way for the
first time and as guidance for the next step where we apply this formalism to the
system of a collapsing spherical dust cloud.

10.3 Spherical Dust Collapse

Over the years, there has been much interest and a great deal of effort has been
expended in studying the spherically-symmetric collapse of a cloud of dust. There
were good reasons for doing so. First, spherical symmetry greatly simplifies analysis,
particularly in General Relativity where this symmetry guarantees the absence of the
complicating factor of gravity-wave emission. Second, the added factor of choos-
ing dust, the zero-pressure fluid, adds a further layer of simplicity. Third, this model
affords the luxury of yielding exact solutions, an advantage not to be under-estimated.
This study was initiated in the 1930s by J. R. Oppenheimer and his collaborators,
leading to the concept of the black hole. As mentioned earlier, we attacked the prob-
lem from a more realistic point of view, noting that with the densities mounting
precipitously during gravitational collapse, the assumption of the continued mainte-

2 For observers positioned closer and closer to r = 2m, it becomes more and more difficult for them
to remain at rest relative to the central body. They require greater and greater rocket thrust directed
towards the central body to do so. At r = 2m, an infinite amount of thrust would be required,
underlining the impossibility of doing so. This explains why, on the one hand, the local velocity of
the falling particle is seen to be approaching the speed of light as r approaches 2m yet there is no
contradiction with the demand that non-zero rest-mass particles can never attain the speed of light.
118 10 The Motion of Galaxies in Galaxy Clusters

nance of zero pressure is physically unrealistic [10]. Accordingly we considered the

addition of pressure P to the model of gravitational collapse with material equation
of state P = kρ, where ρ is the density and k is a constant. We found the criteria for
the eventual formation of black holes and naked singularities, the latter being, as the
name implies, singularities that are not hidden from view by a horizon.
All of these studies had the goal of dealing with very strong gravitational fields.
Up until the time that we did our more recent work on galactic motions, researchers
would not have felt the motivation for studying the gravitational fields of spherically-
infalling matter during the phase when the gravitational field was weak. To that point,
the common wisdom was that Newtonian gravity is well-suited and surely adequate
for that task. Einstein’s gravity, it was believed, could only add insignificantly minor
corrections. However, having seen with the stationary galaxy model that the com-
mon wisdom is limited in scope, we were motivated to examine the weak-field dust
collapse problem more closely, looking ahead to the aim of modeling an idealized
cluster of galaxies.
The complicated set of exact non-linear differential equations for a collapsing
spherically-symmetric ball of dust as well as the exact solution to these equations is
given in [3] (see also [1]). Given the complexity of the equations, the simplicity of
the solution is quite remarkable, a tribute to the advantage of employing spherical
symmetry. However, the extraction of the velocity dr/dt of the dust as viewed by
distant observers, is a more difficult technical task. We have found this velocity,
whose exact form is [24]

  −1
dr (α + β)(1 − β 2 ) α M  1 ∂ρ
=− + β − (10.5)
dt 8πr 2 ρ2 2M 2(M  )2 4M ∂t

where 
r M 2M
α= , β= . (10.6)
3M r

In this expression, which is far richer than that deduced on the basis of Newtonian
gravity,3 the symbol M  indicates the rate at which the mass up to radius R changes
with radial distance and M  is the rate at which M  changes with radial distance.
The density is ρ. With the new dependency factors now involved, there is scope for
greater possibilities than had existed under the Newtonian gravity assumption.
We were able to show that within this model, the velocities of the galaxies as
observed in the Coma Cluster were consistent with a total mass measure that does
not include the vast extra mass requirement as computed on the basis of Newtonian
gravitational theory. Thus, we see another layer to the power of Einstein’s gravity
to account for higher-than-expected velocities with lower-than-expected amounts of
mass.

3 The Newtonian-based expression depends only upon the value of the amount of mass within the
sphere beneath the observation radius and the observation radius itself.
10.4 Summing Up the Dark Matter Picture 119

10.4 Summing Up the Dark Matter Picture

The idea that the material content of the universe is dominated by a mysterious kind
of matter, now labeled “dark matter”, that shows its presence through gravitational
effects arose from Zwicky’s observations of the galaxy motions within the Coma
Cluster in the 1930s and Rubin’s observations of the stars within galaxies in the
1970s. Dark matter is generally regarded as constituting approximately 95 percent
of the matter in the universe and hence the specific understanding of its nature is justly
regarded as one of the most fundamental issues in present-day physics. In turn the
questioning of its very existence by a small group of researchers, the “questioners”,
including ourselves, ties in to the urgency of seeking out the essential truth of the
matter, if the pun can be excused. While the other questioners have followed non-
conservative tracks in their researches, it must be said that we have been uniquely
conservative in our dynamical analyses, applying the most trustworthy gravitational
theory known, namely Einstein’s General Relativity. In doing so, we should be seen
to be even more conservative than the researchers of the status-quo majority who
profess unwavering faith in the reality of dark matter. After all, the latter base their
findings on Newtonian gravitational theory with its known limitations.
So it all boils down to the issue of limitations. It is repeatedly claimed that for
systems with weak gravity and non-relativistic motions, Newtonian gravity is per-
fectly adequate, that the theory’s limitations are irrelevant in the domains of galaxy
and galaxy-cluster dynamics. Yet in our simple model of a spiral galaxy, a model that
incorporates weak gravity and non-relativistic motion, we have seen that the essen-
tial field equations are non-linear. With these field equations, we have demonstrated
the higher-than-expected velocities can be accounted for with modest amounts of
matter, obviating the assumed need for vast quantities of dark matter to encircle the
galaxies in massive halos. An essential point is this: researchers agree that the math-
ematics of non-linear differential equations is very different from the linear kind and
in all the criticism that we have had directed against our work, there has never been
any suggestion that we have erred in the correctness of our equations. As well, in
our modeling of a cluster of galaxies, there has never been any questioning of our
application of the well-known procedure of velocity deduction by external observers.
Be that as it may, let us further review the current ideas pro and con framing the
dark-matter hypothesis. Up to 95 percent of a galaxy’s mass is conjectured to be in
the form of dark matter by the dark-matter proponents. This measure is based upon
an application of the Newtonian virial theorem to the diffuse interstellar gas at the
galaxy’s edge. However we do not yet have the General Relativity equivalent of the
Newtonian virial theorem and we know that General Relativity can alter the picture
remarkably. Furthermore, according to the dark matter theory of galaxy formation,
there are 10 to 100 times as many small galaxies predicted to exist than are observed.
Gravitational lensing observations are frequently proffered as supporting the dark
matter hypothesis. In this phenomenon, light from distant background objects such
as galaxies is bent by intermediate masses on its way to observers on Earth. The
gravity from the intermediate objects is the agent for the bending and as a result,
120 10 The Motion of Galaxies in Galaxy Clusters

Apparent Galaxy

Galaxy Mass
Earth

Apparent Galaxy

Fig. 10.1 Light from a star in a distant galaxy is viewed multiple times at different positions in
the sky

the same distant galaxy can be viewed multiple times and seeming to be located at
a variety of positions in the sky. In what is termed “strong lensing” stemming from
large intermediary masses, the background galaxies appear to be distorted into arc
shapes by the gravitational lens effect. In “weak lensing”, a large number of very
small distortions due to small intermediary masses, are observed and statistically
analyzed. It has been claimed that General Relativity has been applied for the strong
lensing and the results are consistent with the presence of dark matter. However the
effects of rotation of the intermediary masses has not been taken into account and
we have witnessed the importance of rotation in General Relativistic analyses.
The Bullet Cluster has been invoked as support for the dark-matter hypothesis.
This cluster is regarded as the end-state of a collision between two previously sep-
arated clusters and is claimed to harbour large quantities of dark matter outside the
central visible region. The idea is that while normal electromagnetically interacting
matter was braked in the course of the collision and remained within the bounds of
the collision site, the dark matter, immune from electromagnetic interaction, contin-
ued to follow its initial course and be located beyond the collision site, essentially
oblivious to the effects of the collision. This is often seen as evidence for the exis-
tence of a new kind of matter. However it is amusing to note that subsequently, a new
post-collision source, Abell520 [25] has been observed and analyzed. This source is
seen to exhibit the exact opposite behaviour with a large dark massive region in the
center of the collision region!
Various researchers claim that the strongest evidence in favour of the dark matter
(and dark energy) hypothesis derives from the observed density fluctuations in the
CMB (cosmic microwave background) observations. The claim is that these obser-
vations point to a k = 0 spatially flat (or at least very-nearly flat) universe. Yet
we recall that current observations point to a far-from-adequate amount of normal
baryonic matter in the universe to provide the necessary density for this flatness.
While the estimated numbers vary with advances in research and observations, the
rough values are 73 percent dark energy, 23 percent dark matter and a mere 4 percent
normal matter. Later we will discuss the dark energy hypothesis but here we are
concentrating on dark matter. Researchers are seldom at a loss to conjure hypotheses
and for dark matter, they have conjectured three varieties: “hot dark matter”, “warm
dark matter” and “cold dark matter”.
10.4 Summing Up the Dark Matter Picture 121

Hot dark matter, ultra-relativistic particles, is the least-favoured of the three

hypothesized forms for dark matter because it is inconsistent with the currently
favoured galaxy formation theory after the big bang. For this theory, slower particles
are required to enable a clumping to occur to provide seeds for galaxy formation.
It is ironical in that while the hot dark matter hypothesis is least favoured, such dark
matter is the only form of the three types with known presence in the universe. This
is in the form of neutrinos, the nearly massless highly relativistic leptons that are
produced when neutrons decay into protons and electrons. These particles (as well as
their more massive cousins, the mu- and tau-neutrinos) are not present in sufficient
numbers to fill the large conjectured mass demand for dark matter.
More favoured and supported by a very small cadre of researchers is the warm
dark matter conjecture. These warm dark matter particles are believed to be of very
high velocities but not beyond the capacity to provide the necessary clumping for
galaxy seed requirements. Since there are no such particles in the Standard Model
of particle physics, the warm dark matter particles are conjectured to be found in
the framework of “supersymmetry”, the theory that for every lepton in nature, there
exists a massive supersymmetric baryonic partner and for every baryon in nature,
there exists a massive supersymmetric leptonic partner. This enables nature to recover
its symmetry. In various universities in Europe, the theory of supersymmetry is
considered so natural that it is taught as if it were established reality. In the case of
warm dark matter, the hypothesized particles are the supersymmetric partners of the
photon and the graviton, referred to as the “photino” and the “gravitino” respectively.
The most favoured form of dark matter is “cold dark matter”, CDM. This is
conjectured to consist of massive particles that are non-relativistic at the early stages
of the universe evolution, thus enabling the clumping to form seeds for galactic cores.
The COBE and WMAP satellites have revealed very small deviations from perfect
isotropy in the very early universe, the evidence that is presented for the existence of
such galactic formation seeds. Built into the assumption of the existence of cold dark
matter particles is that prior to the appearance of the small deviations, there were
no small deviations. We have no knowledge of the nature of the period before the
maps revealed by COBE and WMAP so this is a form of compounded speculation.
Particularly troubling for the cold dark matter conjecture is that the required particles,
the so-called WIMPS, or weakly-interacting massive particles, have suffered two
recent setbacks in their quest for revelation. First, to compose the required density
of dark matter, there should be trillions of WIMPS passing through the Earth each
second as it follows its path around the Sun. In ordinary circumstances, the WIMP
collisions with normal matter would go undetected as the very few collision products
would be noticed. However an experiment, the Xenon100 experiment, has been in
operation in a deep mine in Gran Sasso, Italy. A cylinder of xenon has been suspended
in vacuum in the depths of the Gran Sasso mine. Collisions of WIMPS with xenon
nuclei would produce detectable emergent particles. This experiment has recently
reached a very high level of sensitivity and to date no WIMP-xenon collisions have
been detected; this in spite of the vast multitude of collisions that should have occurred
had there been WIMPS present in the first place.
122 10 The Motion of Galaxies in Galaxy Clusters

Even more recently, a second experimental result has been announced at CERN.
This involves B-mesons in the LHC, the large-hadron collider. The result of this
experiment has pointed to a failure of the theory of supersymmetry, a major blow to
dark matter theorists. In our view, it is now time for researchers to turn back to well-
established physics, in particular to Einstein’s General Relativity and to probe further
into the seemingly mystifying phenomena that might be explained more naturally
using the favoured theory of gravity.
Chapter 11
Dark Energy

In very recent times, a new major issue has arisen: Is the true story of the very
distant galactic motions, that is in terms of the universe as a whole, the one that had
been assumed? Is it simply a matter of gravity slowing down the expansion of the
universe? It certainly seems reasonable. However in more recent years, data has been
presented to suggest (and most recently claimed to affirm) that rather than slowing
down, the expansion of the universe is actually speeding up! This data comes from
the observations of certain supernovae, massive exploding stars in distant galaxies.
The observations of these distant supernovae are dimmer than they would have to be
on the basis of a universe that is decelerating in the course of its expansion. Instead,
the currently favoured view is that the universe is now undergoing an expansion with
acceleration. This runs counter to our fundamental experience with gravitation, that
gravity attracts rather than repels, the latter going hand-in-hand with acceleration
accompanying the expansion.
How can physics account for this? Not easily, to be sure, and there remains con-
siderable skepticism regarding the claim of accelerated expansion, this in spite of
the recent awarding of Nobel Prizes for the observational evidence leading to the
accelerated expansion conclusion. To introduce the “cure”, let us return to Einstein’s
early view of the universe, the static universe. We know that gravity is attractive and
free bodies do not want to stay put relative to each other. To remain at rest, they
require outside assistance. Einstein provided this with his so-called “cosmological
constant” , a term that he added to his field equations of General Relativity, in the
form
1 8πG
Rik − gik R + gik = 4 Tik . (11.1)
2 c
This  acts as a repulsion between bodies to keep them at rest relative to each other.
Einstein saw this  term as part of the spacetime geometry, belonging on the left
geometry side of his famous equations. However, as we know from our algebra, a
term on one side of the equation can be readily moved to the other side as long as
one introduces a minus sign. It remains the same equation in essence but here the
interpretation is important. In our earlier work (see for example [1]), we argued that

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 123

DOI: 10.1007/978-3-642-30385-2_11, © Springer-Verlag Berlin Heidelberg 2012
124 11 Dark Energy

Einstein’s  term really belongs on the right-hand “matter” side of his equations,
and then the  term is most naturally interpreted as a kind of very weird matter.
It is matter that provides tension, and for Einstein, it had provided the right amount
of tension to keep his universe static, opposing the gravitational attraction. After
learning about the galactic red-shifts, Einstein abandoned his . It had been widely
reported that he regarded the introduction of  as the worst mistake of his life.
The great physicist, R. P. Feynman once remarked to us that physicists are not
sufficiently imaginative. (Actually he used a more colourful expression but that was
the idea.) Once the  genie was out of the bottle, it could not be put back in as
Einstein had wished, at least not for long. Seeing the assumed need for accelerated
expansion, why not bring back our old friend ? After all, it did the trick for Einstein
when he needed it. Thus it is now widely believed that the universe is pervaded by a
tenuous material called “dark energy” that provides accelerating thrust to the matter
of the universe as required by the demands of the current data from supernovae. And
 is the prime candidate to do the job of dark energy.
But the situation is more complicated than this. The current upper-limit to the
magnitude of  is a very tiny number. After all, we recall how well Einstein’s theory
without any  at all had worked to give us the required 43 s of arc extra precession for
the orbital motion of the planet Mercury. Too large a cosmological constant would
have destroyed this wonderfully reliable result. But quantum field theory, that very
successful theory of the micro-world, demands that the vacuum is not really without
energy, but rather is imbued with an incredibly enormous amount of energy. This
level of energy translates into a value that is actually 10120 times as large as the upper
limit demanded by the observed value! This is not a misprint. It is not meant to be a
hundred times or a thousand times or even a million times larger. It is worse- it is 1
followed by 120 zeros times larger! It makes a googol look like a very tiny number.
Some have rationalized this by proposing that Einstein’s theory provides a “bare”
λ (note uncapitalized) which itself is enormous, almost exactly the same magnitude
as that demanded by quantum field theory and that these two enormous numbers
cancel each other so incredibly accurately as to leave behind the mere wisp of the
observed net  of today! Most researchers, ourselves included, regard this idea as
rather unbelievable.
Over the years, a number of alternative ideas have arisen. A variety of researchers
have proposed that the cosmological term is not constant at all, but rather it evolves
in tune with the evolution of the universe in such a manner as to be large when needed
to be large and to be consistent with the present value today. (See [32] for a review
of these many forms as well as some new forms). The problem with this approach is
one of rationalizing the form of this variation on a fundamental level.
An intriguing alternative approach was provided in [33]. These authors argued
that metallic vapours are ejected in the course of the kind of supernovae explosions
that have indicated accelerated expansion and that these vapours are pushed out of the
galaxies in which the supernovae reside, by shock waves. Such debris would have the
tendency to diminish the intensity of the radiation that would ordinarily come directly
to us, thus explaining the dimmer supernovae without any need for an accelerated
expansion explanation. However others would argue that the CMB observations
11 Dark Energy 125

provide another avenue of support for the necessity of a cosmological term. Still
others would argue that making deductions on the basis of these very-early-universe
observations is akin to the reading of tea leaves. As you can now appreciate, research
in cosmology is unlike the precision research related to laboratory experimental
verification that one generally ascribes to scientific research.
So we see how various researchers have dealt with the observations. One, E. W.
(“Rocky”) Kolb has colourfully described dark energy as the ether of the 21st century.
How this will all play out in time is of course a great unknown. But it is the unknown
that keeps us engaged in the fascinating pursuit of understanding our miraculous
universe.
Chapter 12
Time Machines, the Multiverse
and Other Fantasies

12.1 Closed Time-Like Curves

It is our everyday life experience that we re-visit previous places. We take it for
granted that we always do so at later times. Even with the wonders of Special Rela-
tivity, when we considered the experiences of Alicia and Beatrice in Chap. 3, we saw
that Beatrice re-visited her old home at a later time for her and the re-visit was at a
much later time for Alicia, so much later in fact, that she was no longer a member of the
living. We can accept such phenomena, however incredible they may seem, because
they do not violate the demands of causality, that if A causes B, A will always be
seen to precede B in the ordering of events in spacetime. While the perceived time
differences for Alicia and Beatrice are very different as a consequence of Special
Relativity being in accord with the workings of nature, Special Relativity maintains
the necessary requirement of causality preservation. However, human imagination
does not always display such constraints.
In works of science-fiction and with increasing frequency in the movies, we wit-
ness somewhat outrageous plots in which characters routinely take trips into their
past and back to the future again, with often predictable consequences. Some plots
skirt around the causality issue by having the time-traveler merely observe certain
goings-on without being allowed to interfere with them. Significantly, even some
prominent physicists have been swept along in the excitement of the idea. Some
writers realize that their targeted audience might ask some embarrassing questions.
For example, if the time-traveler, in the course of a stint during which he is re-visiting
his past, were able to kill his grandmother in a fit of rage when she was just a very
young child, then how could the time-traveler have ever been born to be able to be
there to make the voyage in the first place? Good question.
The renowned mathematician, Gödel, had taken an interest in General Relativity
and found a solution of the Einstein field equations of General Relativity for a rotat-
ing universe [34]. This solution, he excitedly proclaimed, exhibited what are called
“closed time-like curves”. Real observers follow time-like curves in spacetime, those
physically permitted curves that lie within the observer’s future light-cone, and if

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 127

DOI: 10.1007/978-3-642-30385-2_12, © Springer-Verlag Berlin Heidelberg 2012
128 12 Time Machines, the Multiverse and Other Fantasies

these curves are closed as a result of the light-cone tipping over sufficiently, it means
that the observer, when his spacetime path touches the closure point, is re-visiting
his past. Thus, Gödel reasoned, a person in this universe could be a time-traveler.
Gödel was so excited about this result that he gave a seminar at the renowned
Institute for Advanced Studies in Princeton which was attended by such luminaries
as Einstein, Oppenheimer and Chandrasekhar [35]. We can only wonder whether
Gödel was seriously challenged as to the physical significance of his work, but at
least Chandrasekhar took it sufficiently seriously to write a paper that made note of the
fact that Gödel’s closed time-like curves were not geodesics [36]. Now we recall that
only freely gravitating particles follow geodesics, so Gödel’s time-travelers would
have required some assistance in the form of rocket propulsion or the like to get back
into their past. While this might add some extra degree of complication, nevertheless
the basic assertion that this time-travel experience could be an element of reality was
maintained as a possibility to be reckoned with. Gödel’s work did not garner much
attention for many years until the appearance of a highly technical book that featured
this work [37] (Fig. 12.1).
More recently, we set ourselves the task of considering both Gödel’s solution and
the entire issue of closed time-like curves. We came to a very simple conclusion:
The realization of a closed time-like curve is a matter of a mathematical choice of
the identification of spacetime points and has nothing to do with physics. Let us
construct a very simple example. Suppose a person is walking around a big circle.
At the start of the person’s journey, a point marked x on the circle, it is 1 o’clock.
One quarter of the way around the circle, his clock reads 1:15, half way around it is
1:30, three-quarters of the way around it is 1:45. When he finally reaches the starting
point x, it is 2 o’clock. Now suppose we were to identify the start spacetime point
(x, 1 o’clock) with the finish point (x, 2 o’clock), in other words to say that 2 o’clock
equals 1 o’clock. We could then proclaim that the traveler has gone back in time! He
has not only revisited his previous spatial position, a routine of our experience, but
he has also revisited the same time at which he had been at the position earlier. Now
you might object to such a cavalier statement of identification but there is nothing to
prevent one in terms of mathematics from making such an identification.
Now when the traveler went from the 0◦ position in the circle at 1 o’clock to 90◦
at 1:15 to 180◦ at 1:30 to 270◦ at 1:45 to end up at the starting spatial point x, we
have no problem in identifying the final angular value of 360◦ with the start value 0◦ .
This is our experience in nature, the revisiting of spatial points. And it is precisely
this aspect that enters into the Gödel example.
It arises in the following manner: in Gödel’s solution there is an angular coordinate
φ which has the standard characteristic of identification, wherein the advance in
angular position from 0 to 360◦ brings one back to 0, the original angular position.
This does not raise any concern for us in those regions where φ is a space-like
coordinate. However, in the Gödel solution there is a region where φ becomes a
time-like coordinate. That is where the problem lies. Gödel, and researchers since
then, have continued to regard φ in this time-like region in the same manner as they
had when φ was space-like and have continued to regard 0 and 2π, i.e. 360◦ , as
relating to the same spacetime point. In doing so, lo and behold, a closed time-like
12.1 Closed Time-Like Curves 129

Fig. 12.1 A person walks

around a circle carrying a
clock. He begins at 1 o’clock at
the top of the figure and com-
pletes one cycle at 2 o’clock.
By identifying 2 o’clock with
1 o’clock, a closed time-like
curve is realized

curve is born! It is important to emphasize that this is a choice that has been made.
It is not a physical necessity.
Einstein’s field equations lead to certain solutions. Hand in hand with a given
solution, we assign a meaning to the coordinates of that solution, for example that
r runs from 0 to ∞, that φ runs from 0 to 2π with 0 and 2π being identified, and
z and t run from −∞ to +∞. This tells us that the coordinates have the character
of cylindrical polar coordinates. The process in its entirety, the given solution to
the field equations plus the characterization of the coordinates, gives us a particular
spacetime. If, as part of the specification, we continue in this same vein even when
φ becomes a time-like coordinate, we have Gödel’s spacetime. However, when we
choose instead not to identify the points in φ when φ becomes time-like, we have a
different, and we would argue a more reasonably physical spacetime. It is as simple as
that. Surprisingly, much has been made of the supposed inclusion of closed time-like
curves in General Relativity and it is likely that this has served as a spur towards the
infiltration of the bizarre time-machine plots in more recent fiction offerings. These
aspects and various examples are developed in some detail in [38].

12.2 Topological Twists

As in various areas of human endeavours, physics has had its share of fads. In recent
years, topological structures have come to the fore as supposed elements of physical
reality. A favourite of this genre is the wormhole. The wormhole, as the figure shows,
presents a means of getting from A to B by the alternate path of entering the wormhole
and emerging from the other end. Those with a science-fiction bent have seized upon
the wormhole as a means of drastically cutting the time required for space-travel
between distant galaxies: find the wormhole in the spacetime that offers a greatly
shortened travel-time requirement and space-travel becomes a feasible option for
human beings.
130 12 Time Machines, the Multiverse and Other Fantasies

Fig. 12.2 The normal path

from A to B is along the straight A
portion from A, then around
the flap at C, continuing to B.
However, the shorter path C
from A to B is through the
wormhole

To get a rough picture as to how this works, consider a normal path in spacetime
between points A and B. This is the course that follows the undistorted path from A
around the flap at C and back to B. However, should A and B be connectable through
a wormhole as in Fig. 12.2, one could connect A and B via the shortened path through
the wormhole.
It becomes even more engrossing when the set-up is envisaged wherein the journey
through the wormhole leads to travel to the observer’s past, a newly-minted time-
machine. Those of us who prefer to have science clearly separated from science-
fiction are left perplexed as we witness actual scientists debating whether or not
nature will intercede to pinch off the travel of any mischievous wormhole-sniffer
who tries to enter the alluring wormhole time-machine.
We are left wondering: where do these scientists glean even the slightest hint that
such topological structures are actually a part of physical spacetime? Without such,
we would suggest that there is no motivation to reach in such directions. It is natural
to ask for an explanation as to why a considerable body of discourse in science has
verged towards science-fiction. Perhaps the most cogent explanation derives from
the increasing difficulty to achieve experimental results. The experiment has always
been the ultimate arbiter of truth. As theory demands greater and greater energies for
experimental confirmation and as these energy scales become increasingly difficult
(and expensive) to achieve, the disconnect between experiment and theory grows.
In our view, it has grown to the point of irresponsible speculative pronouncements
often motivated more by human than scientific needs.

12.3 The Multiverse

The picture of the universe that we have developed to this point precedes the radical
theorizing developments that have accrued over the past two decades. The earlier
picture was based upon our most trustworthy theory of gravity: General Relativity,
its homogeneous isotropic solutions incorporating isotropic expansion, in line with
12.3 The Multiverse 131

our observations of the distant galaxies. As well, the solutions provide that any
observer would witness this same isotropic expansion, a reasonable assumption.
After all, otherwise one would ask: “Why should we be so lucky as to see an isotropic
expansion?” We are just ordinary observers circling an ordinary star in an ordinary
galaxy populated by billions of similar galaxies. While the solutions provide for
the possibility of three kinds of basic form, an observable criterion for the right
selection exists with the determination of average matter density. This is science at
its best, elegant substantiated theory with exact mathematical solutions leading to
observable verification within our reach. Unfortunately, the more recent cosmological
developments have not followed the scientific path, as we will discuss below.
The first sign of trouble is in vocabulary. The word “universe” refers to the totality
of all that exists. Proponents of the new cosmologies hypothesize that just as there
are many galaxies, there are many universes, collectively labeled the “multiverse”.
However, since the universe is already the totality of all that exists, introducing the
new word “multiverse” is most charitably seen as an act of injecting redundancy. The
proponents of the multiverse idea might object in that their definition of our universe
is all that is within our cosmic visual horizon, the furthest distance to which we could
possibly probe, based on the distance that light could reach us from the time of the
big bang. Their other universes comprising the multiverse are those which are never
within our possibility of observation. However, let us use their new vocabulary and
briefly describe a few of the various options being proposed.
What one might regard as the conservative among the new cosmologists accept the
multiverse with many, if not infinitely many universes much like our own, obeying
the same laws of physics that we observe to hold sway. On the other hand, one
might view as the more venturesome cosmologists, those who propose the possible
existence of many or an infinite number of other universes obeying different physical
laws and even different numbers of dimensions, even many with copies of ourselves
(the “parallel universes” idea). However, since the external universes of both the
conservatives and the more adventurous are all forever beyond our reach, both groups
of cosmologists have a key element in common: Neither one is engaged in the activity
that we would normally call “science” but rather in the sphere of science-fiction. This
is because their theorizings are a priori not verifiable.
Some have been very impressed by the values of the fundamental constants of
nature vis-a-vis their unique suitability for the existence of life. They argue that the
existence of this special set points to the viability of a multiverse with individual
copies having a wide spectrum of values for the fundamental constants. With a wide
choice available, to their way of thinking, the existence of one with the crucial values
becomes plausible. Implicit in this line of reasoning is that the multiverse repeats
structures with physical laws of our own, differing only in the numerical values of
their physical constants. But how are we to know? A means of testability is called for.
Some cosmologists have even skirted with the idea that they could bring their
speculative theorizing closer to real science by seeking evidence in the cosmic back-
ground radiation for a prior collision of an alien universe with our own. Seek-
ing evidence is welcome news indeed. However, with the essential uncertainty
132 12 Time Machines, the Multiverse and Other Fantasies

(or perhaps “unknowability”!) of what preceded the big bang, one could hardly
rule with any degree of confidence as to what were the sources of any anomalies
found in the background radiation. Could they derive from our universe or from an
alien universe? Who’s to say?
Chapter 13
Concluding Commentary

Very recently, a new observation has been announced: A burst of neutrinos has been
generated at the CERN facility in Geneva and directed through a slice of the Earth’s
crust and towards the Gran Sasso mine in Italy. Remarkably, the experimenters have
measured the speed of the burst as travelling slightly faster than the speed c of
light in vacuum. Assuming this measurement was accurate, and that it harboured no
experimental flaws,1 this is indeed a very stunning result for different reasons.
First, the entire structure of Relativity rests upon c as the universal speed limit in
nature and that only zero-rest-mass particles such as photons can travel at speed c,
and this, only in vacuum. Moreover the theoretical structure of Lorentz-invariance is
built into the entire theory and the theory works brilliantly. Second, in recent times,
it has come to be accepted by particle physicists that neutrinos, even the mundane
variety of electron-neutrinos, actually do have some mass, albeit very little and the
muon-neutrinos and tau-neutrinos have even more so. We recall that according to
Special Relativity, it would require an infinite amount of energy to make a non-zero-
mass body reach the speed of light c, let alone exceed it.
What are we to make of this truly remarkable claim from Europe? To begin, as the
claimers have said responsibly, attempts should be made to replicate their claim at
different sites in different experimental configurations. Let us assume it is replicated.
What does it say to us? In our view, it points to a general problem that has festered
for a very long time, that of our limited understanding of the relationship between
the micro-world and the macro-world, between the world of quantum mechanics,
quantum field theory and macroscopic Relativity. With electromagnetism, at least
we have the well-defined and well-established classical theory of Maxwell. We deal
with this in our daily lives in so many ways. As well, in those areas where the
classical theory is left wanting, the quantum theory of electromagnetism comes to
the rescue. This theory offers extremely precise verification even though it leaves
the issue of vacuum energy unresolved and some irritating infinities carefully tucked

1 The latest news suggests that a loose cable was responsible for displaying a false reading of the
speed.

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 133

DOI: 10.1007/978-3-642-30385-2_13, © Springer-Verlag Berlin Heidelberg 2012
134 13 Concluding Commentary

away. The picture of the quanta of the electromagnetic field, the photons, fit well
into our appreciation of the nature of electromagnetism at the quantum level. The
photoelectric effect, the spectrum for black-body radiation and the Lamb-shift of the
spectral lines in hydrogen underline this excellent fit.
By contrast, we do not have such a familiarity with neutrinos. They were brought
into the physical fold as the required quantized particles to rescue the conservation
of energy-momentum and angular momentum in phenomena such as the decay of a
neutron. At first they were seen to be the massless spin-1/2 analogues of the spin-
1 photons but more recently, were said to have mass. The theory evolved to the
form wherein the neutrinos of the different kinds, the e-, the μ- and the τ -neutrinos
transform into each other under the process referred to as “neutrino oscillations”,
in the course of their propagation. Now we have the latest claim of their speed being
greater than c. All of these results have come in rapid succession by the standards
of fundamental advances in physics. On top of all of these results, we have to keep
in mind that unlike photons with their classical electromagnetic wave connection,
we have no classical neutrino phenomena to back up our confidence in their innate
character.
While all of the foregoing is very interesting and the recent observations poten-
tially of epic importance, it is time to step back and take a breath before we jump
to conclusions. Our approach from the outset has been the conservative one, in line
with Occam and his razor. We see the enormous challenge inherent in the incorpora-
tion of dark energy in line with the demands of the vacuum in quantum field theory.
As well, we ask whether more mundane explanations for the apparent acceleration
of the expansion of the universe have been adequately pursued. And of particular
interest for us, we reflect upon the strong resistance mounted against our displays
that General Relativity can account for the motions of the galaxies, both individually
and in ensembles, with little or no dark matter. We marvel at how General Relativity
almost magically predicts the optical edge of galaxies simply from the knowledge of
the galactic rotation curves. This is an achievement unattainable within Newtonian
gravity theory. And this achievement underlines the necessity of applying General
Relativity to galactic dynamics.
History has shown us that the evolution of sober scientific thought often proceeds
at a snail-pace while facile “fixes” are often embraced rapidly with insufficient critical
analysis. Eventually the truth emerges.
Appendix A:
Proving That the Spacetime Interval
is an Invariant

Our proof is along the lines of that in [3] but it is somewhat simpler. Noting that
when ds ¼ 0; ds ¼ 0 and that these elements are infinitesimals of the same (first)
order, they must be proportional. On what can the proportionality factor depend?
It cannot depend on where one is located in space because no point in space is
privileged over any other point. Space is assumed to be homogeneous. It cannot
depend upon the time because no point in time is privileged. Now we turn to the
only element that relates Alicia to Beatrice; it is their relative velocity v. Now it
cannot depend upon the direction of their relative velocity because no direction is
favoured over any other direction; the spacetime is isotropic. All that is left is the
magnitude of their relative velocity, modðvÞ. So we write
ds ¼ KðmodðvÞÞ ds ðA:1Þ

where K is the proportionality function. Suppose we were to consider the

relationship from Beatrice’s viewpoint. For her, Alicia has velocity v relative to
Beatrice so the connection between the intervals is
ds ¼ KðmodðvÞÞ ds: ðA:2Þ
But modðvÞ ¼ modðvÞ so the K functions in (A.1) and (A.2) have the very
same value. Thus a substitution of ds from (A.2) into (A.1) gives
ds ¼ KðmodðvÞÞ ds ¼ KðmodðvÞÞKðmodðvÞÞ ds ¼ K 2 ds ðA:3Þ

and therefore K 2 ¼ 1. When we square both sides of (A.1) and substitute K 2 ¼ 1,

we have
ds2 ¼ ds 2 : ðA:4Þ
Thus ds ¼ ðþ=Þds and we choose the plus to preserve the direction of the
flow of time. From the equality of the infinitesimals, it follows that the finite
intervals are equal as well, s ¼ s .
The important result that emerges is that the spacetime interval between any
two given events is an invariant.

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 135

DOI: 10.1007/978-3-642-30385-2, Ó Springer-Verlag Berlin Heidelberg 2012
Appendix B:
Deriving the Einstein Field Equations

In Newtonian gravitational theory, there is one simple equation for the

gravitational field that connects the gravitational potential / to the distribution
of matter density q. This equation is of the type that gives solutions for density
changes that imprint upon the entire universe instantaneously. This is inconsistent
with the basic premise of Special Relativity, so if Special Relativity is correct,
Newtonian gravity must be replaced by a relativistic theory of gravity.
For those readers who are interested in developing a detailed understanding of
the mathematics underpinning General Relativity, they would profit from reading
the next level of treatment in [1] with extra tensor calculus detail in [7]. They
could then follow up with more sophisticated treatments in the classic book of
Landau and Lifshitz [3] as well as the various other more advanced books on
General Relativity. In the present book, we are aiming at a very basic level of
development with just enough mathematics to go beyond those books which rely
solely upon descriptions in words.
Without getting into more advanced mathematical detail, we present a narrative
that leads naturally to Einstein’s General Relativity. To do so, we first consider the
very basic constructs of energy and momentum. What makes them basic is that
they have the valuable property of being conserved when systems undergo
changes. Much of physics through the centuries has been driven by the new
constructs that had been invoked to maintain this primary attribute. For example,
in electromagnetism, we express the conservation of the important physical
attribute that we label ‘‘charge’’ in the form
oq
r:J þ ¼0 ðB:1Þ
ot
where q here is the charge density, the amount of charge per unit volume and J is
the charge-current density vector, qv with v being the velocity of the charge at the
point under investigation.
To see why (B.1) expresses the conservation of charge, we require the
fundamental theorem of Gauss that the integral of the divergence of a vector over a

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 137

DOI: 10.1007/978-3-642-30385-2, Ó Springer-Verlag Berlin Heidelberg 2012
138 Appendix B

given volume is equal to the flux of that vector over the surface that bounds the
given volume:
Z Z
ðr:JÞ dV ¼ J:ndS ðB:2Þ

where n is a vector of unit length perpendicular to the bounding surface and

pointing outward at every point from that surface. When we integrate (B.1) over a
given volume and apply (B.2), we have
Z Z
d
qdV ¼  J:ndS: ðB:3Þ
dt
This equation states that the time-rate of change of the total charge within the
volume V is given by the negative of the rate at which charge flows out through the
bounding surface S. In other words, what is lost in the volume is necessarily
compensated by what flows out of its bounding surface. Thus we see that (B.1) is
an expression of conservation locally.
In Relativity, the natural arena for investigation is spacetime rather than space,
and we generalize the three-dimensional current vector J ¼ ðqvx ; qvy ; qvz Þ to a
new four-dimensional current vector J i ; J i ¼ ðq; qvx ; qvy ; qvz Þ, i.e. we add a fourth
‘‘time-component’’ q in going from the three-vector J to the four-vector J i .
In doing so, we are able to express the conservation Eq. (B.1) as
oJ i
¼ 0: ðB:4Þ
oxi
This is a very useful and important result: The vanishing of the divergence of a
four-vector (or tensor) expresses conservation.
In a similar manner, energy and momentum conservation in Special Relativity
is expressed by the vanishing of the divergence of the ‘‘energy-momentum
tensor’’ T ik :
o ik
T ¼ 0: ðB:5Þ
oxi
Here, instead of dealing with a vector, we have to use a tensor (of second rank-i.e.
two indices i and k instead of the single index that represents a vector.) Actually a
vector is also a tensor; it is a tensor of first rank. While the vector J i has four
components (i taking on the values 0; 1; 2; 3), a second rank tensor has 16
components in general but because this tensor is symmetric, i.e. T 01 ¼ T 10 ;
T 12 ¼ T 21 :::, etc., only 10 components of this energy-momentum tensor are
distinct in general.
Now let us consider working in Special Relativity using non-Cartesian
coordinates, for example polar coordinates, or even transforming to a reference
frame that is under acceleration. It is frequently stated that the latter case is no
longer one of Special Relativity, that one is dealing with General Relativity
whenever acceleration enters the picture. This is quite untrue. One can apply
Appendix B 139

Special Relativity in an accelerated reference frame. In fact Bondi [2] and others
have done so. General Relativity enters the discussion when real gravity is present,
namely spacetime curvature generated by matter and/or fields, not the pseudo-
gravity of accelerated reference frames.
So let us suppose that we are working within the domain of Special Relativity
but not in an inertial Cartesian system of coordinates. The formalism of tensor
calculus tells us how the usual partial derivative of a vector or tensor transforms
(see, e.g. [1, 7]). The usual partial derivative of a vector J i for example, changes to
what is called the ‘‘covariant derivative’’, denoted by a semi-colon. Specifically,
oJ i
J;ki ¼ k þ Cilk J l ðB:6Þ
ox
where
1
Cilk ¼ gim ðgml;k þ gmk;l  glk;m Þ ðB:7Þ
2
and we have used commas to indicate the usual partial derivatives. Thus we see
that the metric tensor comes into play when we deviate from simple Cartesian
coordinates.
We now have the means of expressing the conservation of charge and the
conservation of energy-momentum in terms of arbitrary coordinate systems in
Special Relativity. Specifically, charge conservation changes from (B.4) to
J;ii ¼ 0 ðB:8Þ

and the expression for energy-momentum conservation changes from (B.5) to

T;kik ¼ 0: ðB:9Þ

Now locally, an accelerated reference frame is like a gravitational field

(Equivalence Principle) so we are led to the expression for energy-momentum
conservation in General Relativity: simply (B.9). There is really no other
possibility!
We are almost there. Recall that the mass density q is the source of gravity in
Newtonian physics and most often, when we wish to go to the limit of weak
gravity and non-relativistic velocities, we should retrieve Newtonian gravity from
the relativistic theory of gravity that we seek. Interestingly, the T 00 component of
the energy-momentum tensor T ik is q in the reference frame in which the matter is
at rest so we seek a relativistic gravity field equation of the form

Gik ¼ constant:T ik : ðB:10Þ

Thanks to our earlier discussion regarding the conservation laws, we know what to
look for in this so-far undetermined tensor Gik . We know that since T ik satisfies
(B.9), so too must the right hand side of (B.10), i.e.

Gik
;k ¼ 0: ðB:11Þ
140 Appendix B

As well, Gik must be constructed from gik and no higher than its second derivatives
so that in the appropriate limit, the left hand side of (B.10) reduces to the
‘‘Laplacian’’ of the Newtonian potential /, i.e. r2 / where r2 ¼ o2 =ox2 þ
o2 =oy2 þ o2 =oz2 when we use Cartesian coordinates.
Einstein worked diligently in search of the tensor Gik and in the early stages of
his search, he believed that the correct form of Gik is the Ricci tensor Rik . This
tensor is a very complicated and lengthy combination of Christoffel symbols and
their derivatives:

Rik ¼ Cnik;n  Clil;k þ Clik Cnln  Cnil Clkn : ðB:12Þ

Unfortunately, this tensor does not have a vanishing covariant divergence as it

must to be satisfactory. Eventually Einstein discovered that the required tensor is
1
Gik ¼ Rik  gik Rjj ; ðB:13Þ
2
now appropriately named the ‘‘Einstein tensor’’. Thus we have developed the
Einstein field equations of General Relativity,
1 8pG
Rik  gik Rjj ¼ 4 T ik ðB:14Þ
2 c
where the constant 8pG
c4 has been chosen in order to reduce to the Newtonian field
equation for gravity

r2 / ¼ 4pGq ðB:15Þ
in the limit when appropriate. Rather than one simple field Eq. (B.15), we have ten
very complicated non-linear partial differential equations. There are in general, ten
distinct equations, counting the permutations of i; k ¼ 0; 1; 2; 3 and taking into
account the symmetry, T 10 ¼ T 01 ; T 12 ¼ T 21 , etc. Given the complexity, while the
task of finding valuable solutions might appear hopeless, with sufficient symmetry,
physicists have found some very valuable solutions. These are discussed in the
text. More solutions can be found in the advanced texts.
Regarding motion in General Relativity, the Equivalence Principle again serves
as a valuable guide. Recall from Newtonian physics, that when there are no forces
or when the forces on a body balance out, the body moves with zero acceleration.
This holds in Relativity as it does with Newton:
ai ¼ 0 ðB:16Þ

for zero force. Here, we have written the acceleration as having four components
instead of three. This is reasonable as after all, the arena of investigation in
Relativity is four-dimensional spacetime rather than three-dimensional space. You
might wish to delve more deeply into the mathematics of Relativity in the
suggested references where four-dimensional vectors and tensors are discussed.
Appendix B 141

Now suppose we were to examine the system from the viewpoint of an

accelerated reference frame. Then the mathematical transformation of this vector
ai turns it into a different form, aiintrin , called the ‘‘intrinsic acceleration’’, in terms
of the new coordinates, and it is equal to zero as before:

aiintrin ¼ ai þ Cikl uk ul ¼ 0: ðB:17Þ

Here, the C is the Christoffel symbol as discussed above and uk is the four-vector
velocity, appropriate to Relativity, just as ai is the previous four-vector
acceleration. From the Equivalence Principle, gravity is locally like being in an
accelerated reference system. Therefore it is natural to take (B.17) to be the
equation of motion of a free body in full-blown General Relativity, i.e. when the
body is moving in curved spacetime, not only when it is moving under pseudo-
gravity. But you might well object to the effect that when there is real gravity
present, like the pull of the earth, it is no longer a force-free situation. This is
because you are so conditioned to thinking of gravity as a force. General Relativity
tells you that you must think differently about gravity. Gravity is now removed
from the collection that we call ‘‘forces’’ and is identified with spacetime, that is,
its attribute that we call ‘‘curvature’’. And when you ask: ‘‘How does a body move
under this new expression for gravity when all the forces are absent or balanced
out?’’, the answer is: ‘‘freely’’, according to (B.17), the ‘‘geodesic equation’’.

References

1. F. I. Cooperstock, 2009 General Relativistic Dynamics World Scientific Publishing,

Singapore.
2. H. Bondi in Lectures on General Relativity, Brandeis Summer Institute in Theoretical Physics
eds. S. Deser and K.W. Ford, Vol.1, Prentice-Hall, New Jersey, 1965.
3. L. D. Landau and E. M. Lifshitz, The Classical Theory of Fields Fourth revised English
edition, Pergamon Press, Oxford, 1975.
4. R. P. Feynman, R. B. Leighton and M. Sands, The Feynman Lectures on Physics, Addison-
Wesley, Reading, Mass., 1964.
5. W. Isaacson, Einstein, Simon and Schuster, New York, 2007.
6. J. L. Synge, Relativity: The General Theory, North-Holland, Amsterdam, 1966.
7. B. Spain,Tensor Calculus, Oliver and Boyd, Edinburgh, 1960.
8. R. Adler, M. Bazin and M. Schiffer, Introduction to General Relativity, McGraw-Hill, New
York, 1965.
9. J. L. Synge, Proc. Roy. Irish Acad. A53, 83, 1950.
10. F. I. Cooperstock, S. Jhingan, P. S. Joshi and T. P. Singh, Class. Quantum Grav. 14, 2195,
1997.
11. F. I. Cooperstock, Ann. Phys. N.Y. 282, 115, 2000; F. I. Cooperstock and S. Tieu, Found.
Phys. 33, 1033, 2003.
12. M. Gurses and F. Gursey, J. Math. Phys. 16, 2385, 1975.
13. W. B. Bonnor, Commun. Math. Phys. 51, 191, 1976.
14. A. Papapetrou, Ann. Phys. (Leipzig) 20, 399, 1957; 1, 185, 1958.
15. F. I. Cooperstock, Ann. Phys. N. Y. 47, 173, 1968.
142 Appendix B

16. H. Bondi, M. G. J. van der Burg and A. W. K. Metzner, Proc. Roy. Soc. A269, 21, 1962.
17. J. Madore, Ann. Inst. Henri Poincare’ 12, 365, 1970.
18. F. I. Cooperstock and D. W. Hobill, Phys. Rev. D20, 2995, 1979.
19. H. Bondi, Proc. Roy. Soc. London A 427, 249, 1990.
20. M. J. Dupre, gr-qc/0903.5225.
21. F. I. Cooperstock and M. J. Dupre, gr-qc/0904.0469; Int. J. Mod. Phys. D19, 2353, 2010.
22. R. C. Tolman, Relativity, Thermodynamics and Cosmology. Clarendon, Oxford, 1934.
23. F. I. Cooperstock and S. Tieu, astro-ph/0507619; astro-ph/0512048; Mod. Phys. Lett. A. 21,
2133, 2006; Int. J. Mod. Phys. A22, 2293, 2007.
24. F. I. Cooperstock and S. Tieu, Mod. Phys. Lett. A. 23, 1745, 2008.
25. A. Mahdavi et al, astro-ph/07063048.
26. A. S. Eddington, Proc. Roy. Soc. A 102, 268, 1922; The Mathematical Theory of Relativity,
Cambridge Univ. Press, Cambridge, U.K., 1923.
27. W. J. van Stockum, Proc. R. Soc. Edin. 57, 135, 1937.
28. W. B. Bonnor, J. Phys. A: Math. Gen. 10, 1673, 1977.
29. B. Fuchs and S. Phleps, New Astron. 11, 608, 2006.
30. J. D. Carrick and F. I. Cooperstock, astro-ph/1101.3224; Astrophys. Space Sc. 337, 321, 2012.
31. S. M. Kent, Astron. J. 93, 816, 1987.
32. J. M. Overduin and F. I. Cooperstock, Phys. Rev. D 58, 043506, 1998.
33. S. M. Chitre and J. V. Narlikar, Astrophys. Space Sc. 44, 101, 1976.
34. K. Gödel, Mod. Phys. 21, 447, 1949.
35. I. Ozsvath and E. Schucking, Am. J. Phys. 71, 801, 2003.
36. S. Chandrasekhar and J. P. Wright, Proc. Natl. Acad. Sci. U.S.A. 47, 341, 1961.
37. S. W. Hawking and G. F. R. Ellis, The Large Scale Structure of Spacetime, Cambridge
University Press, Cambridge, 1973.
38. F. I. Cooperstock and S. Tieu, gr-qc/0405.114; Found. Phys. 35, 1497, 2005.
Index

A CTCs closed timelike curves, 45, 127–129

Abell520, 120 CMB cosmic microwave background
Accelerated expansion, 123 radiation, 120, 124
Acceleration COBE cosmic background explorer, 121
intrinsic, 59, 60, 141 Cold War, 16
of reference frames, 61 Coordinates
Absolute Cartesian, 66, 88
past, present, future, 101 comoving, 73, 108
rigidity, 36 cylindrical, 57, 108, 109
separated, 47 spherical polar, 57, 66, 67, 87, 95, 108
time, 2, 17, 26, 27, 30, 45 Copernicus N., 63
Cosmic censorship hypothesis, 76
Cosmological term, 124, 125
B Cosmological Principle, 94
Bekenstein J. D., 104 Covariant derivative, 139
Big Bang, 2, 101, 102, 121, 131, 132
Big Crunch, 2, 22, 102
Binary pulsar, 69, 80 D
Black body, 84 Dark energy, 6, 120, 124, 125, 134
radiation, 134 Dark matter, 5, 6, 103–106, 119–122
Black hole, 4, 47, 72–76, 117, 118 de Broglie, 84
Bohr N., 53 Doppler factor, 19, 30
Bondi H., 2, 18, 25, 28, 30, 61, 77, 86, 139 shift, 70
Bonnor W. B., 83, 107 Duality0 wave-particle, 84
Brahe T., 63, 64 Dublin Institute for Advanced Studies, 17
Brandeis Summer Institute in Theoretical Dupre M. J., 86
Physics, 30 Dust, 5, 6, 83, 89, 107, 108, 116–118

C E
Causality, 127 Earth, 16, 45, 52–56, 60, 61, 63, 69, 71, 91, 93,
Chandrasekhar S., 128 94, 102, 103, 119, 120
CERN, 103, 122, 133 Eddington A. S., 71, 106
Charge, 15, 79, 85, 137–139 Einstein
accelerated, 15 field equations, 3, 58, 65, 67, 68, 80, 86,
conservation, 137, 139 87, 98, 109, 129, 137, 139
Christoffel symbol, 60, 140, 141 Rosen bridge, 75–76

F. I. Cooperstock and S. Tieu, Einstein’s Relativity, 143

DOI: 10.1007/978-3-642-30385-2, Ó Springer-Verlag Berlin Heidelberg 2012
144 Index

E (cont.) HI radiation, 112

summation convention, 57 Hubble Law, 94
tensor, 58, 140 Hubble parameter, 96, 99
Electromagnetic, 14, 22 Huygens, 12
fields, 15, 79
spectrum, 14
Electron, 15, 16, 81, 84, 121, 133 I
relativistic, 49 Inertial observer, 10, 30
Energy-momentum Institute for Advanced Studies
conservation, 3, 134 Princeton, 128
Energy localization, 4, 82 Interference, constructive,
hypothesis, 83, 86 destructive, 13
Equivalence‘ Principle of, 3, 54, 55, 57, 60, Intervals
139–141 null or light-like, 42
Ether, 12, 104 timelike, 44
Euclid, 95 spacelike, 46
Event horizon, 75, 76, 117 spacetime, 56
Expansion, accelerated, 6 Intrinsic acceleration, 141
Expansion, isotropic, 130 Isotropy, 95, 96

F K
Fantasy, 6 Kent S. M., 112
Feynman R. P., 86, 124 Kepler J. laws, 63
Finkelstein D., 73 Key to the cosmos, 5, 83
Finzi A., 104 Kolb E. W., 125
Free-fall, 5, 108 Kragh H., 94
Friedmann universes, 96 Kruskal M., 73

G L
Galaxy, 5, 6 Least Action, Principle of, 48, 60
cluster, 6, 105, 115, 118 Lamb shift, 134
Coma, 6, 103, 106, 115, 116, 118, 119 Landau L. D., 137
halo, 5 Laplacian operator, 140
Gamma-ray burst, 16 Leibniz, 9
Gauss divergence theorem, 137 Lensing, 119
General Motors, 17, 18 LHC, Large Hadron Collider, 122
Geodesic, 65 Lifshitz E. M., 137
equation, 3, 59, 60 Light cone, 44–46, 75
Gödel K., 25, 127, 128 London, 60, 61
Gradient, 59 Lorentz contraction, 13
Gran Sasso, 104, 121, 133 Lorentz invariance, 133
Gravitational, 47 Lorentz transformation, 27, 33
collapse, 47
redshift, 70
waves, 4, 79, 82, 83 M
Gravitino, 121 Madore J., 86
Graviton, 83, 84 Maric M., 54
Mars, 45, 46
Maxwell, 85, 133
H Mercury, 64, 65, 69, 104, 124
Harrison E. R., 93 Metric tensor, 57, 58, 60, 61
Hawking S., 74 Michelson-Morley, 13
 Index

Microwaves, 15 Q
Milgrom D., 104 Quanta, 14, 16, 83, 134
Milky Way, 5, 94, 107, 110, 112 Quantization
Modulation Quantum field theory, 6, 124, 133, 134
amplitude AM, 15 Quantum mechanics, 14, 53
frequency FM, 15 Quasars, 15, 71
Moffat J. W., 104
Moon, 52, 91, 92
Multiverse, 6, 130, 131 R
Radiation pressure, 82
Red shift, 22, 93, 96, 98
N Reference frame, 10
Neptune, 64 accelerated, 3
Neutrino, 46, 121, 133 inertial, 10, 12
oscillations, 134 Ricci tensor, scalar, 87
Neutron star, 69 Rigidity, 36
Newtonian Rosen N., 53
principle of relativity, 11, 14 Rotation curves, 1, 106, 112, 134
field equation, 58, 80, 140 Rubin V., 103, 106, 119
gravitational potential, 61, 70, 72, 80
laws of motion, 9, 11, 52, 59, 105
law of universal gravitation, 64 S
virial theorem, 119 Schwarzschild, 3, 65, 67, 72, 75,
Nobel Prize, 16, 123 76, 83, 116
Non-linear, 5, 58, 67, 106, 118 singularity, 76
synchronous form
Shapiro I. I., 71
O Silberstein L., 94
Occam’s razor, 134 Simultaneity, 24–26, 34–36, 38, 46
Olbers H. W., 93 Singularity, 72, 73, 76, 118
Oppenheimer J. R., 117 naked, 76, 118
Solar eclipse, 71
Solar system, 63, 64, 91, 92
P Sound speed, 43
Papapetrou A., 83 Space/time reciprocal
Paradox, twin or clock, 2, 18, 28 relationship, 31
Parallel universes, 131 Spacetime energy-momentum, 88
Particle physics, 6 Spacetime, higher dimensional, 11
Penrose R., 76 Spacetime interval, 41–43, 46, 56,
Perihelion precession, 106 68, 135, 136
Photino, 121 Speed of light, 13–15, 23, 27, 28, 33, 41, 42,
Photoelectric effect, 16, 84 44, 72, 104, 117, 133
Photons, 12, 14, 16, 69, 70, 83, 121, 133, 134 Spin, 84, 115, 134
Planck’s constant, 83 Standard Model, 121
Potential, gravitational, 4, 5, 52, 59–61, 70, 80, Stationary, 87, 88, 107, 108, 115, 118
86, 137, 140 Sun, 4, 13, 15, 52, 57, 63–65, 69–71,
Pound- and Rebka experiment, 70 91–94, 105
Proper length, 33, 38 Supernovae, 75, 123, 124
Proper time, 33 Supersymmetry, 121, 122
Proper volume, 87 Synge J. L., 2, 3, 17, 60, 61, 72, 73, 117
Pseudo-gravity, 60, 139, 141 Synge J. M., 17
Pseudotensor, energy-momentum, 82 Szekeres G., 73
Pulsars, 15, 69, 80 Szekeres P., asymmetric collapse
Pythagoras, 40 of dust, 83
146 Index

T Vessot and Levine, 70

Tensor, 57 Virial theorem, 119
energy-momentum, 80, 82 VHF very-high-frequency, 15
metric, 57, 58, 60, 61, 67, 68, 72, 73, 80, 83, Vulcan, 65
84, 87, 88, 94, 95, 108, 109, 116, 139
rank, 57, 88, 138
Test, 1, 3, 16, 36, 63, 69–73, 85, 105, 131 W
Thought or gedanken experiment, 53, 54, 85 Waves
Time machines, 45, 129, 130 electromagnetic, 14, 82
Tolman R. C., 87, 88 gravitational, 79–83, 86, 106
Twin or clock paradox, 2, 18, 28 sound, 12, 44, 79
Wave-particle duality, 86
Wesson P. S., 93
U White dwarf, 69
UHF ultra-high frequency, 15 White hole, 74
Ultra-violet catastrophe, 84 Wifi, 15
Universe, 6 WIMP, 103–104, 121
Inflationary birth, 77 Wirtz K., 94
Worm-hole, 7, 129, 130
WMAP, 121
V
Vacuum, 6
van Stockum W. J., 107, 109 X
Vancouver, 60, 61 Xenon100, 104, 121
Velocity X-rays, 15, 16
definition, 11
four-, 60, 88
in Hubble law, 94 Z
of stars in galaxies, 103–105 Zwicky F., 103, 106, 115, 119
in general relativity, 116, 117, 119
Venus, 71