BEFORE TIME BEGAN

BEFORE TIME
BEGAN
The Big Bang and the Emerging Universe

Helmut Satz

OXFORD
UNIVERSITY PRESS
 OXFORD
UNIVERSITY PRESS

Great Clarendon Street, Oxford, OX2 6DP,

United Kingdom

Oxford University Press is a department of the University of Oxford.

It furthers the llniversity's objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries

©Helmut Satz 2017

The moral rights of the author have been asserted

An earlier version of this book has appeared as

Kosmische Daemmerun8 with C. H. Beck Verlag, Munich 2016.

First Edition published in 2017

Impression: 1

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above

You must not circulate this work in any other form

and you must impose this same condition on any acquirer

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America

British Library Cataloguing in Publication Data

Data available

Library of Congress Control Number: 2017943713

ISBN 978--0-19--879242--0 (hbk.)

Printed and bound by

CPI Group (UK) Ltd, Croydon, CRO iYY

Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Quid faciebat deus, antequam faceret caelum et terram?
What did God do before he made heaven and earth'
AuGusnNus (AD 354-430 )
CoNFESSIONES u/12
 Preface

Ever since humans began to think about the world in which they
lived, they have asked themselves how this world came to be, why
it is as it is, how it was before, how it will be in the future, and what
role we will continue to play in it. All human civilizations try to give
answers to these questions. Wherever we look, we see happenings
which have a beginning and an end, just as our lives do. Did our
universe have a beginning, and will it have an endl When we stand
at night beneath the immense starlit sky, it seems natural to ask how
all that was formed, how large it is, and what will become of it. We
are such a tiny part of something so great, but still we can ask such
questions, and perhaps it is that which makes us special.
The search for answers to these questions has led to religions,
to epics, to a wealth of philosophies, and eventually, of course, to
natural science. In science, cosmology has developed as the relevant
branch of research, and what we want to say here is largely based on
studies in its framework. Nevertheless, even if different answers to
the basic questions have in the course of time seemed to contradict
each other, it appears that these differences were rather superficial,
and that the essentials are more in accord than initially thought.
And we find today that cosmology is entering regions which for
many physicists are no longer science, since they are no longer
experimentally explorable. Nevertheless, they remain challenges to
human thinking.
In this book I want to describe the essential steps in the forma­
tion and evolution of our universe, as they have proceeded accord­
ing to the presently prevalent ideas in cosmology and physics. We
will see that much is indeed in accord with previous, less scientific
views. Already more than two thousand years ago, much of what
today is considered as the latest scientific result was proposed sim­
ply as the outcome of logical thinking. What has been added from
the time of Galilei on is certainly the insistence that all conclusions
viii Preface

must be confirmed by experiment. It is only that which turns met­

aphysics into physics. Nevertheless, we note that in recent times
more and more interesting ideas have been pursued, from a multi­
verse and parallel universes to wormholes through space and time,
even though in our present world these concepts are not accessible
to experiment. The world of the imaginable remains much larger
than that of the testable, and so in the future, even in natural sci­
ence, concepts and ideas may well survive without presently having
a chance to be verified experimentally.
The basic questions we want to address here fall into three areas:

• How and out of what was our universe formed, our world
in time and space? What was before, and what will happen
afterwards?
• What are the basic building blocks of matter in our present
world, and what forces determine their binding?
• How could a uniform, structureless, primeval world lead to
the present multitude of forms and structures?

The answers we have today to these questions are, as already indi­

cated, still of a somewhat speculative nature, and they are certainly
not accepted by all. But I believe that they are interesting enough to
pursue them further. That is the aim of this book.
Less than thirty years ago, the first of these questions was still
ruled out as politically incorrect. The beginning was the Big Bang,
and "before" made no sense; there was no "before." Today, many
cosmologists and physicists picture the birth of our universe as a
rapidly expanding bubble in a hot primordial world, one bubble
among many others. We are witnessing today a second Copernican
revolution: neither our solar system nor our galaxy nor our uni­
verse are the end of all things. Beyond our world there are innumer­
able others, similar or not similar to ours, worlds we can never reach
but which should nevertheless exist. These views make the Big Bang
a physical process and not a singular occurrence--it is not a unique
event; there were and there will continue to be others like it, and it
can also lead to an end.
 Preface ix

The question about the building blocks of our present world

and of its predecessors in earlier stages of development can today
be answered in a more extensive way than ever before, thanks to
progress in particle physics. The dream of a final theory, a theory
in which electromagnetic, strong, and weak nuclear forces are uni­
fied into a single interaction, is still not fulfilled, but has become
more conceivable. Such a theory of grand unification must describe
a primordial world of great symmetry, in which all constituents are
treated equally. The cooling of the universe then leads to the break­
ing of symmetries and thus to the different interactions. The role of
gravity in such a picture still remains enigmatic, however.
The development of the complexity of our world has also led to
different and perhaps even contradictory ideas. The crucial point
here is the famous second law of thermodynamics, insisting on a
development toward increasing entropy, and thus giving a direction
to time. Does that mean that our universe should become more and
more disordered and thus head toward an end without form and
structure? Here two ways out have been noted. First, the increasing
expansion of the universe can in the long run prevent thermal equi­
librium from ever being reached. Second, the role of gravitation as
the dominant force implies that in a cooling gas an equidistribution
of matter is not the stable form of the medium. As long as gravity
rules, a world of galaxies and radiation in empty space is thermody­
namically preferable to a uniform gas of particles.
So, for all three questions, much has happened in the past 30
years, and our thinking has been thoroughly modified. Two basic
approaches of physics-reduction (What are the smallest possible
building blocks of matter?) and extension (Earth, planets, solar sys­
tem, galaxy, supergalaxy, universe)-seem to have reached their
limits, through quarks and the multiverse, respectively. On the
other hand, the new concept of emergence has appeared on the
scene. We distinguish today fundamental observables and forces
(charges, atomic binding) from emergent observables and forces
(temperature, pressure) that arise from the collective effort of many
constituents. It seems meaningful to present these new views of the
x Preface

world to a general, nonspecialist readership, and that is what I want

to do here. In doing so, I want to concentrate more on the novel
concepts, notions, and ideas that have come up, and less on details
about who did what, when, and why. It could well be that the new
conceptual developments will eventually have profound conse­
quences in other areas of human thought-it would not be the first
time for physics. In any case, I want to show to all those interested
that we are today witnessing the emergence of new ideas in natural
science--ideas that so far have had a much greater impact on our
thinking than they have on our technology. Whether they will ever
result in advances of practical use to mankind-that remains to be
seen. But they have already fundamentally changed our view of the
world in which we live. The universe was reduced to one of the uni­
verses in the multiverse, one among many in a primordial world. We
have become a smaller part of a greater whole.
When writing this book and looking at the various approaches,
it became clear to me that in modern cosmology there still remain
different, even contradictory views of the beginning of our universe.
My guiding line therefore was the approach of the proverbial Amer­
ican baseball umpire who noted, "I calls them the way I sees them."
Others may have different views, and they may be as valid as the one
presented here.
A slightly earlier version of this book has appeared in German,
entitled Kosmische Dammeruns (Cosmic Dawn), from C. H. Beck Publish­
ers, Munich. I am very grateful to various colleagues here in Biele­
feld for discussions on many aspects of the subject; Frithjof Karsch
looked through all the entire earlier version and helped improve it
greatly. Particular thanks also go to Paolo Castorina from the Uni­
versity of Catania, Italy, who was always ready to join me enthusiasti­
cally even in the pursuit of topics that traditional physics considered
as devious. Last but certainly not least, most sincere thanks go to my
wife, who once more found herself in the role of wholeheartedly
supporting the musician without hearing the music.

Helmut Satz
Bielefeld
November 2016
 Contents

1. Before the Big Bang

The multiverse 4
Space energy 8
The Big Bang 13
Inflation 15

2. The First Particles 17

Matter particles and force particles 19

Fermions and bosons 23
Quarks and leptons 27
The birth of matter 29

3. Empty Space 35

The birth of the vacuum 38

Hadrosynthesis 40
N ucleosynthesis 43
Atoms 46

4. Transitions 47

Critical behavior 54
Baryogenesis 59
The birth of light 62
The effective quark mass 65

5. The Light of the Big Bang 77

Spaceship Earth 81
The sound of the Big Bang 84
The form of the space 87
Star rise 96
xii Contents

6. Structure and Form 97

The course of evolution 101

Entropy 105
Structure in the universe 109
Expansion vs. relaxation time 113
Cluster formation 117

7. Dark Corners 121

Black holes 122

Dark matter 127
Dark energy 132

8. The End of Time 137

Three possibilities 140

The ultimate nightfall 141

Appendix 1: How Many Configurations of Balls Are There? 145

Appendix 2: Orbits and Dark Matter 147

Appendix 3: Cosmological Constant and Dark Energy 149

Some references for further readinB 153

Person Index 155
Subject Index 157
 1

Before the Big Bang

In the beginning, God created the heaven and the earth.

THE BIBLE, GENESIS I.I

Before the Big Bang, there was no time and hence also no beginning.
In six days, so the Bible says, did God create the world. Although
omnipotent, he did not do it with one mighty stroke, but instead
spread creation over six days. Light appeared on the first day, land
and water on the second, their separation occurred on the third, and
so on. The theologians ofearlier times called the creation process the
Hexameron, the work of six days. Why did creation take so long, and

why did it pass through different stages? It would be too simple to

blame that on a human-like God who became tired in the evening­
after all, he had just created morning and evening-or who needed
more time for further planning. Instead, it seems more natural for
us to imagine that something as great and as complex as our present
world could simply not be formed at once, that even God could only
have it emerge from something simpler "in the course of time."
That in fact shows the essential result of a sequence of succes­
sive steps of creation. Had the world been made with one stroke, it
would have been timeless. Only the successive occurrence of differ­
ent phases of creation introduced the idea of time, a course of events
and intervals between them, defining a scale. The progression of
happenings also defined a direction of time, and the first event in this
progression then became the beginning.
In a similar way, space first appeared. Only the separation of
heaven and earth, ofland and water, defined a space in which certain
2 Before Time began

Hildegard von Bingen (1098-1179)

The Hexameron

things are here, others are there; some on top and others on the bot­
tom. Thus, the stage for the coming play first had to be created, and
so, in today's thinking, the opening line of Genesis could well be "In
the beginning, God created space and time."
This makes it natural to ask what was before, how and from what
space and time were formed. What did God do before he created
heaven and earth, or as we read it, before he made space and time?
This seemingly heretical question had in fact been put forth in early
Christian theology. Augustine Bishop of Hippo in North Africa and
one of the founding fathers of the Christian church, around AD 400
comments quite critically on the answer that before heaven and
earth, God made hell for those who ask such questions. Augus­
tine himself comes to the conclusion that before creation God did
nothing; the making of heaven and earth was also the beginning of
time and hence the beginning of all actions.
 Before the Big Bang 3

Similarly, today's cosmology is based on a Big Bang as the begin­

ning of our universe. Distant galaxies recede from us at ever­
increasing speed, and if we let this film run backward, we come to
a point at which it all started, a point that the Belgian priest and
physicist Georges Lemaitre proposed as the beginning of our world.
And, just as the work of God before creation was ruled out, it was
for a long time not considered meaningful to ask about the state of
the world before the Big Bang, or how the Bang came to take place.
Not so long ago, even Stephen Hawking, famous for his Short History
of Time, noted that these questions were like asking what lies north
of the North Pole.
The past thirty years, however, have seen a change of paradigms,
leading to a new view of the world, one in which the question of
how and from what our universe appeared makes sense after all.
We do have a frame today, still speculative and certainly not accepted
by all scientists, that permits us to discuss in general terms the for­
mation of our universe. We can now relax and wonder what God did
before he made our world.
Our view of the world began with the Earth as the center of
everything. This geocentric scheme was then replaced by a heli­
ocentric one, with the sun at the center and the Earth only one
of the planets circling around it. Still later, it was found that the
sun, our sun, is only one of the millions of stars making up the
galaxy of the Milky Way. And today we know that there exist mil­
lions of similar such galaxies, spread out in space and moving away
ever farther through the expansion of space. Our own world thus
became an ever-smaller part of an ever-growing universe. The Big
Bang defined the beginning of this immense universe and led to
the idea that this is all there is, the universe. Four hundred years
ago, Giordano Bruno could imagine an endless succession of solar
systems; today's cosmologists can imagine an endless number of
universes, similar to ours or different, with the same or with other
laws of nature. Ours is but one among many universes in this infin­
ity, in what is now called
4 Before Time began

The multiverse.

How, when, and from what could all these universes have appeared
and continue to appear? What are the essential features of the
multiverse?
The world before our time began is the primordial world. At that
time, there was no time in our sense. The course of time requires a
sequence of different events, allowing us to identify before and after.
The primordial world has no beginning and no end, no before or
after, no earlier or later, no here or there, no top or bottom, no large
or small, no form or structure. Two thousand years ago, the Rigveda,
the Indian epic of creation, said that

At that time there was neither being nor not being

Only darkness hidden by darkness,
And invisibly all moved about.
As an accidental fluctuation in this primordial world, like a bubble
in hot lava, our universe appeared, with its space and its time. The
Rigveda says

The One was born from the power of Fire,

but in the same way many other bubbles, many other universes
appeared as well, and they continue to appear. It is not easy to envi­
sion such a primordial world and the eternal creation it provides.
Space and time are such basic features of our world that it is difficult
to imagine a world without them. We can try, however, by starting
with one design, and then keep on correcting that as we go on.
Let us imagine an immense container filled with water, an ocean
at fixed temperature and without any external interference. We dive
deep into this water, far from all confining boundaries. Here we find
only uniform water, yesterday, today, tomorrow; time is meaning­
less, and the idea of space as well, since a shift in position, no matter
in which direction, does not lead to any change in the world around
,
us. And if we picture such an ocean placed in interstellar space,
without any effects of gravitation, there would also not be any up
or down.
 Before the Big Bang 5

As soon as the chosen temperature

approaches the boiling point, small bubbles
ofsteam form, regions oflower density than ••
•
that ofthe surrounding water. In a terrestrial
environment these bubbles rise, escape into • • •
the air above the surface of the water and
• •
continue to expand. That would be a very •
•
simple picture with which we could start to •
•
describe the formation ofa universe: the hot
• •
water would be the primordial medium, the • •
•
multiverse, and the bubbles would later on
become universes of some kind. Boiling water
But we can make this even more interest-
ing. Ifthe water is very pure and the walls ofthe vessel very smooth,
the temperature can be increased beyond the boiling point without
any effect, in practice up to some 110 degrees centigrade; physicists
call this superheatina. The water is now in a metastable state: any small
disturbance will cause an explosion, in which a big bubble of steam
escapes from the medium. Once the bubble is out of the water, its
density will decrease more and more; now there is a time with a
direction. Before, in the water, that was not the case: water is water,
today or tomorrow; the question of time does not arise. Time only
comes into play once the bubble has escaped, once a chain ofevents
appears. The water molecules now separate more and more and
travel into ever more distant regions ofspace. The Big Bang, accord­
ing to today's experts, was a somewhat similar process.
Let's stay with this experiment for a moment. Under fixed condi­
tions, water is in a normal state: under atmospheric pressure at sea level,
it is ice below 0 degrees centigrade, liquid from 0 to 100 degrees, and
steam above 100. The transitions from one normal state to another,
like melting or evaporation, are called phase transitions. But we just
saw that ifwe are really careful, we can heat water some ten degrees
above the boiling point without evaporation setting in. It is still
liquid under conditions when it should really be steam. It is now,
using physics terminology, in a false normal or ground state, and any
6 Before Time began

disturbance will cause it to flip to the right one. Superheated water

thus was artificially brought into an unstable state of too high an
energy, and in the transition from wrong to right it will liberate this
energy: everything flies apart, the bubble explodes. Incidentally, a
similar phenomenon occurs if we cool the water carefully below the
freezing point; any disturbance then causes sudden ice formation.
Supercooled rain coats everything it hits with a layer of ice.
In our daily world, we encounter many other such cases. A well­
known instance is the ball on the hill; here as well, any slight dis­
turbance will cause it to roll down. On top, in the wrong, unstable
state, the ball has a higher gravitational potential energy than it
would have on the bottom. This higher potential energy then turns
into kinetic energy of motion when the ball rolls down.

The ball on the hill

What we have said so far points us in the right direction, but it is

still too much based on our familiar terrestrial world. The universe
has features that cannot really be modeled by water. In particular,
we know today that the universe is not static; it is undergoing con­
tinuous spatial expansion. Our terrestrial world provides a static
stage for the ongoing events, but in the cosmos things look differ­
ent. Distant galaxies are becoming ever more distant and eventually
even disappear; on large scales, everything is expanding, and this has
serious consequences.
 Before the Big Bang 7

Imagine you are on one side of a large room and now walk toward
a door on the opposite side, with your normal walking speed of
about a meter a second. If the room expands by more than a meter
a second as you are walking, you will never reach the other side. In
fact, even though you walk and walk, the door becomes ever more
distant.
A similar fate is in store for an ant trying to crawl on an expand­
ing balloon from the equator to the North Pole. If the balloon is
blown up rapidly enough, the poor ant will never reach its goal; the
pole will move farther and farther away.

The ant on the balloon

We conclude from these examples that a sequence of events can

be seriously modified by spatial expansion. The crucial question is
evidently how the rate of space expansion compares to the time
scale of the process we are looking at.
Before we return to the expansion of the cosmos, we have to
address another essential problem: there has to be some agent caus­
ing this expansion, since gravitation provides an attractive force
between all stellar objects and, as we know from Einstein, even
between clusters of normal energy. The mysterious something that
not only compensates this attraction, but even overcomes it to cause
expansion could be called
8 Before Time began

Space energy,

but in common usage it is now generally referred to as dark enerBY.

From the point of the view of cosmology, empty space is therefore
not really empty; it is filled uniformly with an invisible energy of
very low density, driving it continuously to further expansion. Nor­
mally, such an expansion would cause the energy density to drop;
here, however, that is not the case: it remains constant in space and
time, having everywhere the same value, usually denoted by A.
Such a concept was first introduced by Einstein, who called it the
cosmolo3ical constant. At first sight, such a constancy in space and time
does not seem possible: if the total volume grows and the energy
density remains constant, the total energy of the universe is con­
tinuously increasing. Where does that energy come from? Isn't that
a violation of energy conservation?
The solution to this puzzle is found in the description of gravity
through space deformation, as it is formulated in the general theory
of relativity. We saw a first instance of this in the case of the ball on
the hill. The kinetic energy of the ball is zero on top of the hill, but
clearly not zero when the ball hits the bottom. How is this kinetic
energy created? The topology of hill and valley leads to a difference
in gravity, and that difference in gravitational potential energy pro­
vides the needed kinetic energy. In a similar way, energy and space
are related in the cosmological world: the increase of space energy is
paid for by an increased deformation of space. The bigger the effec­
tive mass of the universe becomes, the bigger the negative potential
energy of the resulting gravitation. Only in this way was it possible
to extract a universe as "the ultimate free lunch" out of the primor­
dial world: what was gained in mass was paid for by the resulting
negative potential energy of gravitation.
The present density of dark energy in our universe is extremely
low, as already mentioned: its mass equivalent in a volume of the size
of the Earth is about one-thousandth of a gram. In stellar dimen­
sions, this does not lead to noticeable gravity effects, and it also
does not affect physical processes. Our planetary system and even
 Before the Big Bang 9

the Milky Way remain unchanged; gravity still wins. Only on cos­
mic scales-the astrophysicists speak of intergalactic scales-do we
find an effect: since the density is constant over the entire universe,
the sheer total amount of dark energy can now overcome gravity
and cause a repulsion. In our solar system, the total amount of dark
energy is negligible, but for the total observable universe it adds up
to three-quarters of the total energy, and because of the expansion,
that fraction is continuously increasing.
According to present cosmology, the dark energy in our uni­
verseafter the Big Bang plays the role of the steam in the escaping
bubble. Before the bubble escaped, it contained a primordial medium
of immensely higher density. This created a correspondingly higher
rate of expansion than the present one; the primordial world
expanded and continues to expand much more rapidly than our
universe. The primordial medium is the counterpart of the super­
heated water; it is in a wrong normal state, and any small local per­
turbation can create a bubble in which the transition to the correct
normal state sets in. The high density of the primordial medium
lets the bubble initially expand with the dramatic primordial rate,
until the wrong normal state has been converted into the right, and
the dark energy density has dropped from its primordial value to its
present one. From this point on, the expansion rate also decreased
correspondingly. The energy difference between the high false pri­
mordial density and the lower stable one today is liberated in the
transition, and eventually gives rise to all the matter in our present
universe. All this makes the Big Bang a "normal" physical process;
nevertheless, for us it is the beginning of what we consider as space
and time.
The origin of our universe as a bubble of primordial medium is, as
indicated earlier, a new view of the world, only some thirty years old;
a major role in its formulation was played by the American physicist
Alan Guth and the Russian-American physicist Andrei Linde. Until
then, our universe was considered as ultimate and final, not as part
of something bigger, and for that reason, the question "What was
before the Big Bang?" had been ruled out.
10 Before Time began

Alan Guth Andrei Linde

Photo: Jenny Guth Photo: Linda Cicero, Stanford
News Service

Nevertheless, it had been found that different observations-we

will soon come back to them-required that very shortly after the Big
Bang, the universe must have gone through a phase of extremely rapid
expansion, an "inflation." Attempts to justify such an inflation then
led to the new primordial scenario presented here. This scenario has
other unavoidable consequences as well. Our universe emerged from
a bubble of the primordial medium; however, that medium continues
to exist and expand further, so that new bubbles continue to appear
forever. Hence, the primordial medium is a multiverse, now and forever
giving rise to new universes; ours is only one of all these uncounted
worlds. In the following picture, we have for simplicity made the new
universes round; in reality, they can be arbitrarily irregular.

time time

Two universes appear out of the primordial multiverse

 Before the Big Bang 11

Because the other universes remain forever inaccessible to us,

we will never know if the same laws of nature hold there, or if
they might even hold intelligent life. For us, they are beyond any
form of investigation, and for this reason many physicists consider
the idea of a multiverse to be heretical, more metaphysics than
physics. One of the most eloquent critics is the American theorist
Paul Steinhardt, even though he initially helped to formulate the
idea of a multiverse. On the other hand, a single unique universe
also leads to problems such as inflation, which are difficult to com­
prehend in a conventional scheme and seem to require further
explanation.
On the largest cosmic scales (the mentioned intergalactic scales),
our own universe is, as far as we can tell, quite uniform: it is homo­
geneous as well as isotropic, i.e., it is of the same form in different
regions and in all directions. This feature is generally referred to as
the cosmoloBical principle. Of course, it only holds for an average over
immense regions; in our solar system and on to the Milky Way, our
world is certainly not uniform. It only becomes so once we aver­
age over cosmic distances; different spatial regions and directions
then become indistinguishable. While that applies to space, it is
certainly not the case for time: ever since the Big Bang, our world
has been steadily and dramatically changing; there is a direction,
an arrow of time that leads from a hot primeval gas to a cosmos
of galaxies.
Things look different in the case of the multiverse. Here as well,
one cannot distinguish sufficiently large spatial regions, but they
are far from uniform. In the false normal state of the primordial
medium, one finds everywhere expanding bubbles undergoing tran­
sitions to the right normal state, creating future universes. And the
multiverse itself continues to expand dramatically; it is a bit like a
superheated soup, from which bubbles are continuously escaping.
This behavior does not change with time; it goes on forever, so that
for the multiverse, space and time are now on equal footing: the
bubbles of the future universes appear here and there, bigger or
smaller, now or later. The superheated multiverse itself undergoes
an ever-increasing expansion as well.
12 Before Time began

A medium showing such irregularities is today called fractal.

This concept was introduced by the French mathematician Benoit
Mandelbrot; it refers to complex structures formed through the
repetitions of a given form at different scales; it is also designated as
self-similar. In the case of mathematics, one considers such structures
in space, but in the case of the multiverse, they appear in both time
and space: bubbles appear here and there, now and later, larger and
smaller. For illustration, we show in the following figure one such
structure; it is a triangle formed through the repetition of black and
white triangles of different sizes and named after its inventor, the
Polish mathematician Wadaw Sierpinski.

The Sierpinski triangle

In the case of the multiverse, one has to picture the whole struc­
ture to be in continuous expansion, both in space and in time. And
even then the picture is not quite correct: if we take the white tri­
angles to be new universes, then, as far as we know today, the black
primordial world expands much faster than the white, leading in
turn to more new universes.
 Before the Big Bang 13

In the end, that means our entire universe, from the Big Bang
until today, is only one of an ever-growing number of universes,
all created in the same way. Copernicus had abolished the idea that
the Earth was the center of the world and thus taken away our spe­
cial cosmic status. Today's cosmology takes such a status away even
from our universe. It is only one of the many little triangles; we
don't know and never will know what is happening in the others.
Such a scenario of course leads to various questions. Our present
world is possible only because our constants of nature have the
values they have. If, for example, electrons were much heavier or
protons could easily decay, a world such as ours could not exist. So
why do they just have the values required for our existence? One
proposal-not really all that satisfactory-is that the values of
the constants of nature in the different universes of the multiverse
have randomly distributed values, and ours just happen to have the
ones they have.
Up to now, we have considered the formation of our universe
"from the outside"; let's now return to our own world. In the
twenties of the last century, Edwin Hubble made astronomical
observations at the Mount Wilson Observatory in California; they
showed that distant galaxies are moving away from us, and that
their recession speed increases with their distance from us. The
reason for this, as already mentioned, is the expansion of entire
universe, driven by what remained of the dark space energy in its
present normal state. If we now have the film run backward, we
come to a beginning, to

The Big Bang.

Subsequent measurements indicated that this must have been about

fourteen billion years ago. In 1964, the American physicists Arno
Penzias and Robert Wilson found what has become the most essen­
tial confirmation of the Big Bang theory: the cosmic microwave backwound
radiation as the surviving light from the Big Bang.
14 Before Time began

Edwin Hubble (1889-1953)

Part of the energy liberated in the transition from wrong to right

normal state eventually turned into photons, particles of light. In
the early stages of development, these photons suffered interactions
with all other charged particles that were created. But sometime
later on-we shall return to this in detail-many of the charged
particles annihilated each other, bringing still more photons into
the game, and they all were in interaction with the remaining
charges. In the long run, however, those remaining charges com­
bined, forming neutral atoms from positive protons and negative
electrons. For the photons, that was the ultimate liberation: they
were now decoupled from matter and since then, they could travel
on freely. This decoupling or last scattering occurred some 380,000
years after the Big Bang. The photons present at that time have ever
since been on their way, flying freely through space. Space, how­
ever, kept expanding, and, as a result, the wavelength of the pho­
tons increased correspondingly. Because the wavelength determines
the temperature of the photon gas, the implication is that the tem­
perature of the cosmic background radiation dropped from about
3000 degrees Kelvin at the decoupling time to about 3 degrees today.
It was just this 3 degree radiation that Penzias and Wilson found.
By now, there are many ever more precise measurements of that
radiation, in all regions and directions of the sky. And wherever one
measures, one finds the same 2.72548 ± 0.00057 degrees Kelvin, up
 Before the Big Bang 15

to the fourth decimal position. The universal remaining light ofthe

Big Bang is thus known very well and it is with a high degree ofpre­
cision the same everywhere. That, as we shall see shortly, creates a
serious problem.
At the time of decoupling, a small region of the universe at the
eastern rim of the sky was many light years distant from one at the
western rim. Between the two regions, no communication was pos­
sible; each was outside the other's causal horizon. So how could they
synchronize their radiation temperatures with such precision?
How could the conductors of two orchestras in absolutely separate
concert halls, one in London and one in Sydney, manage to start
the identical music at precisely the same instant? This puzzle, this
horizon problem, has bothered cosmologists for decades; today it is
thought that it is solved by the idea of

Inflation.

At the beginning, there was a small bubble of hot, dense primordial

matter in thermal equilibrium, like a gas at fixed temperature and
pressure. Then, suddenly, the inflation set in, the transition from
wrong to right normal state. In an unbelievably short time, space
expanded by an immense factor, thus creating regions that could
no longer communicate. And in this expansion, even the slight­
est irregularities were largely smoothed out. Such a process lets us
understand how causally disconnected systems can still have the
same temperature: they were together in the same pot and under
the same conditions, until the explosion-like inflation pulled them
light years apart. What is separate today was once together and could
communicate. The relevant scales are illustrated in the following
picture. It shows that the inflation process had lengths increase by a
factor of 1026 in a time interval ofonly 10-34 seconds. This means that
the space of the primordial bubble expanded at a rate that surpasses
the velocity oflight by a factor of 1050• So it was indeed a very special
process, difficult to accommodate in the previous views ofphysics.
16 Before Time began

10-34 s

Cosmic inflation

The new view of the origin of the universe thus solves two puz­
zles at the same time. We see how our universe appeared as a bub­
ble in the primordial world, and how the inflation of that bubble
allows causally separate regions to exhibit the same conditions. The
cosmic background radiation thus contains more information than
was thought at first. It provides our only tool to look for something
remaining from the flash of the Big Bang, of the transition from
wrong to right normal state. And the universal temperature value
of the radiation implies that the transition must have been accom­
panied by a dramatic spatial expansion, by cosmic inflation. That
brings us to the next question: how could the transition terminate,
in these first instants of the Big Bang, with the formation of some­
thing like matted
 2

The First Particles

I call them elementary particles,

because they were the first and everything is made of them.
LUCRETIUS, THE NATURE OF THE UNIVERSE, ABOUT 55 BC

The first particles appeared when the metastable medium of the

bubble in the primordial multiverse had fallen from its hot but false
ground state into the much cooler right one. This drop suddenly
liberated a lot of energy. The transition-which in our picture of
superheated water corresponds to the explosive escape of an expand­
ing bubble of steam-leads here as well to a very violent explosion,
the "inflation" of the cosmologists. In an incredibly short time, some
10-34 seconds, a spatial expansion took place by a factor of 1026 or
more. At the end of this process, the system found itself abruptly in
the correct ground state, together with an excess of energy deter­
mined by the energy difference between false and right ground state.
There is no comparable "moment" in the physics of our world, and
no comparable expansion rate. The whole process, though a phe­
nomenon within the scope of physics, was an absolutely unique
event in the history of our universe, its birth: the Big Bang.
Let's look at the event in a little more detail. A density fluctua­
tion occurs in the boiling primordial medium of the multiverse;
quantum mechanics tells us that such fluctuations randomly appear
again and again, here and there, with a certain probability. Now, it is
sufficient if this disturbance creates a bubble in which the system is
driven from the false to the right ground state. As long as the energy
density in the bubble corresponds largely to the false ground state,
the bubble continues to expand with the dramatic rate of the multi­
verse. Then, suddenly, the transition, the crash into the energetically
18 Before Time began

much lower right ground state occurs, and the expansion is essen­
tially stopped: inflation is over. Our universe is born, and from now
on what becomes relevant is what so far was called the Big Bang cos­
mology: the development of the universe after the Big Bang.
The crash into the right ground state means, as mentioned,
that the energy difference between false and right ground states is
suddenly thrown into space. So what happens to this immediately
available excess energy? The example of the ball on the hill showed
that the potential energy present on top was transformed into kin­
etic energy at the bottom. And this could be employed in different
ways. If the ball were made of glass, the impact could lead to an
"explosion" into many fragments. In the case of the primordial bub­
ble, the liberated energy provided enough fragments to make up the
entire world to come.
To understand that, we first have to consider what happens when
energy is deposited in empty space. What actually is empty space? Here
we leave aside the dark or space energy mentioned in the last chapter;
it shapes space, makes it expand, but it does not provide its contents.
From the work of the British physicist P.A.M. Dirac, we know that
empty space is empty only at first sight: it is in fact actually a sea of
"submerged" virtual particles, which so far could not come into reality
because they lacked the necessary energy. Empty space is like a silent
lake, under whose surface all sorts of particles only wait to get the
energy for a jump upwards, into reality. What kind of fish are swim­
ming there? What are particles, and what species of particles exist?
In today's world, so particle physics tells us, there exists a multitude
of different species of particles, interacting with a number of different
forces. But this diversity has developed only in the course of time after
the Big Bang, just as the different forms of living beings have arisen
through evolution from a simpler primeval form. It is therefore not
surprising that physicists would like to find one such basic primeval
form, from which all others emerged. This ambition has led to various
attempts, but so far not to a definite solution; we return to that shortly.
Here, however, we note that shortly after the Big Bang two basic forms
had already arisen through the input of the available energy:
 The First Particles 19

Matter particles and force particles.

The former are the basic building blocks of matter; everything

is made of them. The latter convey the interaction between these
building blocks. They are the cement holding the blocks together,
and they are also emitted or interchanged when two matter par­
ticles collide, as a signal of such an event.
If we divide matter into ever- smaller constituents, we expect to
find a limit: the smallest building blocks of matter. Such a limit,
however, only makes sense if no further reduction is possible--if
there is an end to divisibility. Because of this, many early attempts
failed. The atoms of the chemists were found to consist of nuclei and
electrons; the nuclei of the physicists are made up of protons and
neutrons; and these, we believe today, in turn are bound states of
quarks. The Roman philosopher Lucretius, author of the celebrated
early treatise on The Nature of the Universe (source of the quote at the
beginning of this chapter), had therefore insisted over two thousand
years ago that the truly ultimate constituents of matter could never
have an independent existence, but that they could appear only as a
part of something bigger. If they could exist individually, we could
continue to ask what they were made of. Therefore, he concluded,
the ultimate building block can never exist by itself, but only as a primordial
part of a larBer body, from which no force can tear it loose. That would pro­
vide a logical end to divisibility. And today, quarks indeed fulfill this
requirement of Lucretius: they are, in the terminology of present
physics, subject to "quark confinement"-one quark can never exist
on its own. Nucleons consist of three coupled quarks, and no force
can ever split them. That was shown by numerous experiments, and
our basic theory of their interactions ("quantum chromodynam­
ics") contains such unbreakable binding as an essential prediction.
But right after the Big Bang, that was not a problem. There was
no empty space, no nothinB of any kind. The liberated energy was so
huge that it led to an immensely dense cloud of buzzing particles,
of quarks. No quark had ever to fear being left alone; everywhere,
in the most immediate neighborhood, there was a multitude of
20 Before Time began

other quarks. That leads to a rather amusing situation. As part of a

nucleon, a quark is forever bound to its two partners; no force can
ever tear them apart. In contrast, in the dense primeval crowd, any
quark is completely free: it can go wherever it wants to; never is
there a vacuum; a quark always finds more than enough of the com­
panions required by quark confinement. It can move over arbitrarily
large distances, always accompanied by new, ever-changing part­
ners. Such a primeval medium is today denoted as "quark plasma,"
and different projects of large-scale experimental research are pres­
ently attempting to produce such a plasma in the laboratory.
But these primeval building blocks possess another crucial prop­
erty. Why does a nucleon consist of three quarks and not of many
more? Why does the size of a nucleus increase with the number of
nucleons it contains? Apparently, it is not possible to accommodate
arbitrarily many particles in a given volume. Each matter particle
seems to insist on having its own spatial volume, no matter how
small-but a finite space just for itself. The sum of all these volumes
with the particles living in them then gives us matter. Indeed, we
have in today's physics such "territorial" particles, which do not allow
another identical particle in its space: electrons, nucleons, and-as
constituents of the nucleons-also the quarks. They all are subject to
the celebrated exclusion principle formulated in 1925 by Wolfgang Pauli:
it is not possible that two completely
identical particles of this kind can exist
at exactly the same point of space.
An immediate consequence of this
principle is that atomic nuclei must
grow in size with increasing atomic
weight, that is, with an increase in the
number of nucleons they contain. A
gold nucleus contains 200 nucleons, a
helium nucleus only four. Since each
nucleon insists on its own space, the
gold nucleus must be correspondingly
larger than that of helium. Wolfgang Pauli (1900-1958)
 The First Particles 21

Hydrogen Helium Lithium

Nuclei growing in size with the number of nucleons

The exclusion principle, like many concepts in physics, is based

on symmetry arguments. One finds that in the underlying quan­
tum theory an interchange of two identical particles at the same
place does not lead to exactly the same state, but instead to its mirror
image. Now, a given state cannot be the same as its mirror image-­
in the mirror, your right arm becomes your left arm! But the inter­
change of two identical particles leaves the world unchanged, and
therefore the presence of identical particles at the same place is not
allowed. Particles of this kind, with such a symmetry property, are
today called fermions, named after the famous Italian physicist Enrico
Fermi, who developed the basic theory for their physics. Fermi was
probably one of the last physicists who made seminal contributions to
experimental as well as to theoretical physics; in the 1940s, he played
an essential role in the development of the first nuclear reactor.
Matter thus consists in principle
of fermions, but they are interacting
with each other. Since Einstein, we
know that there is no instantaneous
interaction at a distance; if two elec­
trons interact, one has to send a signal
to the other, and the transmission of
that signal takes place with a finite
velocity, the speed of light. One elec­
tron sends out a messenger, a light
particle or photon, which reaches the "IO'.'�V 7 <. � ., <

other after a certain time and passes ...._....._..�aaaa&&&aaaa �

on the transmitted information. Enrico Fermi (1901-1954)

22 Before Time began

,.,,..----­

�h
-
-

<lJ ot o n
,
,

E
·.;:;

space

Electron interaction through photon exchange

Thus the second basic form of particles are the force particles
mentioned earlier, which provide the interchange of information
between fermions, and which also act as cement in the construc­
tion of matter. They are not a countable structural part of matter,
and so they don't require any proper space. The number of force
particles in a given volume is not restricted, and if one carries out
an interchange of two such particles, one obtains again the same
state. Particles obeying this form of symmetry are called bosons,
named after Bengali physicist Satyendra Nath Bose, who as a young
researcher in 1927 had determined this essential property of photons
and sent his results to Einstein. Einstein immediately recognized
the importance of Bose's results, translated them into German and
made sure that they were pub­
lished in Germany under Bose's
name. Besides the photons, we
know today the vector bos­
ons of the weak nuclear force
as related force particles; that
force makes possible the radio­
active decay of heavy nuclei,
such as uranium. In the realm
of strong nuclear interactions, Satyendra Nath Bose (1894-1974)
 The First Particles 23

for the interaction of the quarks, the corresponding role is played

by the sluons; they make possible the binding of quarks to protons or
neutrons, to which we return soon.
The basic particle forms for matter and force thus are.

Fermions and bosons.

Fermions are the building blocks of matter; bosons transmit and

announce their interaction. At this point, we want to return briefly
to the idea of having at the beginning just one primeval form of par­
ticle. Is it not possible that fermions and bosons are the descendants
of a single particle species? The two present forms are distinguished
by their different behavior under the interchange of two particles at
the same place: fermions turn into a mirror image, bosons into the
same state. The constituents of a greater, unified theory would have
to allow both, and so the attempts to formulate it are referred to as
supersymmetry, SUSY for short. If such a theory was valid in the very early
universe, then it must have been possible at that time to convert fer­
mions into bosons and vice versa. Subsequently, there must have been
a transition in which this possibility was destroyed. How something
like that can happen we will discuss in more detail in the chapter
on transitions. So far, there are various attempts to construct such a
supersymmetric theory, but they have remained attempts. However,
one result of the considerations up to now is that if the primeval world
was indeed supersymmetric, then there should remain even today,
after the breaking of the symmetry, for each particle a very heavy
"supersymmetric partner." For the electron, there would be a bosonic
S-electron; for the bosonic photon, a fermionic S-photon or photino,
and so on. One has not only given the missing partners names; they
were also searched for intensely in high-energy theory and in high­
energy collisions, and the search goes on. So far with no success.
As mentioned, we have today various species of matter particles,
interacting through different forces; hence there are also various spe­
cies of force particles. But in the early stages, just after the Big Bang,
the world apparently contained only two kinds: primeval fermions
24 Before Time began

with a universal interaction, and primeval bosons, which mediated

that interaction. Both were massless, and there was no way to iden­
tify or distinguish different species of fermions or of bosons. But the
information, which later on led to the different species, must have
already been present in some latent form. The seemingly identical
particles must even at that early time have had properties which then
were unimportant, but which later became crucial. As an example of
such a situation, consider two heavy balls of equal size, weight, and
appearance, one made of stone, the other of iron. As long as we can
only weigh them, use only gravity, we cannot distinguish them. But
when we bring in a magnet, one will be attracted, the other not.
In the course of time, the primeval fermions would give rise to
all the elementary building blocks of the present world, and the
primeval bosons would do the same for the force particles. To see
which hidden properties primeval particles must have had, let us
elaborate a little on what basic properties the ultimate particles can
in fact have.
Besides the requirement of indivisibility and that of its own spa­
tial region, the essential aspect of matter particles is that they should
have some property that one can enumerate or count. In the case of
electricity, we call this property the char3e of the particle. If we con­
sider a system of two electrons, each one has an electric charge -1,
the whole one of -2. It is precisely this charge that allows us to
speak of one or two or three particles. Such charges are immensely
important, because innumerable experiments have shown that
without outside interference, the total charge of a system will not
change, that it is a conserved quantity. If a system has total electric charge
zero, then the addition of energy to the system can lead to the pro­
duction of a pair, a positive and a negative charge, but never to just
one charge. The positive partner of the electron, the positron, is its
antiparticle. When the two meet, they can annihilate each other and
turn into radiation. Besides the electric charge, there must thus also
be some kind of matter charge, which is opposite for electron and
positron, so that the sum leads to zero. This charge is expected to be
conserved as well.
 The First Particles 25

We might note at this point that while the electromagnetic as well

as the weak and the strong nuclear forces lead to discrete charges
and their conservation, there does not seem to exist a fundamen­
tal charge for gravity. This could be an indication that the force of
gravity is indeed of a different nature, as has been proposed recently
by the Dutch theorist Erik Verlinde, arguing that it is emergent, the
result of a collective effect of many particles. This would presum­
ably rule out a theory of everything, in which all the interactions
are unified.
Charges have an immediate and fundamental consequence. The
primeval fermions were created out of the liberated energy of the
false ground state, and the primeval medium had, as far as we know,
no charge of any kind. For this reason, the energy thus provided
should have led to equally many fermions and antifermions. Such
an inherent duality of opposing concepts seems ingrained in many
parts of human thinking; it appears to be completely natural: there
has to be both plus and minus, left and right, good and bad, heaven
and hell, ying and yang. In the beginning, the world was neutral,
and on the whole it should remain that way.
The law of conserved charges seems to have suffered one excep­
tion, which has created trouble for physicists for a long time: why
does our universe consist of matter, containing, as far as we know,
no antimatter? We can of course just assume that it started that way
from the very beginning, but explanations of the kind "it always
was that way" are considered unsatisfactory, unpleasant by most
physicists. So one assumes that the
bubble from which our universe was
created initially contained no parti­ •
cles and thus had zero charge. That
means when the provision of energy
through the transition to the right
ground state allowed various kinds
of fish to emerge from the sea of vir­
tual particles and become real, the
fish always had to appear in pairs, fish Yin and yang
26 Before Time began

and antifish. That would conserve all possible charges. One thus
assumes that before the Big Bang, there was no matter in our pre­
sent sense; afterward, matter and antimatter in equal quantities.
How did the antimatter disappear, where did it go? W hy is the elec­
tric neutrality of our present world achieved in such an asymmet­
ric way, with light electrons and heavy protons? We shall return
to this issue several times, but we can note a warning even at this
stage. The Russian theorist Mikhail Shaposhnikov concluded in a
talk some ten years ago that he knew of exactly 44 explanations for
the asymmetry, and that he would much appreciate hearing about
possible others. That shows how nicely one can say in physics, "We
just don't know."
Before we list the fish created in the Big Bang, we have to engage
in a little particle zoology: what kind of particles exist today? After
what we just said, that means asking what kinds of charges exist in
nature. This question has been for quite some years a central topic of
particle physics, and so we have to see what that has led to. We have
already mentioned electrons, the smallest units of electric charge,
and their antiparticles, positrons, that carry the opposite charge.
The electron was known for quite a long time, from atomic struc­
ture considerations; the positron was predicted in 1928 by Dirac, pre­
cisely on the basis of charge conservation, and it was experimentally
confirmed shortly afterward by Carl Anderson. Both theorist and
experimentalist were awarded the Nobel Prize.
In our present, electrically neutral world, the negative charge of
the electrons is in atoms compensated by the positive charge of the
nuclei. In contrast to electrons, the nuclei are composite objects,
consisting of nucleons-that is, positively charged protons-and
neutral neutrons. Nucleons are, as mentioned, in turn bound states
of still smaller entities, quarks. The underlying scheme for the bind­
ing of quarks is a little more complex, however, again based on the
conservation of charges, but now on charges of the strong nuclear
force. This requires the existence of two kinds of quarks, denoted by
u (up) and d (down) to account for our normal world. For each spe­
cies, there have to be the corresponding antiquarks, out of which one
then constructs antiprotons and antineutrons. These antinucleons
 The First Particles 27

can be produced in the laboratory (so they really exist), but other­
wise they are hardly encountered in our world.
At this point, the attentive reader might ask what this "hardly"
is supposed to mean. In interstellar space, one finds once in a while
single unbound nucleons flying around, lonely cosmic wanderers. If
two sufficiently energetic such nucleons should happen to collide-­
and that happens quite rarely-the collision can result in the pro­
duction of a proton-antiproton pair. The antiproton thus produced
can now fly off on its own, until it has the bad luck of colliding with
another proton. Then the two annihilate each other and turn into
electromagnetic radiation. Such cosmic antiprotons exist-not
many, but they can even be observed in terrestrial detectors.
Quarks and antiquarks don't just interact with each other-to
form nucleons, antinucleons, or mesons (quark-antiquark bound
states)-they also interact with electrons. When such interactions
were studied, a further, somewhat spooky particle was encountered:
the neutrino. It is almost or completely massless. What happened was
that free neutrons were found to decay into a proton and an elec­
tron. However, in such decays the energy of the proton and that
of the electron did not add up to the neutron mass; their sum gave
less. This missing energy had been carried off by the unseen neu­
trino; Pauli invented the process, Fermi the name. The right way to
write neutron decay thus is n � p + e-+ anti v. Electrons and neutri­
nos are collectively denoted as leptons, from the Greek leptos =light.
They carry a unit lepton charge, which takes on the opposite value
for antileptons. That's why the neutron decay produces an antineu­
trino: then the overall lepton charge remains zero.
So the essential species of matter particles are

Quarks and leptons.

In both cases, one has the corresponding antiparticles, antiquarks, and

antileptons. As mentioned, leptons have a lepton number, antileptons
the opposite. Quarks also have an additional charge derived from the
strong nuclear interaction, the baryon number (from barys =heavy in
Greek); the name arose because nucleons are so much heavier than
28 Before Time began

electrons and was invented before the quark infrastructure of nucle­

ons was known. Antiquarks have the opposite baryon number, so that
in the creation of a quark-antiquark pair, both electric charge and
baryon number are conserved. The same holds for leptons and lepton
number, in addition to electric charge. The conservation of baryonic
as well as leptonic charge today seems to be an absolute law. That,
however, cannot have been the case at earlier times: if at the beginning
there was matter and antimatter in equal amounts, the corresponding
symmetry must have been broken at one stage, in order to produce
our universe consisting of nucleons (baryons, not antibaryons) and
electrons (leptons, not antileptons) .

To obtain a common origin of matter particles, we can imagine

that in the very early universe, only one kind of primeval fermion
existed, which then later split up into quarks and leptons. This pri­
meval form would be massless-there would be no measure, no
scale, and only one kind of interaction, apart from gravity. A the­
ory for such a unified world, a "grand unification theory," or GUT
for short, is up to now still more dream than reality, not yet quite
reached by much intensive research. Since in such a theory, quarks
and leptons would simply be different possible states of a primeval
fermion, quark-lepton transitions would now be possible: a quark
could turn into a lepton and vice versa. At such an early time, there
could have been fluctuations leading to more quarks than anti­
quarks or to more leptons than antileptons. An excess of quarks and
leptons at the time when the GUT era suddenly came to an end,
when all bridges between quarks and leptons were broken down­
that would provide an explanation for the observation of the pre­
sent asymmetry between matter and antimatter. The symmetry,
valid at the time of the Big Bang, was broken at the end of the GUT
period. That made possible our world in its present form: a world
containing much matter, but, as far as we know, not a correspond­
ing amount of antimatter.
We have thus found a further important threshold in the evo­
lutionary history of the very early universe. There was a point in
time at which quarks and leptons went their separate ways. From
 The First Particles 29

now on, in each reaction, the total number of quarks (i.e., quarks
minus antiquarks) and the total number of leptons ( leptons minus
antileptons) had to remain unchanged. Whenever a quark-fish
surfaced, it had to be accompanied by an antiquark-fish, and the
same held for leptons. Since our present world contains so much
matter and so little or no antimatter, the particle-antiparticle sym­
metry was broken at that time, and so we can identify the end of
the GUT era as

The birth of matter.

At this point, the transition to the next era led to a few more quarks
than antiquarks, a few more leptons than antileptons. This excess
was not dramatic: estimates indicate that for 30 million antiquarks,
there were 30 million plus one quarks. But this tiny excess now had
to be preserved for all times, and it finally was enough to provide all
the matter found today in our universe. Small causes can have big
effects; how that took place, we shall see soon.
Up to now, all fermions were massless, quarks as well as lep­
tons. Today, both have masses, but at that time they did not. When,
where, and how did these masses appear? What exactly are masses?
On one hand, they specify the resistance to forces, as inertial masses;
on the other hand, they measure the effect of gravity, as wei3hts. Are
masses perhaps something like a charge for the force of gravity? Is
there a smallest mass as the fundamental charge of gravity, like the
electric charge in electromagnetism? Recent considerations seem to
indicate that this not so. Masses are an emer3ent property, not an intrin­
sic one. They were not always there, but rather were dynamically
created, through interaction, in the course of time. How can such a
dynamical mass 3eneration occur?
How can a massless particle suddenly acquire mass? That is a rather
common problem in physics, and so we will first look at it in gen­
eral. The simplest way to gain mass is shown by a snowball: a small
light snowball becomes larger and heavier as we roll it in the snow.
It attracts the flakes of the medium and makes them part of the ball.
30 Before Time began

Something like that can always happen when there is an attrac­

tive interaction; a certain amount of the medium is then converted
into additional mass of the particle. A well-known case of this effect
is found in the polarization of electric charges. A plasma of equally
many positive and negative particles, which are otherwise identical,
eventually reaches an equilibrium in which the opposing charges
just compensate each other-in each local volume, there are as
many positive as negative. If we now introduce into this medium a
strong negative charge, positive charges will feel attracted to it and
compensate the new charge, creating a larger object, which is again
neutral. The overall result is thus the formation of a new, larger and
heavier "particle"-the introduced negative charge with its cloud of
positive charges-and this object, being on the average electrically
neutral, can move freely through the plasma. The polarization of
the plasma charges around the intruder thus has increased its mass.
And this new mass is indeed an inertial mass: if we try to move it, we
have to move both the original intruder and the cloud of the clinging
charges. Polarization thus has really led to a mass increase, and we will
encounter other examples of such a mechanism of mass generation.

•0•0•0•0
8e8e8e8e
e8e8e8e8
8•8•8•8•
•8•8•8•8
8•8•8•8•
•8•8•8•8

Mass generation through charge polarization

In different and more complex versions, such a mechanism is

today quite generally considered as the origin of mass. Particles have
various inherent labels-electric charge, baryon number, spin, and
more. Mass is not such a property, and neither is spatial size. Before
the quark mass was created, was there a force of gravity, however
 The First Particles 31

weak, between two quarks? That question, though somewhat aca­

demic, so far does not really seem to have a final answer.
We have not yet mentioned what the medium is that plays the
role of the snow or the plasma, giving rise to the masses of quarks
and leptons. We shall return to that issue in the next chapter; here
we only note that besides the primeval fermions and bosons, there
must exist some further field which acts as the mass-generating
snow. But even at this point one aspect becomes evident: if we roll
the snowball too fast, it cannot collect further flakes. And the new
intruding strong charge in the plasma also needs some time to col­
lect the cloud of opposite charges. In the same way, the primeval
universe had to cool down enough after the Big Bang, so that the
speed of the primeval fermions was sufficiently reduced for mass
generation. When that was the case, the quarks collected clouds of
the primeval snow to acquire small intrinsic masses, and so did the
leptons; their intrinsic masses were of comparable size, from half to
a few MeV, in the units of particle physics.
To have an overview, let's combine all the mentioned species of
particles to form a genealogical tree, a family tree. As in all such
trees, some members are closer to us than others. Quarks and gluons
combined to form nucleons, and these determine our own mass.
With the help of photons, nucleons and electrons combined to cre­
ate atoms, and those are the basic constituents of the present world.
At this point, we may wonder if there are particles around today that
have been with us ever since the Big Bang. We shall see that that is in
fact quite unlikely. All particle forms have passed through various
stages of development, in which they could have been destroyed and
created again, or in which the combination of then-existing parti­
cles led to new forms.
We should also note that the family tree shown is meant to be
very schematic. Just as in the case of biological species, the develop­
ment with time also led here to more and more different subspecies.
But there is a difference: in the evolution of the early universe, the
appearance of new species occurred through well-defined transi­
tions. In the very beginning, primeval fermions interacted through
32 Before Time began

The family tree of the particle species

primeval bosons; there was only one primeval force. Why then did
these primeval particles split into the multitude of the particle forms
we know today? The only thing that changed was the temperature:
the universe expanded and cooled off more and more. The decreas­
ing temperature must have modified the structure of the world,
just as it turns steam into water or water into ice. In Chapter 4, we
shall consider such transitions in more detail.
At his point, we want to return briefly to a question that has
occupied many physicists in the past few years. If we imagine that
a certain transition turned the primeval fermions into a number of
different quarks and leptons, why and how do they differ? We have
labels to identify the different forms, charges and more, but we have
not shown why there are different forms. Why are there six kinds
of quarks or three kinds of electrons? This question reminds us of
a classical earlier puzzle: why are there so-and-so many kinds of
atoms? Could it not be possible that the basic constituents-quarks,
leptons, and bosons-are just different combinations of something
smaller and more "basic"? That is the aim of the so-called string
 The First Particles 33

theory, which has been studied in different forms over the past 30 or
40 years. It assumes that all our different particle species are just dif­
ferent vibrational excitations of tiny primordial strings. Such a pic­
ture is a continuation of the familiar reductionist approach, which
so far has been very successful. It seems indeed tempting to identify
six different quark species and six different lepton species as twelve
different vibration forms of a single elementary string. That explan­
atory aspect, together with the fact that string theory is based on
a challenging new mathematical formalism, is the reason why so
many theorists have devoted their efforts to this approach. Unfor­
tunately, however, these efforts have not yet led to a final and truly
satisfactory answer. Present versions require a spacetime structure of
more than the familiar four dimensions. The most popular form is
based on eleven dimensions, of which seven appear much "smaller,"
just as the Earth seen from an airplane appears flat, since the heights
of trees and buildings is so much less than the visible spatial exten­
sion. It is not so obvious, however, that a reduction in the number
of particle species at the cost of an increase in the number of space­
time dimensions really is a genuine simplification. Moreover, such
schemes predict, as we mentioned in the context of supersymme­
try, the existence of further particle species so far not observed. The
observation of such supersymmetric partners in future accelerator
experiments would of course provide an immense boost to string
theory as well.
We now have to return to two features so far left dangling: what
is the snow we need to give our massless primeval fermions some
mass, and where does the whole process take place? Our picture of
the world so far lacks something that today is a crucial part of our
universe: the so-called empty space, the stage for the whole play.
The transition from the false to the correct ground state had liber­
ated so much energy that the result was an extremely hot and dense
medium of primeval particles. Everywhere there was something­
the dominant, almost empty interstellar space did not yet exist. Let
us now see how it came into existence.
 3

Empty Space

In reality there exist only atoms in empty space.

DEMOCRITUS, 400 BC

Empty space is only an idealization, for various reasons; it is never

truly empty. We are here not referring to the virtual particles sub­
merged below the surface of the vacuum and waiting for the energy
necessary to let them come up. They do not yet exist, and thus leave
space indeed empty. But the empty space of our universe does con­
tain the remaining dark space energy causing it to continue its expan­
sion; beyond this, it does not create any structures or forms. This dark
energy is, moreover, of extremely low density, much too low for any
kind of particle creation, and it is distributed at constant density over
the entire space. The counterpart of its density would correspond to
about four nucleon masses smeared over one cubic meter-that's as
empty as it can get. In addition, there is the cosmic background radia­
tion, which also permeates all of space, though at a yet much lower
and steadily decreasing energy density, decreasing because of the
expansion of the universe. And the resulting "empty space" finally
still contains us: it contains matter, which, however, is very unevenly
distributed. If we average over the entire visible universe, we end up
with a density of about one nucleon per cubic meter, less than that
of the dark energy. Moreover, since matter is concentrated largely in
stars and other heavenly bodies, the interstellar space is indeed very
empty, as far as matter is concerned. In any case, it is the most empty
space we can find in our universe, the physical vacuum.
At the start, however, in the very early universe, there was no
form of empty space; the density of particles was immense; there
36 Before Time began

was something everywhere. Even a spatial volume of a present

nucleon, such as a proton, contained an immense number of quarks
and antiquarks. There was no physical vacuum-even empty space
first had to be created. The possibility for such a creation is due to
the nature of the strong nuclear force, which binds quarks to form
nucleons. Electrons are "real" particles: they can be isolated; it is pos­
sible to keep one single electron in a cubic meter of vacuum. With
quarks, that is not possible: they are fundamental constituents and as
such behave like magnetic poles. There is a plus pole and a minus
pole, but these cannot be separated.
Quarks behave in a similar way. There are nucleons, as we have
seen, in which three quarks are coupled together tightly. In addition,
there are quark-antiquark pairs-in today's nomenclature, they are
called mesons, strongly interacting particles that can be produced in
high-energy proton-proton collisions. The symmetry determining
the coupling of quarks to form nucleons or mesons is somewhat
more complex than the plus-minus form of electromagnetism. The
charge responsible for strong interactions can take on three rather
than two different values, which are usually called color, such as red,
blue, and green. Nucleons and mesons then arise through a super­
position of such colors.
The fundamental theory of strong nuclear interactions, quantum
chromodynamics, insists that only colorless objects can have an inde­
pendent existence. Such color-neutral objects can be created either
by combining a red quark with an anti-red antiquark (and simi­
larly for blue and green) or by superimposing the three basic col­
ors, which again leads to a colorless entity. An even simpler form
of colorless state is a very dense plasma containing equal numbers
of quarks of the three colors, as well as of quarks and antiquarks, all
evenly distributed throughout. And that, we believe, was indeed the
state of the very early universe: quarks and antiquarks everywhere,
nowhere a spot of empty space.
The smallest possible color-neutral state then is either a quark and
an antiquark, or three quarks, combined in a small spatial region so as
to make that region colorless. In other words, quarks can exist in our
 Empty Space 37

present world, in empty space, only as quark-antiquark pairs of the

same color/anticolor (mesons), or as three quark states whose colors
are neutralized by superposition (nucleons). These are the smallest
strongly interacting entities allowed to have an independent exist­
ence. They are collectively referred to as hadrons, from the Greek hadros
(thick), to distinguish them from the very much lighter electrons,
belonging to the set called leptons, from leptos (light). The interaction
between quarks and between quarks and antiquarks is mediated by
the force particles of the strong interaction, called sluons, since they
"glue" quark constituents together to form hadrons. These gluons
also have to be colored, so that they can bind different colors to form
nucleons. Thus a red-green gluon is needed to couple a red quark to
a green one, and so on, resulting finally in colorless hadrons.

. . . .
. . . . .
.
.

.
: .
: .
. .

. .
. . .
· . .
. .
.
. .
. . .
. .
. .
. . . . "
. .
. .
. .
. .
.
. . .
. . .
. . .
.
. .
. . .

Quark states of existence: quark plasma, meson, nucleon

Quarks can only exist, as we have noted, if there are other quarks
or antiquarks in the immediate vicinity, so that a local colorless
combination is possible. Hadrons generally have universal size of
15
about one femtometer (1fm=10- m); from this, we can conclude
that this is the largest possible separation between quark compan­
ions. As Lucretius said, no force can separate them further. That also
tells us what the lowest possible density of the quark plasma is: we
have to have at least one quark per cubic femtometer. What happens
if the density falls below that value? Then a new state of existence is
created: our present world.
The early universe found itself in a state of continuous expansion
and continuous cooling. The number of quarks could not increase,
and so their density was necessarily decreasing. Sooner or later, a
38 Before Time began

breaking point had to come: the quark density fell below that criti­
cal value of one per cubic femtometer; an unbound quark existence
was no longer allowed. The overall situation now became a kind of
cosmic "musical chairs": at the critical point, each quark had to grab
an antiquark, or two other quarks, in order to survive as part of a
colorless entity. And the hadrons thus created were now truly freit-­
between two such hadrons there could be any amount of empty
space; so this transition was indeed

The birth of the vacuum.

This birth also defined for the first time in our universe a measurable
scale: the critical density at which a quark plasma becomes a gas of
independent hadrons. This scale is defined by the limit of one fem­
tometer (lO�ts m), the largest possible separation between quarks. A
larger separation is not possible; now hadrons appear, and between
them there is what we call empty space. Hadron formation is thus
also the first occurrence of structure formation in the universe: there are
regions of empty space, and in it, from time to time, a bound state
of three quarks. This type of process later was repeated a number of
times: when nucleons coupled to form nuclei, when nuclei coupled
with electrons to form atoms, when clusters of atoms led to the for­
mation of stars, and when these eventually formed galaxies. In each
case, structures appeared in empty space as a stage. How this can
happen, we will address in detail in Chapter 6.
But the transition from quarks to hadrons has further conse­
quences. At high density, the medium consisted of quarks and
antiquarks, with gluons as force particles. The quarks had by then
acquired a very small but still finite mass through interaction with
the primeval snow given by some remaining underlying field (the
dark energy or, as we shall see, more likely the Higgs field) perme­
ating the entire universe--that was the mentioned "first" mass
generation. The gluons remain massless and form for the quarks a
much stronger background field, but one that exists only within the
interaction range of the quarks. The same process of dynamical mass
 Empty Space 39

generation is now repeated-when the quarks move slowly enough

through the gluon field, the gluons cling to them and give them a
new, much larger mass. The gluons thus have to fulfill two func­
tions: they give the quarks a rather large effective mass, and they
mediate the interaction between these quarks, eventually binding
them to create hadrons.
The quark mass created by the primeval space field was just a few
MeV; the new mass, due to the cloud of the gluon field, is much
larger, around 300 MeV, about a third of the proton mass. It is these,
now much more massive quarks that combine to form hadrons. The
nucleon mass thus lies around 900 MeV; the typical meson mass is
around 600 MeV. In other words, the quark mass determined by the
gluons as well the nucleon mass formed in turn by the heavy quarks
are again emergent quantities, created by interaction; they are not
inherent parameters of the system. And it is these masses we have
in mind when we talk about masses and weights. The 80 kilos of a
human due not arise from any primeval quark masses; they are due
to the effective quark masses formed by the gluon field and by the
nucleon masses resulting from them.

• •

m=3 MeV m=300 MeV

(a) (b) (c)

Evolution of the quark mass: (a) Primordial (b) Primeval field (c) Gluon field

In principle, the decreasing energy density of the medium could

thus lead to two steps of evolution: first, the polarization of the gluon
field results in new, larger quark masses; then, these massive quarks
couple to form hadrons. All present calculations indicate, however,
that these two phenomena occurred simultaneously in the early
universe--it is not really well understood why that was the case. But
40 Before Time began

as a result, we have a hadronization transition in which the quarks

get their larger mass and are simultaneously bound to form had­
rons. The plasma of colored quarks thus became a hadron gas: the
hadron era began. In contrast to the colored quarks, which could never
be further apart than one femtometer, there could now be arbitrar­
ily much empty space between hadrons: the physical vacuum.
The birth of the vacuum was undoubtedly one of the four crucial
events in the formation of our universe.

• The first was the fluctuation of the primordial world, leading

to the exploding bubble subsequently becoming our universe.
• The second was the transition from the false to the true ground
state, which liberated the energy needed to form the contents
of our universe.
• The third was the breaking of the matter-antimatter sym­
metry, turning a world of equally many matter and antimatter
particles into one in which matter dominates.
• And the fourth was the binding of massive colored quarks into
color-neutral hadrons, making possible the existence of empty
space.

The last step, the creation of effective gluonic masses for the quarks
and their simultaneous binding to colorless states through

Hadrosynthesis,

led to what today is the major part of our world: space without any
kind of matter. That is why we identified it earlier as the birth of the
vacuum. But we have to be careful what we say-because space is
continuously expanding, even at an increasing rate, it is clear that
"empty" does not mean "nothing." There is, as noted several times,
the dark energy inherited from the primordial world, out of which
ours was formed. But this mysterious medium is in a sense a prop­
erty of empty space--in any case, it is not something we would con­
sider as matter.
 Empty Space 41

What then was the form of the matter in our universe, just before
and just after hadronization? The strongly interacting constituents,
quarks and gluons, were dominant at first, making up some 80% of
the total energy. It therefore makes sense to call the time before
hadronization the quark era.
In the hadronization transition, the gluon field contribution
turned into the effective quark masses and into the binding of the
hadrons; we now had a mixture of hadron gas and lepton gas in
empty space. At first, the hadron gas dominated, but hadrons and
antihadrons could annihilate each other, leading to electromagnetic
radiation. As long as the gas was hot enough, the inverse process
brought new hadrons back into the system. With decreasing tem­
perature, the process became more and more one-way, hadrons to
radiation.
The hadron era ended as soon as the temperature was no longer
sufficient to create hadron-antihadron pairs. Somewhat later, the
leptons suffered the same fate, as we shall see shortly. And these
annihilation reactions indicate a great potential danger: if the
universe had contained an identical number of hadrons and anti­
hadrons, in particular as many nucleons as antinucleons, the anni­
hilation would have been complete, and today's universe would
have contained no atoms, no matter in our present form. Our ter­
restrial world would never have appeared.
The tiny excess of quarks remaining at the end of the GUT era­
one quark more among thirty million quarks and antiquarks-in
the long run formed the whole basis for our world. The thirty mil­
lion quark-antiquark pairs first formed hadrons and antihadrons,
such as protons and antiprotons. These then annihilated each other
and turned into radiation. But that one excess quark had no one to
annihilate, so it had to find two further survivors to form a color­
neutral nucleon, which now could irrevocably exist in empty space.
The nucleons formed in this way were a tiny minority in space,
which contained mainly electrons, positrons, neutrinos, and pho­
tons. The hadron era had ended; the lepton era started.
42 Before Time began

As we have already mentioned, the leptons shared the same fate

only a little later on: electron-positron pairs annihilated each other,
and only the minor excess of negative electrons remained from the
end of the GUT transition, when it had compensated the equally
small excess of positive quarks and thus kept the overall charge zero.
The space of the universe now contained mainly radiation, photons
and neutrinos, besides the few surviving nucleons and electrons. At
this time, there were about a billion photons for each nucleon, so
one could now rightly speak of the radiation era. And this ratio has
remained roughly the same until now.
But in spite of its numerical majority, radiation no longer forms
the main part of the universe. Space continued to expand, which
increased the wavelength of the photons and thus lowered their
energy, while the masses of the nucleons and electrons remained
constant. The energy density of the photon gas thus decreased more
and more, falling after some 50,000 years below that of the nucle­
ons. The radiation era was now over; the domination of matter
began. Today the energy density of nucleonic matter is more than a
thousand times greater than that of the photon gas. In the follow­
ing picture, we summarize the epochs of the universe mentioned;
the size of the boxes does not, however, indicate the time extensions
involved. The quark era was over after some 10-5 seconds, the lepton
era after a further 10 seconds; the radiation era, as mentioned, per­
sisted for about 50,000 years. And the matter era will be addressed in
more detail shortly, since its fate involves us almost directly.

GUT Quark Hadron Lepton Radiation Matter

Eras of the universe after the Big Bang

Let us return once more to the beginning of the radiation era. The
content of the universe now consisted largely of photons and neu­
trinos, and in addition a few electrons and nucleons. The latter were
protons and neutrons, with as many protons as electrons, to keep the
9
overall charge zero. The temperature was now around 10 degrees
 Empty Space 43

Kelvin; at this value, collisions with photons could still prevent the
binding of protons and neutrons to form nuclei-any bound states
were immediately destroyed again. Such break-up processes came
to an end when the universe was more than ten seconds old. The
energy of the photons was now no longer sufficient for break up, and
nucleon binding became possible; in other words,

Nucleosynthesis

began, the binding of nucleons to form nuclei. At first, a proton

and a neutron combined to form deuterium, then two deuterium
nuclei formed a helium nucleus, and so on. Such fusion processes
always involve two opposing effects: two nucleons attract each other
through the strong nuclear force, but only at very short distances,
while two protons repel each other electromagnetically, since they
have the same charge. So to form a nucleus, the kinetic energy of
the protons must be large enough to overcome the repulsion, and
the overall density must be high enough to bring the nucleons close
enough together. In other words, nuclear fusion can only take place
in sufficiently hot and dense media. For this reason, all attempts to
build nuclear fusion reactors have so far not succeeded-it has not
been possible to create the necessary conditions in the laboratory. In
the hydrogen bomb, nuclear fusion does become possible, because
a prior atomic bomb explosion as a fuse provides the needed heat
and density. In the early universe, the necessary conditions prevailed
long enough to form at least helium. Before any abundant forma­
tion of much heavier nuclei, the world had already expanded and
cooled off too much, so that after some minutes, the nucleons could
no longer reach each other.
An essential feature of nuclear fusion, essential in particular for
us here on Earth, is that the process liberates energy in the form of
photons-electromagnetic radiation, sunshine. When two protons
and two neutrons combine to form a helium nucleus, the mass of
the nucleus is less than that of the sum of its constituents. The for­
mation of a nucleus is therefore energetically more favorable than
44 Before Time began

the persistence of the individual nucleons. That defines the tem­

poral beginning of nucleosynthesis: it began, as mentioned, when the
energy of the photons was no longer sufficient to break up the newly
formed nucleus. Nuclei could now survive. And it ended when tem­
perature and density had decreased enough: even when two protons
now met, their electric repulsion would prevent nuclear binding.
Thus, there was only a small time window for primeval nucleo­
synthesis, nuclear fusion in the early universe; after a few minutes,
the conditions were no longer suitable; and until then, essentially
only deuterium and helium were produced. Because the require­
ments for nuclear fusion are known today, we can estimate what
fraction of the nucleons were at that time bound to form nuclei: we
expect about 25% helium, the rest remains as free protons, that is,
hydrogen nuclei, and a very small amount of deuterium. It is found
that those are indeed the relative abundances of nuclei observed in
interstellar space; this provides one of the three pillars for the Big
Bang theory.
The photons, which now made up a major part of the constitu­
ents of the young universe, thus arose from three sources. The first
appeared when quarks and leptons went on separate ways, when the
electromagnetic force split off from strong and weak nuclear forces.
At this time, the bosons of the weak interaction acquired their huge
masses, leaving the massless photons on their own. A further frac­
tion was created in the annihilation of particles and antiparticles,
hadrons as well as leptons-in both cases, the result was electro­
magnetic radiation. And a final fraction was due to the photons
emitted in nucleosynthesis, the binding of nucleons to form nuclei.
These latter are for us of particular interest. The process of their
creation, which first occurred as primeval nucleosynthesis in the
chronology of the early universe, forms today the basis of our exist­
ence: nuclear fusion taking place in the interior of the sun sends
photons down to our Earth and thus provides the light and the heat
needed for our lives. What first happened in the early universe is
today recurring in all the stars, including our sun: two protons and
 Empty Space 45

two neutrons combine to form a helium nucleus, whose mass is 6%

less than the sum of the nucleon masses. This mass difference is con­
verted into radiation, which becomes our sun light.
We were thus really lucky a couple of times. At the end of the
GUT era, the specific fluctuation of our universe produced a few
more quarks than antiquarks, together with a few more electrons
than positrons. Otherwise, no matter could have survived in the
further evolution-it would all have been converted into radi­
ation. The next time, the expansion of our universe was just slow
enough to provide the time window in which nucleons could bind
to nuclei. A part of the nucleons were now united into nuclei, deu­
terium (2), helium (4), and very rarely some larger ones like lith­
ium (7) and beryllium (9); some 75% remained as free protons, the
future hydrogen nuclei. The mass anisotropy created by the heav­
ier nuclei and their subsequent clustering later on allowed gravita­
tion to form stars.
When nucleosynthesis ended, all the photons were still in con­
tinuous interaction with the charged constituents of the remaining
medium, the electrons, protons, and nuclei. The universe was a hot,
electromagnetic plasma. But in contrast to the quark plasma of still
earlier times, there now existed empty space; there were regions
containing "nothing." Nevertheless, a photon could not travel very
far before it was scattered or absorbed by some charge. The electro­
magnetic plasma was completely opaque to light; our view back is
stopped by this plasma just as our view into the sky is stopped by a
layer of clouds.
The turning point came a little later, with a cosmological "lit­
tle while": after about 380,000 years, the continuing expansion of
the universe had lowered the temperature sufficiently to prevent
photons from breaking up the atoms appearing through the bind­
ing of electrons with protons or with nuclei. The photon energy
had decreased as their wavelengths increased, and they were no
longer powerful enough to dissociate the newly formed constitu­
ents, the
46 Before Time began

Atoms.

The plasma of charged electrons and nuclei now turned into a gas
of electrically neutral atoms. So as far as matter was concerned,
there now were really only atoms, as Democritus has insisted. But
besides such matter, there was lots of radiation, the remaining light
of the Big Bang, to which we will return in a separate chapter. But
first, we have to look in a little more detail at the different kinds of
transitions.
 4

Transitions

Jose Sabra! de Almada Negreiros

Crossing the River Lima, Tapestry
Pousada de Santa Luzia, Viana do Castelo, Portugal

Commanded by Decius Junius Brutus, the Roman legions arrived in the year 135 AC
on the left bank of the river Lima. Because of the beauty of the region, they thought
to be on the banks of the legendary river Lethe, which destroys the memory of whoever
crosses it. The soldiers therefore refused to go on. CarryinB the standard of the Roman
Eagles in his hand, the commander rode across the river to the other hank and from
there called each soldier by his name, thus provinB that this river was not the river
of oblivion.

Transitions occur at the borders between different states of exist­

ence. You can change the temperature of water over quite a range,
and nothing specific happens: water stays water. But then, suddenly,
between -0.1 and + 0.1 degrees centigrade, the homogeneous,
48 Before Time began

isotropic liquid becomes a solid, a crystal of ice with well-defined

crystal axes and with completely different properties than those of
water. Something similar happens between 99.5 and 100.5 degrees:
now the smoothly flowing liquid becomes steam, in which the indi­
vidual molecules have little or no communication, and which again
has totally different properties than water.
In such transitions, the memory of the previous state is indeed
completely eradicated. The water arising from melting ice has no
information about its previous state, and no measurement can tell
if the steam we observe was half an hour ago still water. The tempo­
ral evolution of our universe must have followed a similar pattern,
with transitions between the different eras. We remember the prob­
lem of the vanishing antimatter: shortly after the Big Bang, matter
and antimatter were still present in equal amounts, and then sud­
denly the antimatter was essentially gone. What kind of transition
made that possible?
Before we address the transitions that took place in the early
universe and led us to its present state, we want to consider in a lit­
tle more detail the subject of transitions as such, as they are studied
and described in physics. There are quite different forms of transi­
tion from one state of matter to another, and their description is
a very interesting and rather novel area of physics. It's only fifty
years ago that the American theorist Kenneth Wilson was awarded
the Nobel Prize in physics for developing the foundations of the
field, and a considerable part of its further progress became possible
in the last decades through pioneering work carried out on huge
supercomputers, that is, at large-scale high performance comput­
ing facilities. Let us look at some familiar concrete examples of
transitions.
If with decreasing temperature precipitation turns from rain into
snow, there are no sharp borders. Eventually a few flakes of snow
begin to appear in what started as pure rain, and with time their
fraction increases in comparison to that of rain, until finally it is just
snowing. The constituents of the medium were transformed from
rain drops to snowflakes. The transformation occurs with decreas­
ing temperature, but continuously, over a whole range of values-in
 Transitions 49

spite of the fact that the meteorological experts in Scandinavia are

quite capable of distinguishing "snow-blended rain" from "rain­
blended snow." But in this case we never have an abrupt confronta­
tion of two different states, and so we prefer call transitions of this
kind transformations: rain is continuously transformed into snow, and
vice versa.
The evaporation of water, on the other hand, is something quite
different. Up to 100 degrees the system is in a liquid state--we ignore
the few escaping steam bubbles. If we now continue to add heat (and
also ignore the quite unusual case of superheating), the tempera­
ture does not increase right away, but instead more and more of the
liquid is turned into steam, until all the water has evaporated. The
amount of heat required for this process, the "evaporation heat," is
a well-defined quantity, and only after it has been delivered to the
system does the temperature increase again. Until then we have a
mixed state of water and steam; adding heat changes their relative
fractions, brings the system at constant temperature from "only liq­
uid" to "part liquid, part steam" to "only steam." The situation is
quite similar for melting: the temperature remains constant at zero
degrees, until the sufficient amount of "melting heat" has been sup­
plied to turn all the ice into water.
A third transition form appears when metals become magnetic,
and this form has had a decisive influence on our understanding of
transition phenomena. Iron consists of atoms having their own spin;
they seem to have an axis around which they rotate. If a material is
magnetic, if it forms a magnet, it means that on average the spins of
all the atoms point in a certain direction. At sufficiently high temper­
ature, metals are not magnetic-the axes of its atoms are randomly
oriented, so that an average over the directions of all atoms leads to
zero-there is no preferred direction. For each atom, the spins of its
neighbors point in random directions compared to its own spin ori­
entation. If we now lower the temperature, clusters of atoms start to
form, first small ones and then ever-larger ones, three-dimensional
"islands," in which all spins point in the same direction. The overall
spin directions of the different islands are not coordinated; they are
again random. But if we now lower the temperature still further,
50 Before Time began

something almost miraculous happens. At a certain point, at the

Curie temperature, an island appears whose dimensions are as large as
the whole system and in which all spins point in the same direction.
There still exist other, smaller islands with randomly oriented spins,
but with decreasing temperature they all join their big brother in
their spin orientation. The final state is reached at zero temperature:
now all spins point in the same direction. The arbitrary piece of iron
has become a perfect magnet.

The process of magnetization

This transition occurs, as mentioned, at a certain well-defined

temperature, bearing the name of the French physicist Pierre
Curie. If we measure the average spin orientation of the atoms
below this temperature, it is no longer zero; the overall spin points
in some arbitrary but well-defined direction. There is a polar axis,
from south to north, for the piece of iron as a whole. Pierre Curie
started a famous, albeit tragic Nobel dynasty. He and his wife Marie
Sklodowska were awarded the Nobel Prize in physics in the year
1903; it was the third ever given. Marie later on also received the
Nobel Prize in chemistry in 1911; Pierre was no longer alive then,
having died in a traffic accident in 1906, caused by a horse-drawn
carriage. Their daughter Irene married one Frederic Joliot, and this
couple was awarded the Nobel Prize in physics in 1935. Apart from
Pierre, all members of the Curie family became victims of leukemia,
caused by their work with the radioactive substances they had dis­
covered in their pioneering research.
 Transitions 51

Marie, Irene and Pierre Curie

As we saw before this little detour, there are essentially three

kinds of transition. The first, a transformation such as rain turning
into snow, appears under numerous conditions and was also quite
essential in the evolution of the universe. Nevertheless, it has not
attracted the same interest in physics as the other two, because it
happens continuously, as the relevant parameters-temperature,
pressure, humidity, and more--pass through a range of values.
Nothing happens abruptly.
Things look very different in the case of magnetization. This
can best be illustrated by using a simplified model. Consider a reg­
ular grid, a lattice on whose intersection points there are spins of
unit length pointing either up (i) or down (.J,): s= +1 or s= -1.
Each spin is connected only to its nearest neighbor, and if we flip
all spins simultaneously, the interaction remains unchanged. This
model was studied by the German physicist Ernst Ising in his 1924
doctoral thesis, and it remains forever the Ising model, even though
his thesis advisor Wilhelm Lenz had the idea and the actual solution
was provided only in 1944 by the Norwegian-American physicist
Lars Onsager. In any case, the Ising model today still constitutes one
of the most important models in physics, one that has decisively
formed fundamental concepts in wide areas of research.
At high temperatures, equally many spins point up and down,
and both orientations are randomly distributed. With decreasing
52 Before Time began

temperatures, the "islands" of parallel spins mentioned earlier start

to form; there are some in which all spins point up, some in which
all point down. The average over all spins remains zero. At the Curie
point, however, at the temperature TC' the system must decide
spontaneously what to do: from now on, either up or down wins;
both have the same chance. From now on, for all temperatures
below the Curie point, the average over all spins is no longer zero;
the majority points either up or down. In the following picture, we
have illustrated the situation in the two-dimensional case.

The Ising model: (a) above Tc, (b) just below Tc, and (c) at T = 0

This process is referred to as spontaneous maBnetization, and it is clearly

an emerBent phenomenon: the individual spin is unimportant, and also
its interaction with its neighbors alone cannot determine what hap­
pens. The effect sets in as a concerted action of all spins, and there is
no specific trigger, no general giving a command. The observation of
such a phenomenon is not quite as new as it may seem at first. In the
biblical proverbs of Salomon (Proverbs 30:27) it is said

The locusts have no kinfj, yet all of them fiO .forth in ranks.

This is perhaps the first mention of something known today as

swarm intelligence; it has become a new field of research in biology,
and it plays a decisive role in the behavior of bird and fish swarms.
In the inanimate world, it is in fact a basic phenomenon, as we
just saw.
Another impressive example of the sudden change of behavior
is considered under the heading of percolation. The simplest form is
employed in the Asian game of Go. One randomly places stones
on the cross points of a checker board. These first form small and
then ever-larger islands. But suddenly, when some x-th stone is
placed, the islands extend from one side to the other, and instead
 Transitions 53

of randomly placed islands in a lake we now have randomly placed

lakes in a connected land. Before the last stone was placed, there was
no connection between the shores: this last stone suddenly created
it. Such a transition is referred to as percolation-it is responsible for
the fact that in a coffee percolator, the coffee all of a sudden begins
to flow as we add more and more water. In the same way, the transi­
tion from the dominance of water to that of land on our checker
board takes place very abruptly.

. ..
• .. "'
..
.. .

"' ...
... . ..

..

Percolation

In a two-dimensional world, the state is always "either or": either

a sea with islands, or a land with lakes. Once the percolation form­
ing land has taken place, it is no longer possible to reach all points
on shore by ship. That is the case when less than half of the surface
is covered by sea. The surface of our Earth, however, is about 70%
made up of oceans, and so one can in fact get by ship from any shore
to any other, as was shown five hundred years ago by the Portuguese
seafarer Fernando Magellan.
For three space dimensions, that changes: the "either or" now
becomes an "as well as." The holes in a Swiss cheese are bubbles,
but we can imagine some worm eating its way from one side to the
other, so that now there are "air channels" through the cheese, even
though there still remains a cheese structure as well. A fence divides
an area into two disjoint parts, but a tunnel does not do that to a
mountain. We now quite naturally arrive at a mixed state, such as
we had in the case of evaporating water. In the beginning there is the
54 Before Time began

rock; then we drill tunnels through it, and continue drilling, until
the rock collapses and there is no more rock structure, just stones
and gravel.
But let's return to our spin problem. At the onset of magnetiza­
tion, that state is not changing gradually with changing conditions.
The average spin value above the Curie temperature is zero, and it is
still zero at the Curie point; but from there on it is finite, no longer
zero. This kind of behavior mathematicians call singular or nonanalytic.
It always occurs when only a clear yes-no decision is possible. One
cannot be a little dead, or a little pregnant, and so a system also
cannot be a little magnetic; in percolation, the system cannot be a
little interconnected. At a certain point, there is a crucial, qualitative
change. In physics, such singular phenomena in collective systems
are referred to as

Critical behavior.

Up to the critical point, measured observables were zero, and then

no longer zero-or vice versa. In any case, the behavior of the sys­
tem there changes qualitatively, fundamentally, abruptly, and not
just a little or gradually.
At the critical point and in its immediate vicinity, for values of
temperature and pressure close to the critical ones, the system
can no longer be divided into smaller representative subsystems.
Sufficiently far above or below the critical point, it is generally
possible to study smaller parts of the whole in the expectation that
the rest will behave the same way. In the critical region, that does not
work anymore. In the case of spin systems, islands of all sizes appear,
from two spins to clusters reaching from one side of the system to
the other. For arbitrarily large systems, that means at the critical
point the relevant scale, the correlation range, becomes infinite, it
diverges. That is perhaps also the reason why such systems are so dif­
ficult to treat with our usual mathematics. Lars Onsager's solution
of the Ising model was for the two-dimensional case; in two dimen­
sions, the percolation problem is also solvable. In three or more
 Transitions 55

dimensions, however, neither has so far been solved by analytical

mathematics, in spite of many attempts. What we know about these
cases comes from numerical studies on large-scale supercomputers.
Another very useful way of studying such transitions is based on
symmetry. When iron is not yet magnetic, in its "paramagnetic"
state at sufficiently high temperatures, the spins of the atoms are
randomly oriented; the average overall spin value m is zero. In the
Ising model, there are as many spins +1 as there are -1, randomly
distributed. This result is not changed if we flip every spin orienta­
tion into its opposite, +1 to - 1 South Pole to North Pole. That means
,

the interaction connecting the spins with each other is not changed
by such flipping; it cannot distinguish a state from its flipped form.
Below the Curie point, however, at lower temperatures, there now
is a preferred direction: there are more +1 spins than -1, or vice
versa; m is no longer zero. The interaction is not changed at all but
the state of the system has: flipping now changes the state into its
opposite.
This situation is of great relevance for physics in general and in
particular for the understanding of the evolution of our universe.
If an interaction is unchanged by a certain operation, such as by
flipping all spins, we call it symmetric or invariant under this operation.
This does not mean, however, that the actual state of the system
remains unaffected, as we have just seen. For this state, there are
two possibilities: either it is also invariant, or it is in one of two
asymmetric opposite states, either +1 or -1, and both of these are
equally likely. This means that at the Curie point the symmetry is
spontaneously broken, leading to +1 or - 1 , but because of the sym­
metry of the interaction, we cannot predict which of the two it will
be. In some sense, this is like playing roulette: as long as the ball is
rolling, it is with equal probability on a red or a black field. Once it
stops, however, it has to decide, and in an honest game, the chances
for either red or black are equal.
For a compact description of the situation, one defines the aver­
age mover all spins as the order parameter. Above the Curie point, there
is complete disorder; as many spins point up as down; the measure
56 Before Time began

++tt tt++ tttt ++++

I>- I>-
+t +t Flip t+t+ tttt Flip
++++
t+t+ +t +t tttt ++++
m=O m=O m=+1 m=-1

Flipping all spins above (left) or below (right) the Curie temperature

of order is zero: m = 0. If we now flip all spins, it remains zero. Below

the Curie temperature, however, the order parameter is no longer
zero; there is some kind of order. The average over all spins defines
a specific direction, and when the temperature is zero, all the spins
point in the same direction. Since the interaction does not depend
on the spin direction, that can be up as well as down, but the system
has to make a choice. If we now flip all the spins, + 1 becomes -1 and
vice versa: the actual state of the system is not invariant under flip­
ping, even though the interaction is. One calls the process occurring
at the Curie point spontaneous symmetry breakin&. It occurs even though
no one or nothing intervenes. There is no change whatsoever in
the interaction, but nevertheless the system suddenly falls into an
asymmetric state. This self-induced character is what distinguishes
the phenomenon from explicit symmetry breakin&, which occurs, for
example, if we break off one point of a six-cornered star.
We have so far not paid much attention to the form of transi­
tion encountered when water evaporates or ice melts. But in the
discussion of percolation, we saw how such a process can occur: at
the transition point, both states persist, water and steam, or water
and ice. There we can, as an example, consider water as ordered
and steam as disordered, using the density of the medium as order
parameter. In this case, the system passes gradually, at fixed temperature,
from an ordered to a disordered state. While the order parameter for
magnetization decreases with temperature continuously to zero, it
drops in the case of evaporation from a finite value to zero at a fixed
temperature. Below the evaporation temperature, it is different
from zero, and it remains finite as further heat is added. The input
 Transitions 57

of energy increases the fraction of steam relative to that of water and

thus reduces the value of the order parameter, which finally van­
ishes when everything has become steam. In contrast, for magneti­
zation the order parameter vanishes smoothly as we approach the
critical point from below, and so one calls such transitions continuous,
whereas evaporation and melting are discontinuous transitions. In the
following picture, we illustrate the behavior of the order parameter
in the two cases. Herem(T) is average value of the spin as function
of the temperature, while d(T) is the density of the medium minus
the steam density.

m(T) d(T)

ferromagnetic -- all water

coexistence
paramagnetic ..-- all steam
0 T T

Order parameter for the continuous magnetization transition (left) and for
the discontinuous evaporation transition (right)

With that, we have a summary of the different forms of transition

from one state of matter to another. It can happen gradually­
transformations without any specific indicative parameter values; we
considered the change from rain to snow as an example. Transitions
can occur at a fixed temperature, at which two different states coexist;
further energy input then shifts an ever-larger fraction from the old
state to the new; here we looked at the evaporation of water as a
typical case. And, finally, the transition can occur unambiguously,
at a precise parameter value, as seen at the onset of magnetization:
at the Curie temperature, the whole piece of iron, paramagnetic up
to that point, from now on is a ferromagnet: the metal suddenly
becomes magnetic.
What does all that have to do with the evolution of the universe?
We have seen that the early universe was hot and dense, and that it
58 Before Time began

subsequently cooled off and became more dilute. That originally it

contained equally many particles and antiparticles, but at the end
only particles. That originally all particles were massless, but some
suddenly acquired masses. That quarks combined to form nucleons,
nucleons and electrons to form atoms. So the chronology of the
universe contains quite a few different transitions, and we now want
to look at some of them in more detail.
First of all, we recall that at a transition, different physical pro­
cesses can occur. The symmetry form of the state can change, from
m = 0 (symmetric) to m -:f. 0 (asymmetric). The building blocks of

matter can change, from massless to massive. And they can com­
bine to new, more complex building blocks, like quarks to nucleons.
If we ignore the presently still very speculative possibility of one
single primeval form of particle, then we start with two particle spe­
cies, primeval fermions and primeval bosons. The first transition of
importance for us, then, is the one at which the primeval fermions
split up into quarks and leptons. With the decreasing temperature of
the continuously expanding universe, a point was reached, at some
1027 degrees Kelvin and about 10-35 seconds after the Big Bang, at
which each fermion had to choose one of the two possibilities: the
"grand unification" was over. Until then, transformations from one
form to the other were possible; from then on no longer. Because of
such transformations, the number of quarks could deviate tempo­
rarily from that of antiquarks, and correspondingly that of leptons
and antileptons. At this GUT transition, the basis for the present
asymmetry between matter and antimatter must have been created:
at this point, a symmetry must have been spontaneously broken,
just as the symmetry under spin flipping in the Ising model men­
tioned earlier. Up to the transition point, the fermion species were
indistinguishable, and their number was equal to the correspond­
ingly indistinguishable antifermions.
From then on, however, there existed two distinguishable species,
quarks and leptons, as well as their antiparticles, and now neither the
quark-antiquark difference nor that of leptons and antileptons had
to be zero: small, but not zero. These numbers therefore constitute
 Transitions 59

order parameters, and their finite values mean that the GUT sym­
metry given by the grand unified interaction was from now on spon­
taneously broken by the actual state of the universe.
Because later on the quarks became the basis for nucleons, the
asymmetry was an absolute prerequisite for our present world. As
we shall shortly see, in annihilation reactions between matter and
antimatter, only a possible excess of one or the other kind will sur­
vive, and so without such an asymmetry, the universe today would
only contain radiation and no matter. That is, so to speak, the other
side of the coin. If energy is deposited in empty space, it gives a pair,
fish and antifish, the possibility to surface to reality. But if subse­
quently fish and antifish should meet again, they can annihilate each
other and thus turn into electromagnetic radiation. For this reason,
a universe containing exactly the same amount of matter and anti­
matter would in the long run dissolve into radiation. There had to
be a point at which the matter/ antimatter symmetry was broken.
We mentioned that the charge distinguishing between matter
and antimatter, between quark and antiquark or between nucleon
and antinucleon, is denoted as baryon number. The baryon number
of the nucleon is +1, that of the antinucleon -1. The transition at
the end of the GUT era thus becomes

Baryogenesis,

the creation of what in the long run became a matter-dominated

universe. The trigger for this abrupt transition is found in the nature
of the force particles. Up to this point, all fermions were treated
equally, and hence also all force particles were equivalent and mass­
less. There were those mediating between quarks, those mediating
between leptons, and those changing a quark into a lepton. All these
force particles were massless, and all the interactions they initiated
were equally strong.
The end of this egalitarian world arose when the force parti­
cles changing a quark into a lepton or vice versa (they are usually
referred to a X bosons) suddenly acquired an extremely large mass.
60 Before Time began

Such a process is possible in the GUT forms studied so far; it signals

the spontaneous breaking of a symmetry of the interaction and thus
corresponds to a critical transition, similar to the onset of spontane­
ous magnetization. How such a mass-creating transition can occur
was discussed in Chapter 2: it is yet another case of the snowball syn­
drome, in which the rolling stone gathers moss after all, provided
it does not roll too fast. But what is the moss in this case? That is a
crucial question, so far not really satisfactorily answered. We shall
therefore put it aside for now and return to it when we discuss the
next, so-called Higgs transition. In any case, the huge mass increase
means that the X bosons are effectively removed from the scene.
Two fermions could no longer interact by their exchange, and hence
the transformation of a quark into a lepton or vice versa was effec­
tively excluded from now on. Neither quarks nor leptons have thus
changed in any way: both kinds are still massless. But by suppressing
the force particles that could change one species into the other, the
unified world of the species was broken. From now on, there were
quarks and there were leptons, and it was impossible to change one
into the other.
We noted that quark-lepton transitions are effectively excluded,
because the large mass of the X bosons suppresses such interactions.
Although they are extremely unlikely, they nevertheless remain
in principle possible: the X-boson mass became huge, but not infi­
nite. So quark-lepton transitions are not strictly forbidden: there is
a minute probability for a quark to turn into a lepton and then sub­
sequently find an antilepton to annihilate and turn into radiation.
That way our world could still eventually disappear-its existence is
based on the stability of quarks and in turn on that of protons, and
the possibility of quark-lepton transitions imply proton decay.
In the framework of GUT, the expected lifetime of protons has
been calculated to be some 1032 years. So one really has to wait a long
time until a given proton decays, particularly if we keep in mind
that our universe is only 1010 years old. But it's not that simple. A
human is made up of about 1029 protons, so that a tank of water
with a volume of 1000 humans should produce one proton decay
 Transitions 61

per year. Such experiments have been carried out and continue
to be in progress, with big tanks of water in old mines and similar
places, to eliminate as far as possible any effects of external radiation.
A well-known setup exists in the Japanese community Kamioka,
where deep underground a tank of 3000 tons of pure water is sur­
rounded by 1000 photo detectors. If a proton should decay in the
tank, the resulting electron would emit an observable light signal.
So far, nothing has been seen.

The interior of the Kamiokande setup. Photo: Kamioka Observatory,

ICRR, U. of Tokyo, Japan

As far as matter particles are concerned, at the end of the GUT era,
the universe contained on the one hand a dense medium of quarks
and antiquarks, and on the other an equally dense medium of elec­
trons, positrons, neutrinos, and antineutrinos. Interactions between
the constituents of the two groups were still possible, but no more
transitions from one to the other, with the caveat just mentioned.
And in both groups, there was a tiny asymmetry, a little more mat­
ter than antimatter, and correspondingly a few more electrons than
positrons, to keep the overall charge zero.
62 Before Time began

How does evolution now continue? First, we should emphasize

once again that the present considerations concerning the unifica­
tion of the particle species so far remain speculative. There are, as
indicated, some 44 or more different models, which more or less con­
tradict each other in some aspects; nothing is yet final. And testable
predictions, such as those for proton decay, have not been confirmed
by experiment. If one should find in the near future evidence for pro­
ton decay, that would of course really mean progress. But until then,
the considerations mentioned indicate only how it might have been.
Things become more definite as we progress to somewhat later,
but still very early times; we now enter a region in which terrestrial
experiments can indeed contribute significantly to our understand­
ing. So far, all matter particles were massless, and the force particles
were as well, except for the X boson, which now appears to be unim­
portant. We are now confronted by the question of what remains of
the false ground state after the transition to the true present one.
There still is the dark space energy, which continues to expand the
space of the universe. That question is even today not answered in a
fully satisfactory way. As we have known for about two years, there
is a further mysterious medium penetrating the entire universe-­
the HiBBS field. It is in nature quite similar to the dark energy, and
there are speculations relating the two. The Higgs field, however,
has a different raison d'etre: it is the cosmic snow that can clump
to the bare primeval particles and thus give them masses. Below a
certain temperature, the Higgs field changes the nature of the force
particles of the electroweak interaction; so far, there were four, all
massless. In the Higgs transition, three of them suddenly acquire
very large masses; the fourth remains massless and is from now on
called a photon. For this reason, the Higgs transition is in a sense

The birth of light.

From now on, there is electromagnetic radiation on one hand and

radioactive decay on the other. The world of leptons is thus split
into two sectors: purely electromagnetic interactions, mediated by
 Transitions 63

massless photons, and the so-called weak nuclear force, responsible

for radioactive decays. It is precisely the new, large mass of the bos­
ons mediating this interaction that makes it so weak and so short­
ranged. Moreover, the Higgs field also clings to the matter particles
and gives them small, albeit finite masses, in a similar way as seen for
polarization in a plasma. The electrons and positrons, so far massless,
now acquire their intrinsic mass, as do the quarks. It is claimed that
the neutrinos suffered a similar fate--but so far, their mass has not
been unambiguously determined. The interaction between quarks,
however, is unaffected-the gluons remain massless.
So after the GUT transition, we have encountered another change
of state, the electroweak or Higgs transition. At the GUT transition,
quarks and leptons separated; at the Higgs transition, electrons and
neutrinos go separate ways, and from now on, all matter particles
have small but finite intrinsic masses. At the Higgs point, photons and
W bosons also separated: the latter share the fate of the X bosons: they
acquire such huge masses that from now on, they are out of the game.
This transition occurred about 10-10 seconds after the Big Bang, when
the temperature had dropped to some 1015 degrees Kelvin. We have to
emphasize here that the masses created in the Higgs transition have
very little to do with what we today refer to as masses in our everyday
world. Our weight is determined by the atoms in our body; more pre­
cisely, by their nuclei. The masses of the electrons don't play any signif­
icant role, and the intrinsic masses of the quarks don't, either; we come
back to that shortly. For the world of the leptons, the Higgs transition
was the last: both electrons and neutrinos have made it unchanged
into our present world. And the photons have as well, of course.
Before we turn to the fate of the quarks, let us come back to those
three mysterious fields that permeate everything: the remaining
dark energy causing the ongoing expansion of the universe, the
GUT-field that gave rise to the large X-boson masses, and the Higgs
field making the W bosons so massive. The latter should actually
be called the Higgs-Brout-Englert-Guralnik-Hagen-Kibble field,
since all these theorists more or less at the same time proposed and
investigated such a snowball form of mass creation. We certainly
64 Before Time began

cannot resolve the priority issue; in any case, the Nobel Prize 2013
was awarded to Englert and Higgs; Brout was no longer alive at that
time. For simplicity, we shall join most others and remain with
Higgs-all the more so because Peter Higgs is an extremely mod­
est, very academic scientist, who never clamored for the fame today
assigned to him. He had concluded that the existence of the field
now named after him must also imply the existence of an associated
heavy boson; moreover, this boson should be observable in high­
energy collisions of elementary particles. By now, it seems clear that
in 2012 this particle was in fact first observed at the European Center
for Nuclear Research (CERN) in Geneva-and that led a year later
to the Nobel Prize for Englert and Higgs.

Fran�ois Englert and l'eter Higgs, 2014

Photo: CERN, c;eneva, Switzerland

The next in the line of transitions is one that already affects us

quite directly. The quarks, through their interaction, give rise to
what we today call the stron& nuclear force---and that leads to the bind­
ing of quarks to form nucleons. This hadronization transition is also, as we
have seen, the basis for the creation of the physical vacuum, empty
space as such.
Before this transition took place, the universe was densely filled
with quarks and antiquarks; no quark was ever more than a fem­
15
tometer (10- m) away from another one. At the hadronization
 Transitions 65

transition, three quarks then combined to make a nucleon, three

antiquarks an antinucleon, and a quark-antiquark pair a meson.
And these new entities, the hadrons, could exist alone, all on their
own, in empty space, without other hadrons nearby. Once more
we encounter here a transition similar to that in the Ising model:
one state, the quark-gluon plasma, becomes a gas of hadrons. The
plasma consisted of unbound color charges, and if we imagine a
"color voltage," then this medium was a color conductor-there
could be color currents. At the hadronization point, that ended;
color conductivity now dropped to zero, so that we again have an
order parameter that is finite in a certain range of temperatures and
then vanishes in another range.
But, as we have already indicated, that was not all. In the plasma,
the quarks were effectively massless: they had only the tiny intrin­
sic mass due to the Higgs field. At the hadronization point, gluon
clouds formed around every quark, and these clouds gave to the
quarks what we now, in our world, call mass. Each quark acquired
about a third of a nucleon mass, some 10-27 kg. Three such quarks
then formed a nucleon, and about 1029 such nucleons make up the
weight of our body. That weight is thus made up almost totally of
the gluon cloud surrounding otherwise very light quarks.

The effective quark mass,

created by the gluon cloud, again constitutes an order parameter:

it is (almost) zero in the plasma and then suddenly becomes finite.
The hadronization transition, at which the universe for the first
time began to look a little like it does today, occurred about 10-5
seconds after the Big Bang, at a temperature of some 2 X 1012 degrees
Kelvin. The determination of the time depends a little on the spe­
cific model for the cosmic evolution, but the transition temperature
is quite well known by now, from theoretical calculations in the
context of quantum chromodynamics, as well as through nuclear
collision experiments, in which a quark-gluon plasma is created in
the laboratory for a few fleeting moments.
66 Before Time began

Our universe now consists of "empty" space---keeping in mind

the caveats concerning "empty"-and of particles, mainly had­
rons and antihadrons, electrons, positrons, neutrinos, and photons.
These constituents form a hot gas in violent interaction, with con­
tinuous annihilation and creation. Before turning to the further fate
of this gas, we first summarize in the next picture the transitions in
the early universe and their effects on matter particles.

Temp (°K) 1027 1 o1s 1012

hadrons
Hadron
?:'-"
�,::,.?;

vacuum
quarks
primordial
electrons
fermions
leptons
neutrinos

Time (sec) 10-35 10-10 10-s

The transitions in the early universe and their effects on matter particles

But, as already mentioned, the different transitions also had their

effects on some of the force particles and modified them, which then
in turn changed the form of the interaction. We summarize these
changes in the next figure. At the GUT transition, the primeval bos­
ons gave rise to the gluons of the strong nuclear interaction and the
leptonic vector bosons, which govern the electroweak interaction.
At the following Higgs transition, the latter set split into photons
and the so-called W bosons. Gluons and photons remain massless,
while the clouds of the Higgs field create the large masses of the
W bosons and thereby assure that the weak nuclear force is indeed
weak and short-ranged. In the last of the transitions, hadroniza­
tion, the gluons play a twofold role: on the one hand, they form
the clouds that give the quarks and hence later the hadrons their
observed masses; on the other, they bind these massive quarks to
 Transitions 67

today's hadrons. The interaction between the hadrons is now taken

care of by the mesons, whose masses determine the observed range
of the strong nuclear force.

Temp (°K) 1027 1015 1012

mesons
Hadron
�c,
:,_00
�
vacuum
gluons
primordial
photons
bosons
e-w bosons
W-bosons

Time (sec) 1 o-35 10-10 10-5

The transitions in the early universe and their effects on the force particles

The world created by the hadronization transition was quite short­

lived. All matter particles were in continuous interaction; hadrons
and antihadrons annihilated each other in collision and were recre­
ated in other collisions; electrons and positrons shared this fate. But
the expansion of the universe let the temperature decrease quickly,
and with it the momenta of the constituents; as a result, the world
soon reached a state in which the production of hadrons through
collision was energetically no longer possible. This meant that in a
very short time, almost all hadrons and antihadrons annihilated each
other, producing photons and electron-positron pairs. The only sur­
vivors in this mass annihilation were the few excess nucleons existing
from the end of the GUT transition. Up to the hadronization point,
the universe was still almost symmetric, as far as matter and antimat­
ter were concerned-but fortunately only almost. Now all antinu­
cleons destroyed each other in collisions with nucleons, leaving only
the few lucky excess nucleons. They were the entire future matter of
our world; since then nothing more was added. Thus the nature of
the universe was completely changed soon after the hadronization:
68 Before Time began

except for those few nucleons, the entire sector of world originally
consisting of quarks was now fully converted into radiation and lep­
tons. The quark era was over; the universe now consisted largely of
leptons and photons: the lepton era had begun.
But its duration was soon threatened by a similar fate as that of the
quark era. In the beginning, the world contained leptons of immense
number, again almost as many electrons as positrons-the excess of elec­
trons was just what was needed to keep the world as a whole electri­
cally neutral, to balance the charges of the few remaining protons. This
meant, as before, one electron among many millions of electron-pos­
itron pairs. Things remained that way as long as the temperature was
high enough to compensate pair annihilation through pair creation.
But the expansion continued, the temperature went on decreasing.
And this meant that eventually a temperature limit was reached,
below which electron-positron pairs could only be annihilated-the
radiation energy was too low for pair re-creation. In a short time, all
the electrons and positrons that had determined the lepton era had
left the scene; the only survivors were again those that made up the
excess at the GUT transition. The excess of electrons and quarks left
over from the earliest phases was thus to become the basis for the
entire future world.
We have here ignored the neutrinos: for a while, they could still
interact with the remaining nucleons and electrons, for example,
converting a neutron into a proton and an electron: v + n � e-+ p.
But soon the temperature became too low for even that, and the
neutrinos were thus the first particles released into complete free­
dom, decoupled from all remaining matter. They could now propa­
gate unconstrained throughout the entire universe, and as such
they became a kind of vanguard for the light of the Big Bang, which
we consider in the next chapter.
At this time, some ten seconds after the Big Bang, and at a tem­
perature of about 109 degrees Kelvin, the lepton era was over. Since
electrons and positrons had largely annihilated each other, as had­
rons and antihadrons had done before, the universe now contained
almost only radiation, and number-wise, this would never change
 Transitions 69

again. Per nucleon there were some 1010 photons, so one can now
truly speak of a radiation era. The universe was still a hot plasma, in
which protons, electrons, and photons continued to interact. The
neutrons, because electrically neutral, could only participate indi­
rectly: since they are more massive than protons, they decay after
about 15 minutes into a proton, an electron, and an antineutrino.
The relative numbers of protons and neutrons had first been deter­
mined at the end of the lepton era, when neutrinos were still able
to convert one into the other. Because neutrons are slightly heavier
than protons, they were at a disadvantage; their ratio converged ini­
tially to an average of about 15 neutrons to 85 protons. But the sub­
sequent neutron decay would have led to complete extinction, had
not a salvation mechanism entered the scene, as we shall see shortly.
Since the hadronization transition, we have seen the universe
pass through a number of different states, from a hadron gas to a lep­
ton gas and then to a radiation-dominated electromagnetic plasma.
In the terminology we introduced at the beginning of this chapter,
these were transformations rather than the sharp critical transitions
encountered in the very early universe. Also, in the last of states
mentioned, the electromagnetic plasma, once in a while a proton
and electron would join to form a neutral entity, a first atom. Not
for long, however: the energetic photons of the plasma would soon
break it up again. But now something else set in: a proton and a
neutron joined to form a first nucleus, deuterium. The binding is
achieved by means of the strong nuclear force, and the mass of the
deuterium is slightly less than the sum of the proton and neutron
masses. The energy gained in this way is then emitted as radiation,
through the reaction n + p � D + y.
As long as the energy of the photons in the plasma was sufficient
to break up such a nucleus, the binding was not for long. But once
the plasma temperature had fallen below 109 degrees, the photons
were no longer strong enough to do that: the era of nucleosynthe­
sis, fusion, had started. And the process continued: two deuterium
nuclei joined to form a helium nucleus, again slightly less massive
than the sum of its constituents. And such a nuclear fusion process
70 Before Time began

could in principle continue, resulting in lithium, beryllium, and so

on. Because the masses of the nuclei were always less than the sum
of the constituent masses, the neutrons now were too light, so to
speak, to still decay: the neutron mass in a nucleus was not enough
to make a proton and an electron. This was the previously men­
tioned way out for the neutrons, the mechanism allowing them to
survive.
But such nucleosynthesis also turned out to have only a finite
time span to proceed, and again it was the expansion of the uni­
verse that was responsible. Nuclei consist of protons and neutrons,
and the positively charged protons repel each other electrically.
Fusion is therefore possible only if the momenta of the nuclei are
large enough to overcome this repulsion. In other words, once the
temperature of the plasma had decreased to a point at which the
proton momenta were no longer sufficient to allow them to get
close enough to each other for fusion to set in, nuclear binding was
over. The process could therefore take place only in a small time
window: the photons of the plasma had to be "soft" enough to pre­
vent them from breaking up the nuclei, but the protons had to be
"hard" enough to get near each other in collisions, to overcome the
electric repulsion. In addition, the expansion reduced the density
of the universe and thus also the chance for nucleons to meet.
Before we turn to the further fate of the photons, we want to
look back once more and list the main forms of transition that we
encountered in our time line of the early universe.

• The first transition occurred when the energy of the prime­

val bubble in the multiverse fell from its false ground state to
its right normal one, comparable to the escape of a bubble of
steam from superheated water.
• The second took place at the end of the GUT era, when the
force particles, which until then could change quarks into lep­
tons and vice versa, suddenly acquired huge masses, ruling out
such transformations. Quark and leptons now were "different."
• The third happened during the quark era, when the force par­
ticles responsible for the weak nuclear force suddenly became
 Transitions 71

so massive that the force in turn became very weak and short­
ranged. In this electroweak or Higgs transition, quarks and lep­
tons acquired small but finite "intrinsic" masses.

All three transitions were triggered through mysterious primor­

dial fields permeating the entire world. In the first, the difference in
energy between the false and the true normal state of the iriflaton field
gave rise to the creation of matter. In the second, one expects that
the energy difference between such false and true states of a field so
far without a name gives rise to the large masses of the X bosons and
thus to the splitting of primeval fermions into quarks and leptons.
In the case of the third, the energy difference between the normal
states of the Higgs field at high and low temperatures leads to the
intrinsic masses of fermions and W bosons.
The Higgs field is the only one among these three fields that now
appears to be quite well established: after the transition, there remains
the heavy Higgs boson, which was finally seen in 2012, created in
high-energy proton-proton collisions at CERN. The production of
a corresponding inflation boson for the first transition seems forever
out of reach for terrestrial experiments. Nevertheless, the continuing
and even-increasing expansion of the universe caused by this field is
experimentally confirmed. In the case of the second, the source of
the X boson, we can only note that there is ongoing intensive search.

• The fourth transition in the time line is of a basically different

nature: it has colored quarks joining to form colorless hadrons,
which can exist individually and thus allow the appearance of
the physical vacuum, space without matter.

In this case, the responsibility is not attributed to some mysteri­

ous medium permeating the entire universe. The crucial role is here
played by the force particles of the strong interaction, the gluons:
they give rise to the effective mass of the quarks through a gluonic
snowball effect, and they bind these massive quarks to form hadrons.
This transition is today being studied with increasing precision on
both sides-before and after hadron formation-and it is investi­
gated experimentally in high-energy nuclear collisions.
72 Before Time began

The four transitions we just considered were indeed critical phe­

nomena, in the sense of critical behavior as we have defined it. In
each case, a symmetry effective up to that point was spontaneously
broken; the universe fell from a state of higher to a state of lower
symmetry. Our picture thus corresponds to an initial, primordial
universe that was maximally symmetric, and in the course of its
evolution this symmetry was reduced more and more through
spontaneous breaking. To have a pictorial illustration: we start with
a three-dimensional sphere turning into a two-dimensional circle,
which in turn is transformed into an octagonal star and finally into
a square. The sphere was invariant under all spatial rotations, the
circle only under rotations in a plane; the octagonal star remains
unchanged under eight rotations of 45 degrees each, and the square
only under four 90-degree rotations. Something similar happens
when water freezes to become ice: the interaction between the water
molecules remains invariant under all spatial rotations, but the ice
state only under a finite number (like four) of rotations through a
finite angle around a crystal axis. The symmetry of the molecular
state of water is spontaneously broken when the liquid freezes.
After the earlier phase transitions and the subsequent transforma­
tions, the world now consisted essentially of light nuclei, including
protons as hydrogen nuclei (the remaining neutrons had decayed),
and in addition electrons, photons, and neutrinos. The latter were
already liberated, because of their extremely weak interaction with
other constituents. What remained was thus a still rather hot plasma
of nuclei, electrons, and photons. Since both nuclei and electrons
carried electric charges, the photons were in continuous interaction
with both-they could not (yet) escape.
Nuclei and electrons were able to combine to electrically neu­
tral entities only when the temperature had decreased still further:
now the first atoms entered the stage, and the universe had reached
a state in which it still is today. The time between the beginning of
nucleosynthesis and the formation of the first atoms was truly an
eternity as measured in the scales of the early universe: only some
380,000 years after nucleosynthesis were all nuclei and electrons com­
bined into atoms, and the universe contained only electrically neutral
 Transitions 73

constituents. Because photons can only interact with electric charges,

they now had no more counterparts; they were free and could
spread out unconstrained. We can thus slightly extend the words of
Democritus: "In reality, there exist only atoms, radiation, and empty
space," including the neutrinos in the radiation. While the amounts of
matter and radiation are fixed forever, the empty space is not static­
thanks to the remaining dark energy, it continues to expand. That has
a crucial effect on radiation, as we shall see in the next chapter.
First, however, we want to combine into one picture all these
many steps taken by the universe since the Big Bang. It is divided
into an early, extremely hot and dense part, before the appearance
of the physical vacuum, and another part, still rather hot and dense,
but already containing what we call empty space, the stage for all
that came since then and still exists today.

multiverse

primordial
particles
t (SUSY)
fermions
. bosons
Big Bang
t (GUT)

fermiogenesis

baryogenesis

electrons
hadrons
Higgs
photons
(intrinsic
electrons
masses)
nucleons
hadrosynthesis
(inertial
masses)

annihilation

The stages of the universe from the gig gang to the formation of atoms; in
the blue region the physical vacuum already exists
74 Before Time began

In this picture, we only show the succession of steps as they

occurred in the evolution of the early universe. But it is of course
also of interest to learn when what happened, at what "time after
the Big Bang," and what the state of the universe was at that time,
"how hot" it was. The answers to these question obviously become
less and less certain the further back we go. We try it anyway, and we
begin with the time line.
As we have mentioned several times, the present atomic uni­
verse was born some 380,000 years after the Big Bang; that was when
electrically charged nuclei and electrons combined to form neu­
tral atoms. Before this time, there was the radiation era in which
unbound nuclei and electrons together with photons and neutri­
nos were the essential constituents. This era had started some ten
seconds after the Big Bang, because at that time most electrons and
positrons had annihilated each other, turning into radiation, except
for those electrons needed for today's world. The hadron era before
this had ended about one second after the Big Bang, when nucleons
and antinucleons suffered this fate, annihilating each other except
for what is left today, the rest becoming radiation. A large part of
today's photons was thus born in these two transitions, nucleon­
antinucleon and electron-positron annihilation. The hadron era
itself began in a genuine transition, 10-5 seconds after the Big Bang,
when the density of the universe had dropped to a point at which
the required quark density risked falling below that needed to assure
each quark its partner. Quarks and antiquarks now were forced to
combine to hadrons and/or antihadrons.
This transition, as mentioned, is the first that is quite well under­
stood both theoretically and experimentally, through calculations
based on quantum chromodynamics (the theory of strong nuclear
interactions) and though experiments studying high-energy
nucleus-nucleus collisions.
Before this, we have the Higgs transitions, which gave quarks and
leptons their small intrinsic mass and the weak bosons their large
ones. The determination of the mass of the Higgs boson in experi­
ments at CERN has also here provided an empirical basis.
 Transitions 75

Time Big Bang Temp. (°K)

1o-34seconds

massless quarks & leptons

Higgs - -
- - Quark- - - -- - - - - - - - - -- - --
10-10seconds massive quarks & leptons 1015
�-----+---hadronization - - - - -

-
10 5seconds 1012

1 second 1010
e•+e- -r's -

10 seconds
nucleus+ eledron
380,000 years -atom 3000

today 3°K

Time and temperature scales in the evolution of the early universe

The situation at still earlier times is quite uncertain, specula­

tive, and dependent on the theoretical basis adopted. One tries to
extrapolate the energy dependence of the couplings for strong and
electroweak interactions to a point, at which they become equal:
that would define the GUT transition, at which primeval fermi­
ons separated into quarks and leptons. Present considerations lead
to the values shown in the previous figures. Comparing the time
of the GUT transition shown there with the end of the inflation
as discussed in the first chapter seems to have the GUT transition
occurring before the end of inflation. That cannot really be possible,
because only at the end of inflation, after the transition to the true
ground state, did the energy for the creation of particles become
available. The numbers simply are not consistent. We can only note
here that neither the 10-34 seconds for the end of inflation nor the
10-35 seconds for the GUT transition are known well enough to
draw any conclusions. And if we believe supersymmetric considera­
tions, there should even have been a yet earlier transition, in which
76 Before Time began

the primordial constituents were split into fermions and bosons.

Our understanding of such a transition, as that of the GUT transi­
tion, is today simply not adequate. The other scales, as estimated, are
included in the following picture.
At this point we should emphasize once again that the picture we
are presenting here is that corresponding to the "new" cosmology,
based on the work of Alan Guth, Andrei Linde, Paul Steinhardt, and
others. It has our universe appearing as just one of the many bub­
bles continuously created in the multiverse. The appearance of the
bubble and the subsequent inflation are now physical phenomena.
The "old" cosmology was based on general relativity and its tempo­
ral and spatial singularity, constituting a unique Big Bang outside
all "normal" physics. Following this singular event came the Planck
era, in which all forces, including gravity, were unified. In such a
picture, the separation of gravity and inflation occurred still sooner
after the Big Bang, after about seconds. On the one hand, this "old"
view suffers from the absence of a quantum theory of gravitation,
a quantum theory combining strong and electroweak forces with
gravity. It is not evident whether such a theory would still lead to any
singularity. On the other hand, it seems not clear today if gravity, as
a possibly "emergent" force, has any place in such a unified theory.
Conceptually, at least the "new" cosmology provides a rather satis­
factory framework, albeit at the cost of introducing worlds outside
our experimental reach.
 5

The Light of the Big Bang

And God said, Let there be light, and there was ligh t .

THE BIBLE, GE:--iESIS 1.3

The light of the Big Bang is today no longer visible to our eyes; never­
theless, it is present everywhere, uniformly distributed throughout
our universe: it the cosmic background radiation already mentioned,
discovered in 1964 by Arno Penzias and Robert Wilson. The wave­
lengths of visible light range from the ultraviolet (4 X 10-7 m) to the
infrared (7 X 10-7 m). The 2.7 degree Kelvin cosmic radiation, in con­
trast, today has an average wavelength of about half a millimeter and
thus falls into the microwave range. But when it was first liberated,
some 380,000 years after the Big Bang, its wavelength was a thou­
sand times less, in the yellow-orange range of our visible light. So at
that time, the sky was not dark at night, but bright yellow! Shortly
afterwards, the expansion of the universe moved the wavelength of
that light into the invisible for us, but if some creatures were to exist
with sensors for these wavelengths, then they would find that even
today the night sky is not dark. In fact, we do have access to such
sensors: in the absence of a signal from some station, our TV screen
registers that well-known milky noise. A small part of it, about 1%, is
due to the cosmic background radiation. So at frequencies between
two stations, we receive the message from the Hig Hang.
In the last chapter, we showed that about 380,000 years after
the Big Bang, when electrons and nuclei had combined to form
electrically neutral atoms, there was nothing left with which pho­
tons could interact in any way. l�rom this point on, for them the
world was transparent. So the cosmic background radiation we
78 Before Time began

400 nm 500 nm 600 nm 700 nm 800 nm

ultraviolet - infrared
' ,
' ,
' ,
,
'
,
'
'
' ,
' ,
' ,
'
'
,
'
,
' ,

radiation X-rays UV light infrared microwave radio

wavelength 10-10 10-7 10-6 10-2 102
(m)

---'l
The spectrum of electromagnetic radiation ( 1 nm = 10 m)

observe today consists of the photons liberated at that time; they

are unchanged since then, except that the subsequent expansion
of the universe has increased their wavelengths correspondingly.
We know from present experiments that hydrogen atoms become
ionized, break up into unbound protons and electrons, at about
3000 degrees Kelvin. So at that time that must have been the tem­
perature, and because the background radiation today is a thousand
times cooler, space must have expanded by a factor of a thousand
since then.

Time:

380 OOO years after BB 14 billion years after BB

3000°K 3°K

The stretching of the photon wavelength through the expansion of space

 The Light of the Big Bang 79

Nevertheless, the distribution of the photons contains otherwise

all the information imprinted on them at the time of liberation, and
that is the earliest direct information we can get about the beginning
of our universe. Before that time, the world consisted of an interact­
ing plasma of electrons, nuclei, and photons; after that, there were
only electrically neutral atoms, so that the photons no longer had
any partners with which they could interact. The formation time of
atoms-the cosmologists call it the time of last interaction or decouplins
time--this time is for us something like a temporal layer of clouds
that stops our vision. All earlier happenings can be addressed only
indirectly, and that is why today we have to search for the tiniest
irregularities in the incredibly uniform distribution of the cosmic
background radiation. These can perhaps give us at least some hints
about what happened before, beyond the cloud surface.
Penzias and Wilson had run into the background radiation just by
accident. They had been asked by the Bell Telephone Company to
look at the possibility of microwave communication based on high­
altitude balloons. In the course of their studies, they encountered a
mysterious interference radiation, noise for which they could find
no origin. This radiation was there day and night, and in all direc­
tions of the sky. Its origin became clear soon afterwards. A group
of theorists at Princeton University had just predicted that there
should be remnant radiation from the Big Bang, and as it turned
out, that was the source of the apparent noise.
Penzias and Wilson only measured radiation of a fixed wave­
length, but subsequent studies covered the entire spectrum. They
found with unbelievable precision the distribution of black body
radiation of a temperature of 2.725 degrees Kelvin. This is the radi­
ation emitted by a container kept at a constant temperature of that
value, and so it corresponds exactly to what one would expect if at
an earlier time a perfectly uniform gas of photons had been liber­
ated. In the following picture, we show the data taken by a special
detector ( COBE) installed on a space satellite. It is compared to the
black body radiation mentioned earlier, although it's hard to see the
comparison. We have here one of the rare cases in physics where
80 Before Time began

the measuring error is considerably smaller than the thickness of

the shown theoretical curve. With incredible accuracy, one observes
throughout the universe black body radiation of a temperature of
2.725 ± 0.001 degrees Kelvin.

400

,., (
T = 2.725 ± 0.001 °K
300
-;:::-
V>

:;:,
� 200
+"
·;:;;
c
<11
+"

E
100

100 200 300 400 500

Frequency (GHz)

The spectrum of cosmic background radiation, theory, and COBE data

The measurement of this radiation was in fact not quite as simple

as it seemed. On Earth, it would be like looking out of a steam-filled
room through dirty windows. In particular, clouds and other gas
effects have a considerable influence on the precision. That is why the
stage of real precision data began only when space satellites became
available, taking data far from any terrestrial interferences. The
beginning was made by the Cosmic Background Explorer (COBE)
toward the end of the 1990s; the results, including the spectrum just
shown, brought the 2006 Nobel Prize in physics to the leader of the
effort, George Smoot, and his collaborator John Mather. The next
step was the Wilkinson Microwave Anisotropy Probe (WMAP); it
was followed by the Planck detector, which in 2009 was brought into
orbit by an Ariane rocket; we will shortly return to its latest results.
But all these measurements first had to be corrected for effects
totally unrelated to the Big Bang, and in the spectrum shown previ­
ously they are already taken into account. These corrections sound
a little like science fiction: we are in fact living on the
 The Light of the Big Bang 81

Spaceship Earth.

Our Earth orbits around the sun at some 30 km/ s; the sun in turn
moves around the center of our galaxy, the Milky Way, at 220 km/s;
and even our galaxy is not stationary relative to the rest of the observ­
able universe. The sum of these effects has us here on Earth moving
at 390 km/s in a well-defined direction through space, through the
container of the cosmic background radiation-the space, in which
the entire matter of the observable universe is at rest.
What this means is well known from the world of sound waves. If
we move toward the source of the sound, the wavelength is short­
ened; the pitch of the tone becomes higher. In the opposite case,
moving away from the source, the wavelength grows; the pitch of
the tone goes down. Physicists call this the Doppler effect, named
after the Austrian physicist Christian Doppler.
And if we now travel on our spaceship Earth with a certain speed
through the waves of background radiation, then the length of the
waves in front of us is contracted, those in the back dilated. That implies
an ultraviolet shift in the direction of travel and an infrared shift toward
the back. These shifts are clearly observed; on the one hand, they tell us
in which direction we are moving through space, but on the other, they
have to be subtracted in order to obtain the correct spectrum of the
radiation. In the following picture, this effect is schematically shown:
here the colors are simply meant to indicate that the Earth is moving
through space in the direction from the upper right to the lower left.

Doppler-shift of the cosmic background radiation

Reproduced with permission from DMR, COBE, NASA.
82 Before Time began

As mentioned, this effect has to be "subtracted" from the observed

temperature distribution before one can draw any conclusions. A
further effect is caused by clouds of gas that circulate in the plane of
our galaxy. Just recently, some dramatic conclusions about the early
universe had to be corrected, because the measured results were
finally found to be due only to interstellar clouds of dust.
When all is taken into account, one obtains the following picture,
recently taken by the Planck detector. We now indeed see deviations
from a uniform temperature, but we have to keep in mind that the
different colors indicate fluctuations of less than a thousandth of a
percent: blue means a little cooler, red a little warmer. The informa­
tion that the background radiation can give us about the universe
before the time of last scattering is thus based on tiny variations.
What can these tell us?

Distribution of the cosmic background radiation (fr o m the Planck detector)

First, we should emphasize once more that this picture leads to

an average uniform temperature of 2.725 degrees Kelvin, with that
precision; and only then do we turn to the previously mentioned
tiny deviations from the average. Why are photons coming from
certain spots of the universe a little warmer or cooler?
Here we should remember that the inhomogeneous structure of
the present universe, with stars and galaxies on the one hand, empty
space on the other, must have come into being at some time. It's not
 The Light of the Big Bang 83

possible that everything was completely uniform and homogene­

ous, and then suddenly the world became "clumpy." Even gravity
cannot create clumps in a perfectly uniform world. For that reason,
cosmologists have always insisted that one must find fluctuations in
the temperature distribution of the cosmic background radiation, if
one only looks carefully enough. And the improvement in obser­
vation techniques, from COBE to WMAP to Planck, did bring the
required increase in precision.
So before the decoupling of the photons, in the hot plasma of
nuclei, electrons, and photons, there were indeed certain, albeit tiny
irregularities in the density of the medium. These must have first
appeared in still earlier stages, as minute quantum fluctuations in
the expanding primeval world, at the time of inflation. In the prime­
val medium, quantum theory tells us, there were again little bub­
bles of higher or lower density, similar to the bubble that originally
led to our universe itself. Through inflation, the spatial extension of
such fluctuations was dramatically increased, but at the same time,
their deviations from the average were greatly reduced-everything
was smoothed out, until today, only the tiny fractions of less than a
thousandth percent remain. The macroscopic areas of higher den­
sity formed in this way later on gave gravity a chance to first create
gaseous clouds, and still later form stars and galaxies out of these.

Expansion of a quantum fluctuation

84 Before Time began

When matter and photons became decoupled, photons coming

from a region of denser matter had to overcome a stronger force
of gravity than those from regions of lower density. The former
were restrained, their wavelength was infrared-shifted, while the
others had it easier and thus appeared faster, ultraviolet-shifted.
The tiny irregularities appearing in the temperature distribution of
the cosmic background radiation are therefore indeed witnesses of
the fluctuations that were the seeds for the subsequent structure
of the universe, the origin of stars and galaxies. What in the earliest
times of the early universe was only a tiny spot of slightly higher
density later on became the Milky Way, long after it had created the
fluctuations in the spectrum of the cosmic background radiation.
But this radiation has still more to tell us. In the plasma, the
attraction of gravity and the expanding pressure of the medium
compete with each other. An area of higher density contracts
under gravity; but this increases the temperature, so that the radi­
ation pressure increases as well, and eventually the bubble expands
again. Such a process, a temporal variation and contraction of a
medium, is in fact quite familiar to us: it happens in sound waves.
So in a sense, at the time of last scattering we encounter not only
the light, but also

The sound of the Big Bang.

The density oscillations in the plasma act similarly to an under­

water bellows that is contracted and released in a certain rhythm.
The resulting pressure pulses create sound waves that propagate
through the water. Whales and dolphins use this effect for commu­
nication, and our sound waves in air are of this nature. In this way,
the small density irregularities created by quantum fluctuations
cause the early universe to "sing." The spatial regions formed at the
end of inflation form "sound boxes," whose size is determined by
the speed of sound waves in the plasma medium. In other words,
the size of the sound box is the region accessible by sound waves up
to the time of decoupling. The speed of sound waves in a relativistic
 The Light of the Big Bang 85

medium is about 57% of the speed of light, and so the volume thus
created has a spatial radius of some 200,000 light years and a lifetime
of 380,000 years until decoupling sets in. The modification of dens­
ity modifies gravity effects, and as a result the spectrum of the pho­
tons is slightly modified by this hill-and-valley structure. Because at
the time of last scattering, matter and photons become decoupled,
the bellows stop working at this point-the song is frozen. And the
challenge for satellites is then clear: measure sufficiently precisely to
identify the frozen melody.
The following picture shows what this requires. Observers have
to take their bearings in the radiation sky at such a small angle that
they cover exactly one sound box and not more, because other­
wise other sounds come in and create a cacophony. As mentioned,
the sound box had a size of about 200,000 light years at decoupling
time; the subsequent spatial expansion increased this by a factor of
1000, so today we get 200 million light years. For the observer, that
means an angle a of a little more than one degree. Up to this angle,
one observes the emission of the primeval sound box; for larger
angles, the result becomes an average over several emitters and thus
approaches eventually an equidistribution. Only for the noted small
angle can we hope to hear a pure sound.

Big Bang observer

decoupling
today

The sound box of the Big Bang

86 Before Time began

The wavelength of the sounds produced in this way is determined

by the velocity of sound in the given medium and by the size of the
sound box, which is here given by the length of time available. Using
the values quoted earlier, we get a wavelength of about 200,000 light
years for the fundamental tone, measured at the time of decoupling.
As noted, since then the universe has expanded by a factor of 1000,
and therefore the waves of the cosmic sound have increased accord­
ingly. As a result, the wavelength of the fundamental tone comes
for us to 200 million light years. In addition to the fundamental
tone, there are harmonics, overtones, of higher frequencies, integral
multiples of the fundamental tone frequency. In the following pic­
ture, this is illustrated schematically for a closed flute.

::

The fundamental tone and the following two harmonics ( top to bottom)

All these waves thus created a topography, a landscape of hills

and valleys, through which the photons had to work their way out
into our world. Because of this work, their spectrum was shifted up
or down and thus led to the spotty carpet of cosmic radiation that
we observe, if we look carefully enough.
To catch the fundamental tone, one has to observe radiation in an
angle of one degree, as we have noted. To get the higher harmonics,
one has to observe at correspondingly smaller angles, at finer resolu­
tion. What emerges in this way as a function of the angle of observa­
tion is illustrated in the following picture. The apparent regularity
 The Light of the Big Bang 87

of the harmonics is, however, an idealization, because in reality, the

remnant medium present in the interstellar space leads to an effec­
tive dampening.

Q
<lJ
<lJ 100
0,
CJ)
"O fundamental note
'+-
0
-:5
c:
g
.s
c: 50
0
·.;::;
harmonics
CU
·:;:
CJ)
"O
�
:J
+"'

�
<lJ
a.
E
<lJ
+"'
goo 10 0.1°
angle of observation (degrees)

Variation of the temperature of the cosmic background radiation as a function

of the observation angle

What can we learn from such measurements? An essential ques­

tion can be answered with the help of the first maximum: what is
the form of the space in which all this takes place? We remember
that a long time ago, many people pictured the Earth as a disk, a
plane--even though in ancient Greece doubts were already being
expressed about such a scenario. We now want to repeat this ques­
tion in a very much larger framework.

The form of the space

in which our universe "lives"-the form of the stage on which

everything takes place-is an issue that has to be clarified. The most
familiar form is of course that encountered in our normal world: a
flat space in which two parallels never meet and in which the angles
of a triangle add up to 180 degrees. We find it normal to picture all
space like this; but then, our ancestors also found the Earth as a disk
88 Before Time began

perfectly normal. We get an alternative if we consider the surface

of a sphere as the underlying space. Now all parallel lines emerging
from the equator at right angles meet at the North Pole---they are
our lines of longitude. And the angles of an equilateral triangle, with
the equator as baseline, add up to 270 degrees.
Such a world can be extended to three (or more) dimensions,
even though everything then becomes a bit more complicated.
But we can readily distinguish between the two possibilities, flat or
spherical space. If two explorers start out at the North Pole, with an
angle of 90 degrees between their lines of travel, then they arrive at
the equator after some 10,000 kilometers, separated by a distance of
again 10,000 km. If the Earth were flat, their distance of separation
would be much greater, some 14,000 km. In other words, in a spheri­
cal world distances "shrink"-the 14,000 km becomes 10,000 km.

North Pole

10 000 km
10 000 ";/,�ooo krr

// ""
L/--·---- ·
- ----""
14100 km
10 000 km

Distances on the surface of a sphere (left) and on a flat plane (right)

Fortunately, we can stick to our well-known view of space as flat.

If we know how large the sound box was at the time of decoupling,
then we find in flat space (for two dimensions we would have a pla­
nar triangle) indeed the angle mentioned earlier, about one degree,
for the first maximum. For a spherical world, the angle would
become smaller, in two dimensions by a third. The observations
show that that is not the case---our universe is flat, just as our closer
surroundings are.
At this point, it is perhaps useful to distinguish two features: the
form of space and its rate of expansion; we will deal in more detail
 The Light of the Big Bang 89

with the latter in Chapter 8. For the form of space, we have three
different types: the two just mentioned, flat or spherical, and in
addition hyperbolic (saddle-shaped). If we project the behavior of
triangles onto a plane for each of these three cases, we get the fol­
lowing picture.

spherical flat hyperbolic

Triangle structures for different forms of space

W hile in flat space, the three angles add up to 180 degrees, their
sum is larger in a spherical world, less in a hyperbolic.
The stage for the actual cosmic background radiation is, however,
not really empty: our universe contains assorted bits and pieces, and
that, as we know from Einstein, has an effect on the spatial structure.
Obviously there is the "visible" matter known to us, planets, stars,
galaxies. In addition, there is a further participant, one that is quite
problematic for present-day astrophysicists. Galaxies such as the
Milky Way are bound together by the force of gravity between the
participating stellar constituents; gravity determines the shape and
size of the galaxy just as the sun and the planets determine the size
of our solar system. For the Milky Way, just as for the solar system,
the motion of all constituents has to be in accord with the gravita­
tional forces active in the system. That, however, does not work out:
to bind a galaxy such as the Milky Way and to arrive at the motion
found for its stars, one needs more matter than is visible, very much
more; in Chapter 7, we will return to this issue in more detail.
So once again there has to be an invisible player in the game, dark
matter, not to be confused with dark energy. Dark matter surrounds
galaxies, while dark energy is uniformly distributed throughout the
90 Before Time began

entire universe. The amount of dark matter needed for the shape
and behavior of galaxies is many times that of the visible--it must
be there, but it cannot be seen; it interacts with our world only
through gravitation. How something like that fits into our frame­
work of physics remains so far an open question; not even a black
hole can do it, because dark matter has to be distributed over a very
large region, whereas black holes are extremely dense and concen­
trated. Whatever the solution, the universe must contain altogether
very much more matter than we can see. And the total amount,
dark and visible, contributes to the effect of gravity, to the deforma­
tion of space. To counterbalance that, we need the often- mentioned
dark or space energy, which has the opposite effect. Based on today's
observations, this space energy must add up to about 75% of the total
energy of our universe: we have 75% dark space energy, 20% dark
matter and 5% visible matter, as illustrated in the following picture.

20%
75%
dark matter
dark energy
5%
visible matter

The composition of the universe

Let's reiterate the underlying logic. Visible matter we can meas­

ure. The structure of the Milky Way requires four times more
dark matter, as we shall see, in order to arrive at the force of grav­
ity needed for shape and binding of the galaxy. The position of the
first maximum in the spectrum of the cosmic background radiation
shows that the space of the present universe as a whole is flat. That
requires the additional 75% dark energy, whose force of expansion
compensates the gravitational attraction of dark and visible matter,
leading to a flat space. As mentioned, we'll return to all this in more
detail in Chapter 7.
 The Light of the Big Bang 91

The further investigation of the cosmic background radiation is

today a major topic of research. Here we want to address a subject
recently discussed quite often, the polarization of the radiation. Light
is an electromagnetic wave, whose amplitude "oscillates" orthog­
onal to its direction of motion; if we stand at a given point through
which it passes, we experience an alternation of peaks and valleys.
Normal light consists of a superposition of such waves, oscillating
in all directions at right angles to the line of motion. Such light is
referred to as unpolarized. If a beam of such light is passed through a
slit allowing only oscillations in one direction, then the light on the
other side oscillates only in the direction of the slit; the other com­
ponents are stopped: the light is now polarized.
In the following picture, we show for simplicity only two waves
oscillating at right angles to each other, of which one is stopped by
the orientation of the slit. The stopping of the other components
not along the line of the slit of course reduces the intensity of the
light, and therefore this effect is often used in sunglasses. The light
reflected from a surface of water is largely polarized to oscillate in
the plane of the water surface; sunglasses made of material with slits
orthogonal to the water thus stop much of the incident sunlight ...
unless you turn your head by 90 degrees.

Polarization of a light beam

What can we say about the polarization of the cosmic background

radiation? The light reflected from a surface of water is not, as one
might think, a "redirected" beam; instead, it is a completely new
beam. The incident photon hits an electron in the water surface
and gives a kick. As a result, such kicked electrons in turn emit
92 Before Time began

"secondary" photons, and the photons created in this way form the
"reflected" light we observe. The motion of the electrons triggered
in this way occurs largely on the surface of the water, and this causes
the newly created photons to oscillate in this plane.
The origin of the photons in the cosmic background radiation
is quite similar. These photons are also "reflected": they were cre­
ated in the last collision of plasma photons with electrons. However,
the electrons now are not in a plane, and the plasma photons come
from all sides. As long as all photons are equally strong, there is no
preferred direction, and the resulting radiation is unpolarized. But
if the intensity of the incident photons depends on their direction of
motion, then the emitted photons will transmit the information of
this direction through polarization.
As a result, the observation of the polarization of the cosmic back­
ground radiation can provide information about eventual bumps or
irregularities in the universe just before the threshold of last scat­
tering. Such anisotropies can arise through fluctuations in density or
temperature, but they can also be triggered as an aftereffect of gravi­
tational waves created at the time of inflation. Even at the much
later time of last scattering, such waves can still lead to remnant tur­
bulences. The polarization formed as a result of such gravitational
waves is quite different from that due to density or temperature
fluctuations, so that their observation would provide a first hint for
the existence of inflationary waves. The polarization due to density
of temperature fluctuations is symmetric, pointing toward the cen­
ter of the irregularity. If there is in addition a spatial wave surviving
from the time of inflation, then this wave will provide direction and
thus destroy the symmetry. The effect of such a gravitational wave is
expected to be very weak and hence very difficult to measure.
It is therefore no surprise that the announcement made two years
ago by the American research team BICEP led to great excitement:
they thought they had found the first evidence for asymmetric
polarization due to gravitational waves. But, as in many other cases,
unfortunately the champagne corks popped here too early. Inter­
stellar dust clouds can lead to a similar effect, and in the direction
 The Light of the Big Bang 93

Polarization scheme for a density fluctuation (left) and the modification

due to a gravitational wave (right)

of the sky studied by BICEP there were such clouds. More extensive
studies in different directions of the sky, carried out with the Euro­
pean Planck detector, showed that when the clouds were absent, so
was the effect seen in the polarization.
Nevertheless, the great interest in gravitational waves persists.
Until now, electromagnetic waves arc our only line of communica­
tion with times long gone. But neither dark matter nor dark energy
can experience electromagnetic interactions-they can only com­
municate through gravity. And just as electromagnetic interactions
lead to the emission of light waves, gravitational interactions must
produce gravitational waves-so predicts Einstein's general theory
of relativity. What is so fascinating is that these waves give us direct
information about the deformation of space and time; what is prob­
lematic is that these spacetime waves are very weak and therefore
difficult to detect.
Let us try to construct a detector for this purpose. We start with
a pipe several kilometers long and evacuated, to eliminate all other
possible disturbances. On one end of this pipe we install a light
source, on the other a mirror. Now we send out a light signal and
measure how long it takes until the reflected signal returns. If dur­
ing the measurement a gravitational waw passes the site, then the
path of the light signal is sometimes longer, sometimes shorter, and
that should show up in the time of travel of the signal.
It is easy to imagine that the time measurement for such an
arrangement is hardly possible on a terrestrial scale: for a length
of ten kilometers, light needs thirty-thousandth of a second, and
94 Before Time began

to measure tiny deviations of this is well beyond the possibilities

of most watches. To avoid this difficulty, the experimentalists use
a setup in which two such pipes come together at a right angle.
The light beams meeting at this corner point create interference
effects, and if the lengths of their paths are contracted or dilated,
these interference effects are modified. W ith the help of such a
setup, it indeed becomes possible to measure wave deformations
of space.

reflecting mirror

light source

( ;ravitational wave detector

The light beam is split by the dividing mirror and sent into two
arms; at the end of each arm, the corresponding beam is reflected.
The reflected beams, when they meet, create an interference pat­
tern, which is measured in the detector. Gravitational waves modify
the effective arm lengths, and thus change the interference pattern.
In February 2016, a team of scientists from the USA and Europe
(LIGO) used two such detectors well separated (one in the state of
Washington and one in Louisiana in the USA), and they received
signals-according to the analysis-due to gravitational waves.
These waves are not remnants of Big Bang inflation; they were cre­
ated in the collision of two massive black holes some 1.3 billion years
ago and have arrived here only now. If these measurements are
 The Light of the Big Bang 95

confirmed-and after the BICEP affair one has become even more
careful-then this direct proof of gravitational waves would consti­
tute an absolute breakthrough.
Information about the collision of the two black holes would
only be the start of a new way of information transmission from dis­
tant times. As we indicated in connection with the BICEP experi­
ment, the inflation at the birth of our universe must have created
gravitational waves that even today cause spacetime to swing, albeit
very weakly. However, from what has been said, it is clear that ter­
restrial experiments soon reach their limits-we have to let light
travel over much larger distances in order to see clear modifications.
The hopes and expectations of scientists therefore rest on the new
eLISA project, a joint effort of several European laboratories. It is
based on an interferometer situated on one satellite in space, reflect­
ors on others, so that the effective length of the "pipes" is increased
from ten to a million kilometers. If all goes according to plan, the
measurement and investigation of gravitational waves will be carried
out with this new tool-perhaps as fruitful as cosmic microwave
radiation has so far proven to be.
In closing this chapter, we want to briefly mention a further devel­
opment of the universe. Nucleons and electrons are now bound to
electrically neutral entities, so that cosmic microwave radiation can
proceed unhindered through space. The small fluctuations in dens­
ity of matter now begin to show consequences. As long as radiation
and matter were in interaction, radiation pressure could break up
eventual clumps of matter created by gravity. But now radiation is
out of the game, since matter has become electrically neutral. Now
denser clouds of gas can form, which through gravity attract other
such clouds and thus become larger and still denser. With time,
these "proto-stars" will become real stars: with increasing density,
the kinetic energy of the nucleons became ever larger, until a point
was reached at which the collision of four nucleons would form a
helium nucleus: nuclear fusion had started. The energy liberated
in this process is sent as radiation into space: the star has started to
shine. This
96 Before Time began

Star rise

began some 500 million years after the Big Bang and thus ended an
era which the cosmologists call the dark ages, since between the last
scattering and the creation of stars nothing was shining, apart from
the cosmic microwave radiation, of course. But now the lights go on
in the universe, stars light up the sky, darkness is over.
The appearance of the stars leads us to our next big question: how
could this universe, this hot atomic gas, uniform up to a millionth
of a percent, produce all the stellar structure of our present world?
We recall that there are laws of thermodynamics claiming that dis­
order, lack of structure will always increase in the course of time.
What does that mean, and how, nevertheless, could stars and galax­
ies appear on the scene?
 6

Structure and Form

The whole is more than the sum of its parts.

ARISTOTLE, METAPHYSICS

Structure and form surround us wherever we look. According to

the Bible, in the beginning, "the Earth was without form and void,
and darkness was over the face of the deep." The physics of the
newly born universe assumes the absence of all forms and scales;
just after the Big Bang, the universe consisted of pure, concentrated
energy, without the least structure. There was not even nothing,
no empty space, and there was not the slightest indication of all the
complexity th at would eventually arise from this simple primordial
state. Nevertheless, the seeds of the coming diversity must already
have been present in the force fields making up the uniform primor­
dial medium. Although there was no way to see it, the structure­
less world must have already contained in a latent form the seeds
for the structures later to emerge. The appearance of structure out
of many simple identical components is today found in many areas
of science, and is in fact generally denoted as emer8ence, as collective
self-organized formation. And not only structures are emergent;
there also exist emergent observables of all kinds: temperature, den­
sity, pressure, to name just a few. A single atom or molecule does
not have a temperature; such quantities characterize the collective
behavior of many individual components.
Emergence is familiar in our daily life: out of an apparent uni­
formity, in the absence of any form, suddenly specific formations
can appear. We have discussed the states of water. Stearn consists of
98 Before Time began

very weakly interacting molecules; it is a gas of almost free particles,

the same in all directions and completely without structure. There
is no way to tell that with decreasing temperature such a system
will first suddenly liquefy and then eventually turn into ice with
its crystal structure. However, the basis for such transitions must
have been imprinted in the form of the interaction of the gas, like
inheritance in genes. The complexity arises spontaneously, without any
external action; in a latent form, it must therefore have been present
all along. In this way, the laws governing the newly born universe,
completely without any structure, just after the Big Bang, before
the existence of gravity or electromagnetism, of nuclei, atoms, or
molecules-these laws must have already contained the prerequi­
sites even for the formation of an ice crystal or a snowflake.
Wherever we look today, we sec a world of infinite diversity, of
structures, colors and sounds, of processes and events. As far as we
know, even animals are aware of such a diversity. We humans, in
contrast to them, can reflect and question. An African friend sum­
marized our drive to understand in the words "We used to see the
sun, the stars, the clouds, the mountains, the animals and the trees,
and we enjoyed what we saw. Today we want more." What is this
more? What do we want to understand? Is the symmetry of the figures
that we see a reason for our enjoyment? Or is it the order, the under­
lying pattern that we recognize when the seasons of the year keep
 Structure and Form 99

reappearing? When did humans first realize that the moon provides
them with a calendar, that the flow pattern of the tides changes, but
periodically returns? At some time long ago, it became clear that
diversity does not mean chaos, that there was some inner order,
some rules governing the occurrence, constructed such as to give
rise to a spatial structure and to a direction of time.
In the beginning, diversity also led to the desire to order things
such that combinations "made sense," were based on common fea­
tures. What are the possible forms of matter? Is it possible to reduce
the observed states to a smaller number of basic forms? In the fifth
century BC, the Greek philosopher Empedocles defined four forms,
four elements:

�
JlVMIDVS :S-

Earth, Water, Air, and Fire

100 Before Time began

The Sicilian earth, the water of the Mediterranean, the winds

along the coast, and the fire of the volcanoes suggested these as nat­
ural forms. Not surprisingly, similar considerations also appeared in
the early Buddhist-Hindu thinking. And in both cases, the question
came up as to what the stage for these forms might be: where does
matter appear, where are its building blocks put together? In Greek
thinking, this led to a fifth form, the quintessence, empty space.
For the Indians, the "Vakasha," the nothing, played a similar role.
The different forms of matter stage their appearance in front of the
background of empty space. These concepts have not really changed
much in the course of the past two thousand years: even today, sol­
ids, liquids, gases, and plasmas are the basic forms of matter. And if
we heat a solid at constant pressure, matter passes with increasing
temperature through these four states; a plasma does so for decreas­
ing temperature. But the empty space of today is no longer just
nothinB. On very large as well as on very small scales, our concept of
the vacuum has become more complicated.
How did these forms appear, and why-why just these? Once
Albert Einstein asked, "When God made the world, did he have a
choice?" That seems to be a philosophical question, or at least a met­
aphysical one. But present-day physics can contribute to the answer,
and that is what we want to show here. Our world was not always as
it is now, and by investigating how it got to be this way, we can learn
to understand many things. The path from the Big Bang to today
evidently was one from a simple. uniform world to all the complex­
ity that surrounds us now. At one point, heaven and earth were sep­
arated, light and dark, wet and dry, and much more. How could that
happen, how could structures develop? How could the hot uniform
clouds of gas just after the Big Bang lead to a solar system of circling
planets? Is it possible that a completely disordered world turned into
one with so much structure?
The direction of the evolution of systems in our present world
is codified, written down as one of the immovable laws of physics,
as a basic law of thermodynamics. It is the thermodynamic arrow that
points toward expansion, mixing, breaking, aging. A broken glass
 Structure and Form 101

will never recombine again on its own, a scrambled egg will never
again become a whole egg, nor will ashes turn into coal. There is a
direction that so it seems, leads from structure to its absence, from
order to disorder.

The course of evolution

is thus specified in today's physics, and it defines a direction for the

progression of all physical phenomena. It took physicists quite a
while to reach this conclusion, or-to be more honest-to admit it.
The equations of mechanics or of electrodynamics don't define
a temporal direction-everything can evolve forward as well as
backward in time. However, if we look at systems consisting of very
many particles, that fact is not of much help.
To see why this is the case, it is useful to consider an experiment
carried out around 1850 by the British physicist James Prescott Joule.
Joule was the son of a brewer, and he later took over his father's
brewery; hence his interest in things like temperature and pressure
was not purely academic. In his experiment, he started with a con­
tainer separated into two compartments, with the help of a dividing
wall, and isolated from the outside world as much as possible. In one
compartment, there was a gas; in the other, nothing, a vacuum. If
Joule now removed the dividing wall, the gas expanded until both
compartments were equally filled. And he could wait as long as
he wanted-the gas in the previously empty compartment never
returned to its initial volume. The expansion of the gas was certainly
a physical process, but it never ran backward: the direction of evolu­
tion was fixed; the evolution was irrevernhle. In principle, the gas was
free to return from time to time to its starting compartment, like a
liberated animal to its cage--but that never happened.
Other features of the process are also quite interesting. Because the
whole setup was isolated from the outside, the total energy of the gas
was conserved in the expansion: the number of molecules and their
momenta were unchanged, and so also the temperature remained
the same. But the available volume was now larger, so the density
102 Before Time began

of the gas had been reduced. Also, the pressure, the total energy of
impact on the container walls, had become smaller: per square cen­
timeter, there were now fewer impinging molecules, because the
total area had grown, but the number of molecules had not.

D
The Joule experiment

The understanding of the situation required an enlargement of

the conceptual world of physics. When you study a gas in a con­
tainer, you measure its temperature and pressure. These quantities,
however, do not arise from one or two molecules, whose paths are
forward-backward, symmetric in time; they are the result of the col­
lective effort of all 1023 molecules of the gas. A few molecules might
once in a while find their way back into the initial container, but that
all the weakly connected particles simultaneously turn around: that just
doesn't happen. If only a single particle on its way back is through
some interaction deterred from its path, then the entire return col­
lapses. The investigation of individual tracks thus becomes meaning­
less. What you can determine is the average energy of a particle, for
which temperature is a measure; the average number of particles per
unit volume, that is, the density; and the average impact energy on
the walls of the container, the pressure. The reversibility of individ­
ual molecular fates becomes insignificant, their paths unimportant;
they become submerged in the mass of all, and for the evolution of
that, there is a temporal direction.
It thus became clear that the description of the collective behavior
of very many particles requires a new postulate. Something had to
be added to the previous laws of physics, which allowed time rever­
sal. Mainly through the work of Ludwig Boltzmann in Vienna and
John Willard Gibbs at Yale University in the USA, these thoughts
led, toward the end of the nineteenth century, to a new paradigm,
statistical physics. Imagine a gigantic catalog containing all the possible
 Structure and Form 103

configurations of all the 1023 molecules allowed when the overall

parameters of the system are fixed, that is, its total energy, its vol­
ume, and the number of molecules it contains. We call these overall
parameters the macrostate of the system.
A specific configuration, sometimes called a microstate, consists of
all positions and momenta of all particles possible at a given time
for a given macrostate; here we assume classical statistical physics.
The basic postulate of statistical physics now says that a system in
thermal equilibrium finds itself in principle, a priori, with equal
probability in any one of the multitude of microstates, provided
there is no interference from the outside. And that means in the
evolution of the system from one microstate to the next, in the
buzzing swarms of particles, the most common states win-unu­
sual configurations are almost never reached, and thus become
effectively eliminated. We shall see shortly that in the Joule experi­
ment, the number of configurations possible for a gas equidistrib­
uted over both compartments is so incredibly much larger than
that for the gas in the initial compartment only, that this indeed
rules out any return.
To illustrate what happens in such cases, let us look at a box with
nine compartments. If we now take four identical balls and distrib­
ute these into the compartments, one per compartment, then there
are in total 126 different configurations. In the following picture we
show three of them, and in Appendix 1 we do the corresponding
counting. What we want to emphasize here is that there is exactly
one configuration in which each ball is in a corner. If we assign to
all configurations the same probability, then the chance for such a

corner arrangement is 1:126. In other words, such very " ord ered "

microstates are rather unlikely.

Configurations of four balls in nine compartments

10 4 Before Time began

Next we want to see what effect a doubling of the size of the box
has, giving it 18 compartments. Now there are 2060 configurations
for the four balls, that is, almost 20 times more than in the previ­
ous case. And this relation between starting size and doubled size
increases more and more with the number of balls. If we start with
nine balls, we fill all the compartments, so that there is just one
configuration. The doubled volume, however, now gives 48,620 pos­
sibilities for the nine balls; the chance to return to the starting posi­
tion is 1:48,620.

Nine balls in nine and in eighteen compartments

That ratio holds for only nine balls. For a box with more com­
partments and a larger number of balls, the chance for a return
to the starting configuration decreases more and more. And if we
more than double the size of the volume, it drops even more. For
an x-fold volume growth and n balls, the number of configurations
increases as x" (again sec Appendix I). In the case of a normal gas,
23
with some 10 molecules, the opening of the new compartment
leads to the incredible increase in the number of configurations
mentioned earlier. If we assign to all configurations the same prob­
ability, then the chance to find the gas again in the initial com­
partment is reduced by the astronomic factor x100 . . . ooo, an x to the

power of a one followed by 23 zeros. In comparison to this, win­

ning the lotto (six out of 49) is an easy task: for 36numbers, there
are "only" about 14million configurations of six different numbers
between 1and 49.
The fundamental quantity for the behavior of systems in statistical
physics is thus the number of possible configurations, of microstates, for
a given total or macrostate. The measure for this quantity is the
 Structure and Form 105

Entropy.

If we know the total energy E of the system, the number N of

molecules it contains, and the volume V of the container, the
macrostate is fixed and we can calculate the number W(E,N, V) of
possible microstates. The entropy S(E,N, V) is then defined by the
famous formula

=I< fog w

Tomb of Ludwig Boltzmann in Vienna

Photo courtesy of Oesterreichische Zentralbibliothek

fuer Physik, Vienna, Austria

inscribed on the tombstone of Ludwig Boltzmann in a Vienna cem­

etery. The proportionality factor k is named after him and provides
the connection between dynamics and thermodynamics. The Boltz­
mann constant k, the gravitation constant G, the speed of light c,

and the Planck constant h then give the four fundamental constants
of physics, for thermodynamics, gravitat i on relativity theory, and
,

quantum theory, respectively. For the definition of entropy, one

uses the loaarithm (log); this has two helpf u l aspects. The number
of microstates is immense, and so the logarithm is more suitable,
because it just gives the number of powers of ten. To illustrate: the
logarithm of a billion, 1,000,000,000 = 109, is simply 9. Furthermore,
the number W of microstates for a system consisting of two parts is
106 Before Time began

the product of the two numbers, W1 and W2, W = W1 X W2. Because

the logarithm of a product is the sum of the logarithms of the fac­
tors, the entropy of the total system becomes simply the sum of the
subsystem entropies, 5 = 51 + 52. For a gas, as in the Joule experi­
ment, the number of microstates is determined by the available
volume, and so the overall entropy becomes the sum of the partial
volumes. The difference in entropy before and after removal of the
divider is thus determined by the resulting increase in volume.
The Joule experiment shows another, essential feature: a system
left to its own, isolated from the outside world, always develops
so as to maximize the number of microstates allowed, to maxi­
mize its entropy. And once it has reached that macrostate, it will
remain there forever, barring outside actions. Before the opening
the divider, that was the case: the system was in a state of maxi­
mal entropy for the initial volume; it was in thermal equilibrium.
The opening of the divider suddenly increased the available vol­
ume, so that the system now was no longer in equilibrium, because
its actual entropy at that moment was much less than that for
the larger volume. By streaming into the new volume, the gas
increased its entropy, until it finally reached the maximum value
for a gas at fixed temperature in the larger volume. The system was
now (again) in thermal equilibrium.
The two basic laws of thermodynamics summarize these con­
siderations. The first says that the eneray of the total system remains
constant, is conserved: without actions from the outside, it can nei­
ther increase nor decrease. The second law of thermodynamics
states that the entropy of an isolated system never decreases, and that in
equilibrium it attains its largest possible value, its maximum. After
that, nothing changes anymore: after all, equilibrium means that
everything remains as it is, provided nothing interferes from the
outside.
It seems worthwhile to elaborate a little further the connec­
tion between entropy and structure. To write down the previous
sentence, you need 81 letters. The English alphabet consists of 26
 Structure and Form 107

letters. So there are 2681 ways of arranging the 81 letters-in other

words, there are 2681 microstates, more than 10114• Only one of all
these produces the sentence in the given form. Structure, and here
information, corresponds to a very low entropy. Conversely, recog­
nizable structures disappear with increasing entropy-the transfer
of information ends. News can be transmitted by Morse code only
because one sends a very specific sequence of all the possible short­
long combinations. The higher the entropy, the lower the informa­
tion contained in the system.
The concept of entropy is in fact older than its explanation in
terms of the motion of atoms or molecules of the medium, as we
have given it here. The French physicist and engineer Sadi Carnot
had noticed around the beginning of the nineteenth century that
heat always flows from hot to cold, never the other way around.
He was presumably not the first to see that, but he recognized this
as one of the basic features of what was to become thermodynam­
ics. Just as the atoms in Joule's container would never stream back
into their initial compartment, a pot of water would never sponta­
neously freeze and thereby heat up the room. Twenty years later,
the German physicist Rudolf Clausius used that to define the con­
cept of entropy, thus establishing the fundamental quantity for a
theory of heat. The subsequent development of statistical physics
then provided the explanation in terms of the atomic structure of
matter.
Up to our time, the second law of thermodynamics, the entropy
of an isolated system never decreases, is perhaps the most profound state­
ment of physics. In a science fiction world, in which it is not valid,
cold water can become hot without heating, the old can become
young again, the dead arise, and much more. There are few state­
ments of physics so directly connected to the experiences of our
daily lives, few that form such a fundamental basis of all our knowl­
edge about the course of events. The famous English physicist and
astronomer Sir Arthur Eddington formulated that as advice to his
colleagues:
108 Before Time began

If someone points out to you that your pet theory of the uni­
verse is in disagreement with Maxwell's equations, then so much the
worse for Maxwell's equations. If it is found to be contradicted by
observation-well, these experimentalists do bungle things some­
times. But if your theory is found to be against the second law of
thermodynamics, I can give you no hope; there is nothing for it but
to collapse in deepest humiliation.
As we shall see, that is no empty threat.
The concepts presented so far allow us to understand the steps,
the direction in the evolution of matter. A glass that is intact has that
one state, corresponding to minimal entropy. The harder it falls or is
thrown, the more pieces there will be, so that the final entropy, the
number of produced pieces, depends on the energy put into its break­
ing. We can further increase this energy by heating. The fragments
now melt and form a liquid of glass molecules, which eventually will
evaporate and form a gas. The addition of still more energy will first
break up the molecules into their atomic constituents and then even
the atoms into nuclei and electrons, leading to a plasma of electri­
cally charged constituents and thus to still more microstates.
Based on the concepts of conventional statistical mechanics, we
thus expect that a system initially not in equilibrium-the falling
glass, its melting fragments, the gas in the initial compartment of
Joule's experiment-will rapidly evolve, so that its entropy, too
low for the given conditions, will increase to attain the maximum
value in each case: many fragments, molten fragments, the gas in
the entire volume. This evolution always proceeds from a state
of low entropy ("structure") to an equilibrium state of maximal
entropy ("uniformity"). That is the course of events prescribed by
"normal" physics. Fluctuations remain possible-some corner of
the world can accidentally be a little more structured-but on
the whole, the direction is set: larger entropy, more disorder, less
structure. On a warm spring day, a winter scene with snow, snow­
men, and icicles is no longer in equilibrium: everything first melts
and then evaporates, until all these structures have become uni­
form water vapor.
 Structure and Form 109

If then in the development (evolution) of all systems entropy can

never decrease, how can we understand that the Big Bang, starting
from a structureless hot gas, has led in the long run to the diver­
sity of our present world, with galaxies, crystals, and snowflakes? In
other words, how can the second law of thermodynamics, insisting
that the entropy of an isolated system can never increase, that order
leads to disorder, be reconciled with the apparent formation of

Structure in the universer

That problem has occupied physicists and cosmologists for quite

some time, and even today there is by no means universal agree­
ment. Here we only want to present one picture of how it could
have been. The Big Bang lies 14 billion years back, and so it is quite
understandable that the picture we have of the earlier stages of the
universe still shows some speculative features. But one thing we
should always keep in mind: Eddington's warning words.
For quite a while, the apparent dilemma between structure for­
mation in the universe and the second law of thermodynamics
was resolved in a way that predicted a rather unpleasant end of the
world. These considerations go back to Lord Kelvin, Hermann von
Helmholtz, and others around the middle of the nineteenth cen­
tury. They assumed that the universe was initially in an ordered state
of low entropy (prepared that way by a Creator? ) and subsequently
proceeded to evolve slowly but surely into an ever more uniform
state of maximal entropy. The predicted result is the "heat death"
of the universe: the whole world would end as a uniform thermal
medium, a gas without any structure and any energy-consuming
processes. In the final chapter of this book, we shall return to these
ideas. Such an end is still not fully excluded, but if at all, it will
happen only in the very long run.
Today we see the development up to the present quite differently.
There are basically two aspects that decisively influenced the evolu­
tion of the universe after the Big Bang. On the one hand, the uni­
verse has been expanding ever since then and continues to do so.
110 Before Time began

We therefore have to check if such an expansion in fact ever allows

a thermal equilibrium to be attained. It takes a certain time to reach
equilibrium, and if the expansion is fast enough, that time is not avail­
able. On the other hand, it is not unambiguously clear how a state
of maximal entropy has to look. In our world of snowmen and ici­
cles, raising the temperature above the freezing point evidently first
produces water and eventually, given enough heat, water vapor-a
uniform, disordered system of water molecules. Here the form of
the intermolecular forces plays a decisive role, and the nature of the
force between the constituents of matter is important in other cases
as well. For a medium of non-interacting or only weakly interacting
particles, the state of maximum entropy is always a disordered, struc­
tureless gas. But if the interaction becomes stronger-for example,
at lower temperatures-maximum entropy can mean crystalline
ice, with a well-defined crystal structure. Something similar hap­
pens if the particles carry electric charges, positive and negative,
in equal numbers. At high temperatures and densities, the state of
highest entropy is again a uniform disordered plasma. For dilute or
low temperature media, however, a positive and a negative particle
combine to form an electrically neutral entity, an atom; that allows
both to escape from further electromagnetic interactions. Now the
state of highest entropy is a disordered gas of such atoms. Later we
will see that for stellar clouds in the cosmos, gravitation plays a still
stronger role. Here we shall first consider in more detail each of the
two factors we have mentioned as crucial in the evolution of the
universe: expansion and interaction.
To understand the role of expansion a little better, we return to
the Joule experiment and consider what happens when the divid­
ing wall is removed. In the moment of the opening, the gas is still
completely in its initial compartment, where it was in thermal
equilibrium, that is, in a state of maximum entropy. The opening
suddenly provides access to a new and empty additional compart­
ment. The gas begins to flow rapidly into this compartment and to
fill up the empty volume. After a certain "relaxation time," both
compartments are filled up, and as soon as the entropy has reached
 Structure and Form 111

its new larger value, corresponding to the larger volume, thermal

equilibrium is once more attained. In the time between the opening
and the re -establishment of maximum entropy, during the relax­
ation time, the system is not in equilibrium; its entropy is larger
than before the removal of the divider, but less than the allowed
maximum. During the entire process, the entropy is increasing,
but it does not reach its new maximum immediately; that happens
only after the relaxation time is over, and from now on, everything
remains stable. The temporal behavior of the entropy is illustrated
in the following picture.

maximum entropy ----i

actual entropy

divider closed divider open

Time

- relaxation time -

The temporal evolution of the Joule experiment

During the relaxation time, the medium is not without struc­

ture--on the one hand, it consists of partially empty, partially
gas-filled regions; and on the other hand, the gas-filled region is
expanding, moving toward the opposite container wall. Shortly
after the opening of the divider, the molecules that have entered the
new volume are not moving around uniformly in all directions­
they are moving mainly toward the opposite wall, the faster ones
in front, the slower further back. So there clearly is some kind of
"order." In the course of the relaxation time, this order disappears
and finally all molecules are equidistributed over the entire volume;
112 Before Time began

the entropy has reached its new maximum value, and the system is
in equilibrium once more. An intermediate state, in which equilib­
rium is not yet attained, is shown in the following drawing.

Molecules in the Joule experiment shortly after the removal of the divider

What's happening there has in recent years led to some remark­

able considerations. It seems as if some mystic "entropic" power is
driving the molecules from left to right. A single molecule would
simply fly around-it would be in either section of the container
with equal probability. It would not experience this power, which
therefore must be a collective, multi-particle effect. In the following
chapters, we shall return to such emerBent forces, which arise only from
the interplay of many individual constituents.
In the Joule experiment, the increase of the maximum possible
entropy was achieved by the removal of the divider, which suddenly
created a larger accessible volume. But we can also do this more grad­
ually, by pulling out a piston (for simplicity without any friction),
and thereby increasing the volume. Here it is crucial that we pull and
not have the gas pressure push: that would cause the gas to do work
and reduce its temperature. What we have in mind is achieved if the
piston is very heavy, so that the gas pressure cannot move it.

The continuous Joule experiment

 Structure and Form 113

In such an experiment there are two rates of change: the relaxa­

tion time, which the disturbed gas needs to return to equilibrium,
and the rate at which the volume is increased by pulling out the
piston. If we move the piston very slowly (adiabatically, in physics
terminology) , the gas can adjust to the changing conditions: it
remains always in equilibrium; the increasing entropy always stays
at its (increasing) maximum. But if we retract the piston very rap­
idly, the gas cannot keep up: it falls more and more out of equilib­
rium; it becomes ever more "ordered," the molecules frantically
try to reach the disappearing opposite wall. Instead of a uniform
medium, we now have oriented flow, attempting to reach the
retracting piston.
In other words, the crucial competing phenomena are

Expansion vs. relaxation time.

If the accessible volume increases faster than the relaxation time

needed for equilibration, then the maximum entropy can never
be attained. On the contrary, the difference between actual and
maximum entropy continues to increase with time--the system
shows more and more structure. This phenomenon-the role of
an expanding volume in the context of the second law of thermo­
dynamics-was proposed around 1975 by David Layzer of Harvard
University as a possible source of structure formation in the uni­
verse. If you increase the available volume for a gas in equilibrium,
then the rate of expansion and the relaxation rate form two critical
opposing factors. If the expansion rate of the volume is slow enough,
then the system has sufficient time to reach maximum entropy, to
achieve thermal equilibrium again. If that is not the case, the maxi­
mum entropy increases faster than the actual momentary entropy
of the system. It therefore deviates more and more from equilib­
rium; it develops more and more order and structure. We should
emphasize that all this is fully in accord with the second law of ther­
modynamics: the entropy is indeed continuously increasing, but not
fast enough to reach the maximum value for the given volume. The
114 Before Time began

resulting discrepancy means order. Requiring an increase of entropy

thus does not always mean an increase of disorder, a reduction of
structure, as sometimes claimed.
We should further underline that these considerations are
schematic and hence certain important details are neglected. The
expansion reduces the density in the given volume, which in turn
generally increases the relaxation time; as a result, the actual entropy
grows slower, and order and structure formation faster-at least, if
the expansion rate is kept constant.
The cosmic result of our considerations is quite evident: start­
ing from an extremely hot gas in equilibrium at initial state, the
expansion leads in the long run to a discrepancy between actual
and maximal entropy and thus creates structure and order in the
medium. In this sense, complete disorder means that the system
is in a macrostate, for which all microstates are equally likely and
whose entropy is thus at a maximum. In the case of order, some of
the microstates are completely excluded, and some have a higher
probability than others. In terms of entropy, order or structure R is
thus defined by the relation

R = Smax - S.

The order of the system vanishes if the entropy reaches its maxi­
mum possible value. Then all states are equally probable--no
particular structure is selected. If R is not zero, we have some
kind of order or structure. What its specific form is depends on
the nature of the interaction between the constituents of the
medium, in particular its density and temperature. It may seem
surprising that disorder is more generally defined than order­
that the latter can be defined only once you know the former.
But order can appear in many forms, to larger or smaller extent,
as rain, hail, or snow, whereas disorder just means that all micro­
structures are treated equally.
At this point, we also note that if we carry out the ultimate Joule
experiment and simply let the initially confined gas escape into
 Structure and Form 115

empty space, then the volume available after removing the confining
wall becomes infinite, and so does the maximum possible entropy.
The actual entropy of the expanding gas increases with time, but it
always remains finite. The order as defined earlier can thus never
vanish: the system can never achieve a maximum entropy; it can
never reach thermal equilibrium.
That brings us to the second aspect mentioned earlier: how does
the form of the interaction between the constituents affect the state
of maximum entropy? We begin by noting that for gases and similar
rather weakly interacting media, volume is the decisive factor. If we
perform a Joule experiment and simply remove all walls of the con­
tainer, letting the gas expand into empty space, then everything flies
apart: now there is no maximum entropy-that would be infinite,
just like the volume now available. In such gas-like cases, the value
of maximum entropy is thus defined through the size of the system,
determined "from the outside." What happens if we allow much
stronger interactions?
Here a classical example is gravity, which has two extraordinary
properties: its range is infinite, and it is always attractive. Electro­
magnetic force is in principle also of infinite range, but there are
positive and negative charges, with equal charges repelling, and
unequal charges attracting each other. That means a cluster con­
taining equal numbers of positive and negative charges appears
to be neutral, uncharged, if observed from far away. The different
charges compensate each other's effects, so that the whole cluster
does not create an electric force. In contrast, a cluster containing
n particles of mass m has a total mass M = nm, and this total mass
determines the force of gravity exerted by the cluster-there is
screening of charges, whereas gravity is additive. For that reason,
gravitational effects of distant bodies can never be completely
neglected. And whereas a cluster of equal electric charges tends
to be driven apart, a cluster of masses is always pulled together,
contracted. So, in a way, gravity itself determines the size of the
system, and the external "box" becomes less unimportant.
116 Before Time began

Up to now, we have considered energy and volume of the sys­

tem as two independent parameters, both being determined from
the outside by the experimenter. Such a picture now finds its lim­
its. According to Albert Einstein's seneral theory of relativity, the theory
addressing the role of gravitation, the presence of mass deforms,
curves the space in which it is contained; more generally, that holds
for the presence of energy as well. This means that an equidistri­
bution of many masses in space no longer needs to be the state of
maximum entropy. Imagine a number of marbles, equally distrib­
uted in a plane, here on Earth. If we now create a dip in the center
of the plane, the marbles will roll into that hollow. In the plane, an
equidistribution provided the state of maximum entropy, but in the
presence of a dip, a pileup at the deepest point constitutes such a
state. If we start from a set of masses initially distributed uniformly
in space, the effect of gravity will have them contracting to a cluster
much smaller than the starting size of the box. The size of the pile
of marbles, the density of packing, is determined by the size of the
individual marbles. The same holds true for the "gravitating" gas,
for which the size of the individual masses determines the size of the
star or the cluster of stars. In addition, there are of course effects due
to motion of the constituents.

• •
• •
• •
• •
• •

The effect of space curvature on the distribution of masses

The structure of the universe on a cosmic level is very largely a con­

sequence of gravity. And in this case, the hot uniform gas, just after
the Big Bang, is, in the long run, after expansion, no longer in the
state of highest entropy. Such a state is achieved through
 Structure and Form 117

Cluster formation,

so that it is precisely the second law of thermodynamics that pre­

vents the classical heat death of the universe. In other words, the
evolution of the universe proceeds indeed from lower to higher
entropy; but for a medium with gravitation, the uniform gas is a
state oflower, and the cluster pattern a state ofhigher entropy. That
constitutes the crucial difference from the short-range gases of the
usual thermodynamics.
In our context, we can then picture the start as a system of total
mass M (i.e., of total energy Mc2), consisting of individual masses
interacting through gravity. According to Einstein's general theory
of relativity, such a system, even if it begins as a uniform gas in a box
V, will soon determine its own fate and contract to a cluster of much

smaller size. This size can be calculated, and it is much smaller than
the starting volume. Up to now, we have considered energy and size
of the system as two separate entities. In the case of gravity, that ends
at some point-if the total mass becomes large enough, it determines
the spatial extension of the system itself. And in this case, the state of
maximum entropy is a distribution ofclusters, a world ofstars, no mat­
ter what the initial size was. In Appendix 2, we shall consider the under­
lying physics in a little more detail. We shall see in particular that in the
presence of gravitation, an equidistributed set of masses will converge
to something like a solar system in order to maximize its entropy. The
masses will then circulate around a center in a way to compensate
attractive gravitational force by repulsive centrifugal force.
We should note that this does not imply that the entropy ofa hot
gas is lower than that of the star into which it eventually collapses;
the entropy of the star is in fact lower. But in the formation of the
star, during the collapse, part of thl' total availabk mass is converted
into radiation, photon emission. And thl' l'ntropy of that photon
cloud is in fact much higher than that of the initial hot gas, because
the density ofthe cloud is much lower and hence the volume ofthe
cloud is larger. So the entropy-increasing transition is one from a
118 Before Time began

uniform gas of matter particles to a compact dense star immersed in

a very much larger cloud of photons.
The general framework for the development of the universe after
the Big Bang is thus provided: expansion can cause a world with
structure, not in equilibrium, and even in equilibrium, long-range
interactions such as gravity make a distribution of stellar clusters the
state of highest entropy.
Some 500 million years after the Big Bang, we thus find a world in
which the tiny irregularities created from earlier quantum fluctua­
tions have led to cluster formation. Clouds of gas float through a
now relatively empty space, collide and unite to form larger clouds
or break up into smaller ones. That leads to some important fea­
tures. When two clouds collide, it generally causes a rotation of the
connected system, so it is not surprising that most galaxies today
have a disk-like structure, with a higher density in the center.

Formation of a galaxy in the collision of two gas clouds

The dynamics of galaxy formation remains not fully understood

even today. Often spiral-like arms directed outwards form at the
edge of the central disk. If within a galaxy two gas clusters collide,
the impact region provides a zone of extreme energy, high enough
to result in nuclear fusion, which in turn leads to radiation. In this
way, shining stars appear, and each galaxy consists of innumerable
stars. The dimensions of galaxies surpass all imagination: in the case
of our own, the Milky Way, the average thickness is about 3000 light
years, with a diameter of some 100,000 light years; it contains more
than 100 billion stars. And in our visible universe there are billions
of such galaxies, which again combine to form supergalaxies. Again,
the dominating pattern seems to be self-similar, with structures of
the same kind at all scales.
 Structure and Form 119

Spiral galaxy Messier 101, Photo courtesy of ESA and NASA.

Perhaps the picture of Messier 101 can convey a little of the beauty
and the infinity of such structures. Our home galaxy, the Milky Way,
is also of this type, and our sun, the center of our solar system, is in
one of the spiral arms, near the edge of the actual center.
We thus find some similarity between a galaxy and a solar sys­
tem such as ours. In a galaxy, stars instead of planets circle around
a center, which however is not determined by one single object. In
our solar system, the mass of the sun is a thousand times larger than
the masses of all planets combined, so it makes sense to speak about
planets circling around the sun. In the case of the galaxy, the center
is-from the point of view of our sun-determined by all stellar
objects within a sphere drawn by the solar trajectory. That includes
not only all stars in that region, but also an extraordinarily massive
black hole at the galactic center; its mass alone is about 1 % of the
total Milky Way mass.
The sun then encircles the center of the Milky Way with a veloc­
ity of about 200 km/ s; that means it needs some 200 million years for
a complete trajectory. It is estimated that the sun is some five billion
120 Before Time began

years old, so that it has made the complete orbit about 25 times. And
our Earth takes part in this voyage. We have already noted one effect
of this motion: because the Earth moves with such a speed relative
to the center of the Milky Way, the cosmic microwave background,
determined by all space, is correspondingly Doppler-shifted when
seen from our spaceship Earth. The picture on page 81 shows the
direction we are traveling through space.
We mentioned the size of galaxies. If we can estimate the number
of stars in a galaxy and the mass of a typical star, then we know the
mass of the total galaxy. The number of stars is obtained by dividing
the light output of the entire galaxy by that of a typical star. From
that we conclude that the Milky Way consists of about 100 billion
stars and hence must have a mass of about 100 billion solar masses.
Using that information, the law of gravity tells us how long it must
take any given star for one complete trip around the Milky Way.
Unfortunately, the result turns out to be very wrong-as we shall
see in the next chapter. Although "seeing" is perhaps not the right
word here, there is more in the universe than what we can see.
 7

Dark Corners

There was a Door to which I found no key,

There was a Veil past which I could not see.

THE RuBAYIAr OF OMAR KHAYYAM

(TRA�SL. EDWARD FITZGERALD)

Dark corners have always been around in physics, and most likely
always will be. But often the key to understanding is not found under
the bright light of the lamp, so we have to search in the unlit dark part
of the world. Dark corners are thus on the whole quite fruitful: they
are nature's way of telling us to keep looking and to keep thinking.
Quite often the continuing search takes a very specific form.
When Rutherford proposed a heavy positive nucleus surrounded by
light negative electrons as the model of the atom, the total mass of
the protons did not nearly suffice to give the mass of the nucleus.
Something was wrong. The way out, Rutherford proposed, was to
introduce new, heavy but neutral constituents that, together with
the protons, would provide the measured mass of the nucleus. The
new constituents simply could not be "seen." Twenty years later, his
student Chadwick identified these neutrons experimentally.
The neutron itself continued in this vein. In an isolated state it
was not stable, but decayed into a proton and an electron, keeping
the overall charge zero. The masses and the kinetic energies of pro­
ton and electron could be measured, and when they were added,
the sum was less than the mass of the neutron. Again something
was missing. This time it was Wolfgang Pauli who blamed the deficit
on another invisible particle, the neutrino. The name--little neu­
tron-was invented by Enrico Fermi, and some twenty years after
Pauli's proposal, it was indeed experimentally identified.
122 Before Time began

Ernest Rutherford (1871-1937)

Looking at the universe, we encounter even now several such
dark corners, which must hide something that cannot be seen, but
established only through its effect on the surrounding world. One
such phenomenon we have already mentioned:

Black holes.

From our point of view, a black hole is basically what remains of a

sufficiently massive star when it is completely burned out. A star
acquires its size from the balance of gravitational attraction, com­
pressing the mass, and the opposing pressure of the heat created
by nuclear fusion. When there is no more nuclear fuel, the stellar
fire is extinguished and the heat pressure that had compensated
the gravitational force of attraction is no longer available. The star
now collapses, undergoing several different reaction processes, until
it consist only of neutrons. These are, as we mentioned, territorial
particles that don't like to be compressed-the resulting Pauli pres­
sure limits the possible shrinkage and leads to the formation of a
much smaller stellar object, a neutron star. That is what happens
provided the overall stellar mass is not too large (less than some five
solar masses), because otherwise gravity is able to overcome even
the Pauli pressure, compressing the neutron star further and fur­
ther and thereby creating a completely new object: a black hole.
The essential property of such objects is that their force of grav­
ity is sufficient to prevent even light from escaping their domain of
attraction. On the surface of the Earth, we have the so-called escape
 Dark Corners 123

velocity, the velocity we have to give a bullet so that it can over­

come the attraction of the Earth and escape into outer space. It is
proportional to the mass of the Earth and inversely proportional to
the radius of the Earth. We can therefore imagine a stellar body suffi­
ciently heavy and sufficiently small that the escape velocity becomes
greater than the speed of light; then even light can no longer get
away. Such an object retains simply everything: it is a black hole
from which nothing can ever escape.
More than two hundred years ago, an English priest and natu­
ral scientist, John Michell, predicted that such objects should exist,
using arguments similar to the ones we just presented. The famous
French mathematician Pierre-Simon de Laplace provided a math­
ematical proof not much later. Today we are sure that such black
holes indeed exist in our universe. One can never see them, but it is
possible to observe their effect on the surroundings: if somewhere a
shining beam of light suddenly disappears, a black hole has eaten it.
The well-known English mathematician and writer Lewis Carroll,
immortal through his Alice in Wonderland, has given a beautiful pres­
entation of such a phenomenon in his epic poem The Hunting of the
Snark. An expedition sets out to find the mysterious creature Snark,
of which there are two species, common Snarks and Boojums. The
latter have a horrible property: whoever sees one, immediately dis­
appears, dissolved into thin air. How can they then be discovered? To
find the solution to this dilemma-and not only for that reason-it
is really worthwhile to read the story.
Things like that indeed exist in the universe. There are twin stars,
orbiting around each other. If one dies and becomes a black hole,
the material of its partner, still shining, is sucked into the black hole
accompanying it. The twin star system Cygnus X-1 seems to be of
this nature.
The black holes created out of dying stars are very massive, some
5--10 solar masses or more, but also very small: for a black hole of ten
solar masses one obtains a radius of about 30 kilometers, while a shin­
ing star of that mass has a radius of more than a million kilometers.
124 Before Time began

Schematic view of the Cygnus X-1: on the right is the black hole, on the left
its still shining companion being sucked in. Photo courtesy of ESA and NASA

We believe today that besides stellar black holes, there is yet

another form of such entities: the so-called supermassive black
holes, with masses up to a billion solar masses. These monsters are
found in the centers of most galaxies; they could have formed in
the early stages of the universe, when gas clouds first contracted
to very massive, star-like structures. Collapse would then lead to
intermediate-sized black holes, which gobbled up all stars around
them to reach their present size. We arc quite sure today that such
a monster galaxy resides in the center of our Milky Way. We will
shortly see how such a conclusion can be drawn.
One of the greatest discoveries in physics is Isaac Newton's con­
clusion that the same force, the force of gravity, determines in a uni­
versal way the attraction between all masses. That holds from falling
apples on Earth to planets orbiting the sun and beyond, and for the
motion of galaxies.
The strength of the force of gravity is specified by a universal
constant G, Newton's constant, that determines how strongly two
given masses separated by a given distance attract each other. If we
know the radius of the Earth, we can use Newton's law of gravity
 Dark Corners 125

to "weigh" the Earth, that is, determine its mass-details are given
in Appendix 2. Applied to the moon, Newton could use his law to
calculate that in balance between earthly attraction and centrifu­
gal force, it would need 28 days to orbit once around the Earth.
Newton's law implies (see again Appendix 2) that the square of the
orbital velocity is proportional to the distance between Earth and
the moon. The weight of the moon is irrelevant for this-any satel­
lite positioned on the lunar orbit would also need 28 days for the trip
once around.
The same law can be used to calculate the orbit of the Earth
around the sun, or the orbit of any other planet. In this way one
obtains a rule the German astronomer Johannes Kepler had already
derived before Newton: the square of the orbital velocity of a planet
around the sun is inversely proportional to its distance from the sun.
We can use Kepler's rule either to determine how long a year is, or­
if we know that already-to "weigh" the sun.

Sun

Saturn Jupiter Mercury Mars

Neptune Uranus Venus Earth

Planets orbiting the sun

And we get a universal curve connecting the orbital velocity of

any planet to its distance from the sun. For the determination of
this curve the mass of the planet is unimportant. To draw the curve
we only have to know Newton's famous G and the mass of the sun.
The weighing process just mentioned led to 2 x J0-10 kg, and the cor­
responding curve is shown in the following picture, labeled Kepler's
rule. We can now get from the astronomers the orbital velocities
of the different planets and their distances from the sun and enter
them in our picture. It is evident that all planets in our solar system
126 Before Time began

fit extremely well into the scheme of Kepler and Newton. Even
Pluto fits, although it has just lost its status as full planet and is now
a labeled dwarf planet.

50 Mercury

� 40 Kepler's Rule
E
� Venus
M5= 2x1030 kg
�
·c:; 30 Earth
0
Qi
>
Mars
]i 20
15
0 Jupiter

10
--2���� •
Neptune
•
Pluto

2 4 6
distance from sun (billion km)

Orbital velocities versus distance from the sun for the planets of our solar
system, compared to Kepler's rule

Given this information, we can now come back to the giant black
hole at the center of the Milky Way. One observes a star there (52
in the astronomical notation) that orbits an invisible center of our
galaxy at a distance of about I 010 km every 15 years. That implies that
the responsible gravitational ccnter must have a mass of more than a
million solar masses. The observed orbit requires that this immense
mass must be contained in a volume of about the size of our solar
system: over a million solar masses in a volume of our planetary
world. No known astronomical structure, other than a black hole,
can lead to such a large mass in such a small volume.
Black holes thus become for us a first instance of a dark, unattain­
able corner, whose darkness we can never illuminate. We can see
what effect they have on the world around them, but we can never
find out what happens in their interiors. Nevertheless, in compar­
ison to what's coming next, they are still in a way quite familiar.
 Dark Corners 127

They are a result of the best-known force of our world, gravity, and
since Michell and Laplace we can imagine how they can be formed, as
a consequence of the force of gravity acting on matter as we know it.
Such considerations reach their limit, however, if we want to
understand the large-scale behavior of galaxies. As already noted
several times, that leads to

Dark matter.

In the following, we want to look at the reason for its appearance in

more detail; but for better understanding, it seems helpful to start
with the conclusion. A galaxy consists of millions or even billions of
stars, which give the system a certain mass; a small part (less than
1 %) is provided by the giant black hole in the center, the rest by the
multitude of stars. The mass of the galaxy determines the motion of
its stars or groups of stars, as specified by the law of gravitation: the
attraction of the mass must compensate precisely the centrifugal
force on the orbiting stars, so that the whole system is kept together.
And the mass of all visible stars in our galaxy is not nearly sufficient
to explain the motion of stars at the edge of the galaxy. As early as
1933, the Swiss astrophysicist Fritz Zwicki concluded that every gal­
axy, including ours, the Milky Way, must contain five to ten times
more invisible mass, to account for the behavior of the stars at the
edges: there must be very much invisible dark matter. It interacts with
the rest of the world only through gravity; for all other forces, it
does not exist. Just as with the nucleus and then with the neutron
and its decay, we again encounter a situation in which we sec that
something is missing; there must be more than we can see, but we
don't know what. Let's look at this problem now in a little more
detail.
In the interior of a galaxy, stars circle around a center, which
also contains the giant black hole mentioned earlier. The farther
we move away from the center, the more stars are contained in the
sphere defined by our position, and therefore the bigger is the mass
that determines the orbit of a star on the surface of this sphere, and
128 Before Time began

correspondingly the greater is its orbital velocity. Kepler's rule gave

us the orbits of planets around the sun, whose mass is fixed. As a
result, the orbital velocity decreased with increasing distance from
the sun. But now, in the case of a galaxy, the relevant mass increases
with distance from the center, proportional to the enclosed volume:
the effective "sun" becomes more and more massive. As a result, the
orbital velocity now increases with the distance from the center.
This increase stops when we reach the edge of the galaxy. The
enclosed mass now no longer grows, so that Kepler's rule in its
original form becomes operative: with increasing distance, the mass
remains essentially constant, and so the orbital velocity for more
distant, outlying stars should decrease, just as it does in our solar
system. For such outlying stars, the bulk of the galaxy is something
like a sun, which keeps objects outside the densely populated inte­
rior in orbit through its force of gravity. We can determine the
effective galaxy mass through the mass of the visible stars, and then
we can predict the orbital velocity of an outlying star in terms of its
distance from the center. The situation is thus completely analogous
to the case of the sun and the planets, with the bulk galaxy as sun
and the outlying stars as planets.
The resulting prediction is shown in the following picture. Com­
paring that to the measured orbital velocities of outlying stars, we
conclude that the prediction is completely wrong: the orbital veloci­
ties of the outlying stars remain constant, independent of their dis­
tance from the center of the galaxy. Planets close to the sun orbit
faster than those far from the sun; hut here all outlying stars, up to
distances of five to ten times the size of the galaxy, all have the same
orbital velocity. How is that possible?
If the orbital velocity does not decrease with increasing distance
from the center, then apparently the force of gravity involved must
increase. However, the visible mass remains constant, and in any
case it is too small. So there must be an additional, invisible form
of matter in which the entire system is embedded. And the total
mass of this dark matter must increase linearly with separation from
 Dark Corners 129

outlying stars
•
•
•
•
•
•
•
Kepler's Rule
e inner stars
•

distance from center

The orbital velocity of stars in a galaxy as function of their distance from

the center, as measured and as predicted by Kepler's rule

the center; that implies that its density decreases with the square of
the separation distance (see Appendix 2). For each outlying star, the
mass responsible for its gravitational attraction is then given by the
shining mass of the galaxy and in addition by the dark mass con­
tained in a sphere of the size of its orbit. And that means the visible,
shining component of a galaxy contributes only in small part to its
total mass-five to ten times more remains invisible. Each galaxy
we see in the sky is embedded in a much larger cloud of dark matter.

The cloud of dark matter around a galaxy

130 Before Time began

So today we can picture a galaxy as consisting of three compo­

nents. In the center, there is a supermassive black hole, around
which with increasing distance an ever-growing number of stars is
orbiting. These orbits define a sort of disk, like a throwing discus,
just as the planets in their orbits define a kind of plane around the
sun, with some fluctuations. This galactic discus in turn is embed­
ded in a sphere of dark matter, which assures that the outlying stars
circle around the center with constant orbital velocities.
While the black holes still are due to gravity effects on normal
matter-that is, effects on the compression of neutrons-dark mat­
ter brings us to the limits of our wisdom. What is it made of? The
search for an answer to this question is today probably one of the
central topics of particle physics as well as of cosmology. In other
words: we don't yet know.
We can, however, determine with measurements of increasing
precision that dark matter does not consist of any known species of
particles. It neither emits nor absorbs light, nor does it show any kind
of interaction with our visible world, except through gravity. It sur­
rounds each galaxy like a huge cloud, whose density, as mentioned,
decreases with the square of the distance from the galactic center.
The most p o pular speculation today is that this cloud consists of
unknown, weakly interacting and very massive particles (Weakly
Interacting Massive Particles = WIMJ's). Millions if not billions of
such WIMPs would have to be contained in each cubic meter of our
present world-but because of their extremely weak interaction,
they remain (so far) not detectable. The experimentalists at CERN
in Geneva try to produce theses ghosts in high-energy collisions,
and the theorists try to find suitable candidates in their supersym­
metric theories. But, as mentioned, so far they are all still searching.
The experimental search is once more based on the scheme that
led to the discovery of neutron and neutrino: something is missing
in the observed picture. This approach was perhaps first illustrated in
one of the stories of Arthur Conan Doyle, which deals with the theft
of a race horse and the murder of its trainer. Sherlock Holmes points
out "the curious incident of the dog in the night-time." When the
 Dark Corners 131

accompanying Scotland Yard detective replies that the dog did

nothing in the night-time, that it did not bark a single time, Holmes
notes that just that was the curious incident. Something like that
must happen with the WIMPs: if a WIMP is really produced in a colli­
sion, it will never be seen in the detectors, due to its extremely weak
interaction. But the production of a very massive energetic particle
leads to a strong recoil, and that in turn produces a jet of many nor­
mal particles, as shown in the following picture. So we have to look
for a jet of many particles, whose recoil partner is missing; we have
to find such unbalanced "mono jet" configurations.

?
.
(:;,.
11

proton 11
proton

known secondary particles

Something must balance the recoil of the observed known secondary

particles: a WIMP?

Since the WIMP mass is expected to reach up to 1000 proton

masses, such a search requires very high collision energies; per­
haps those of the large hadron collider in Geneva are sufficient. The
WIMP search is in fact one of the most pressing topics at CERN, and
a success would be a dramatic breakthrough. But, unfortunately,
there are other, more mundane possibilities that could account for
the lack of a visible recoil-witness neutrinos.
Theorists hope, as we have indicated, that among the many par­
ticle species predicted by supersymmetric models, one might find
some particles that could be the constituents of dark matter. The
mathematical structure of these theories fascinates many theorists,
132 Before Time began

but mathematical beauty alone is not enough, as we well know.

The cycles and epicycles of the Ptolymean geocentric picture of the
solar system provided a complex, mathematically very interesting
scheme; moreover, it made predictions that were extremely well
confirmed experimentally. Nevertheless, it was eventually replaced
by the much simpler Copernican formulation. So we can only wait
and hope.
Besides black holes and dark matter, there is yet another feature
that in spite of much research has remained even darker and more
mysterious:

Dark energy.

It appeared already at the beginning of our considerations, as the

space energy remaining in our universe after the bubble in the mul­
tiverse burst. Albert Einstein introduced such a quantity 100 years
ago in his general theory of relativity, with quite similar arguments.
It seems worthwhile to look a little closer.
Newton's mechanics describes the force of gravity between two
masses, but it does not say how and why this force comes into action.
Einstein therefore went one step further: he explained the effect of
the force as a consequence of the structure of space. The Earth cir­
cles around the sun, according to Newton, because the force of grav­
ity of the sun just balances the centrifugal force experienced by the
Earth in its motion, just as we can tie a stone to a string and make
it circle around us. Einstein tried to understand the nature of the
string and came to the conclusion that such a string does not exist.
Instead the presence of the sun deforms space, just as a heavy ball on
a soft surface causes an indentation, and the Earth then rolls in this
bowl, circling around the sun-see the following picture.
These considerations led Einstein to his famous equations relat­
ing space, time, and matter. In simplified form (Appendix 3 contains
a few more mathematical details), one can summarize these equa­
tions as
 Dark Corners 133

The orbit of the Earth around the Sun as caused by space deformation

where GE is a mathematical quantity (the Einstein tensor) describing

the deformation of space, and TM specifies the contents of space (the
density of energy and matter as well as the resulting pressure). Since
space and time lead to several components (x,y,z;t), the previous
expression is an abbreviated form of several equations. The Ameri­
can physicist John Wheeler, who also coined the term "black hole,"
summarized the statement of these equations as

Matter tells space how to curve and space tells matter how to move.

An essential quantity determined by these equations is the scale for

measuring the separation of distant objects, like galaxies: the so-called
scale factor a(t) , which can vary with time if the universe expands or
contracts. And the solution of the Einstein equations indeed led to a
world which in general was not static. The scale factor could increase
or decrease with time, and its rate of change could in turn also change
with time. However, the Einstein equations
in their original form predicted a continu­
ous decrease of this rate.
Before Hubble's discovery, the idea of
an expanding universe did not fit into the
accepted view of the cosmos: the universe
was considered as eternal and unchange­
able. The equations of the form given
above did not allow this; they led to an a(t)
increasing or decreasing with time, that is,
to a finite rate of expansion or contraction Albert Einstein (1879-1955)
134 Before Time began

v(t) = Lla(t)/Llt. Moreover, they implied that even this rate would
decrease with time b(t) = Llv(t)/Llt'1'0; nothing remained eternal and
unchangeable.
To alleviate this problem, Einstein added a further term to his
equations, the cosmolosical constant A:

GE= TM+ A.

It was intended to compensate exactly the effects responsible for

contraction or expansion, such as the attraction of gravity, and in
this way allow an eternally constant universe. The problem this led
to, however, quickly became evident: A was supposed to be a univer­
sal constant, while all other terms in the equations varied with time.
That implied that the desired compensation could take place only at
one specific time. If the energy density in the universe should subse­
quently change a bit-for example, due to galaxy formation-the
compensation would be destroyed and everything would move
again. For more details, we again refer to Appendix 3.
On the left side of the last equation we have Einstein's curvature
tensor, which is determined by the effects on the right side, the pres­
ence of energy and matter in space. The addition of the cosmological
constant implies that not only the matter and radiation contained
in space play a role in this, but that space itself contains a univer­
sal "vacuum energy," constant in space and time. It only affects
the structure and the evolution of space and has no further conse­
quences. In our present view, this dark space energy is just what is
provided by Einstein's cosmological constant.
Even the dark energy cannot, as we have mentioned, make the
universe static, but it can in fact decide what will happen. The pre­
sent frame is determined by two measurements. The precision data
of the cosmic background radiation indicate, as we saw in Chapter 5,
that space itself, the stage for all happenings, is intrinsically flat.
What curvature there is must have arisen from the contents of
space. Moreover, the investigations of supernova explosions car­
ried out some 20 years ago have shown that the universe not only
expands, but that its rate of expansion is in fact increasing-the
 Dark Corners 135

expansion is accelerating. Such an acceleration can be achieved with

a suitable A; for details, see again Appendix 3. The Nobel Prize in
2011 was awarded to the scientists responsible for this discovery, Saul
Perlmutter, Adam Riess, and Brian Schmidt.
The presently ever-growing expansion of the universe thus
requires the presence of the dark space energy, which uniformly
permeates the entire space and whose density remains constant in
space and, as far as we know today, in time. This density is extremely
small-we shall see shortly how small. Shortly after the Big Bang,
the effective volume of the universe was still small enough that
gravity could overcome the expansion force of the dark energy. The
universe expanded, but its rate of expansion decreased. At some
point, however-cosmologists think some 500 million years ago­
the volume had become so large that gravity could no longer keep
up. Now there was so much space and hence also space energy that
the expansion rate began to increase again, and it continues to do so
today. Because the amount of matter in the universe remained con­
stant, it constitutes today only about a quarter of the total energy of
the universe--the remaining three-quarters consists of structure­
less space energy, as we showed in Chapter 5.
The analysis of the measurements mentioned previously shows
that the dark space energy has a value of about

3 3
10-44 GeV/frn = 10 GeV/rn .

In terms of the units used, the radius of a proton is about one fern­
torneter (frn) and its mass one giga electron volt (GeV). The vacuum
energy of the universe thus corresponds to roughly ten protons per
cubic meter, or to a density that is by a factor 10-44 less than that
inside a proton. In yet other words: an empty space of the size of
the Earth contains as much vacu urn energy as a gram of water. For
this reason, the dark energy becomes relevant only for dimensions
of intergalactic size. But then it becomes essential.
We're already stumbling in the dark, as far as dark matter is con­
cerned. For dark energy, things become even worse. If it would sim­
ply be zero, we could just say that empty space is empty after all.
136 Before Time began

But to understand or explain this specific small value different from

zero-that is today perhaps the most serious issue for cosmology.
There are various ways to arrive at an energy for the vacuum. In
Chapter 5 we saw that the physical vacuum entered the scene when
the quarks combined to build hadrons and thereby allowed the exist­
ence of empty space. Before this time, all space was densely filled
with colored quarks and gluons. In the simplest models, the mass
of a hadron is a combination of the kinetic energy of its quarks and
an inherent energy of the empty hadron "shell" or bag: this is the
energy of a vacuum bubble of hadronic size. Hadron spectroscopy
3
leads to an estimate of about 0.2 GeV/fm for the energy density of
this bubble, so that about four-fifths of the mass of a proton is given
by the kinetic energy of its three quarks and the remaining fifth by
the vacuum energy of the empty bag. This gives us an idea of what
3
the energy density of the vacuum could be: 0.2 GeV/fm look at the
value determined from the expansion of the universe shows us that
33
the hadron bag value is too large by a factor of about 10 . Experts in
astro-particle physics have extended these estimates to earlier stages,
and find that the vacuum energy due to quantum fluctuations at
the Planck scale is too large by a factor of 10120. With some pride
they conclude that this is the largest discrepancy between theory
and experiment ever encountered in physics.
We can thus justly say that more than three-quarters of what­
ever is contained in our universe consists of unknown constitu­
ents. Some 5% of the contents are known particles of the world we
understand; some further 20% is dark matter, perhaps somewhat
similar particles we don't yet know. And three-quarters finally con­
sists of dark space energy, of which we don't really understand any­
thing, except that it drives the expansion of the universe. Neither
its nature nor its strength can be determined in the framework of
our present theories. This clearly leaves more than enough work for
future generations of physicists and cosmologists. And the conclu­
sion that more than three-quarters of the universe is of a completely
unknown form constitutes yet another form of Copernican revolu­
tion. What we know and understand makes up only an extremely
small fraction of all there is.
 8

The End of Time

Yes, God murmured, it was a good play,

I will have it performed again.

BERTRAND RussELL, A FREE MAN's WoRSHTI'

The end of time is, from a human point of view, as difficult to imag­
ine as its beginning. Time as such is really an essential part of our
existence; we can think of different spatial worlds, but in all of
them time always seems to flow in the same way. At the beginning
of this book, we noted that with the Big Bang both space and time
appeared. Yet even if relativity theory combines the two into a space­
time, essential differences remain. In space we can move in all direc­
tions, but not in time. As long as we consider dynamical theories in a
formal world of one time and three space dimensions, we can move
forward as well as backward in both space and time. But as soon as
we include the collective effects of the real world, when we bring the
second law of thermodynamics into the game, all that is over. Time
acquires a direction. We get older, not younger. We may know what
happened in the past, but not what may happen in the future. We
can still influence events in the future, but not those in the past. The
past is a subject for historians, the future one for prophets.
If the order of cause and effects is no longer valid, the world
becomes a strange place. The White Queen in the second part of
Lewis Carroll's Alice in Wonderland can remember the future as well
as the past. She utters a scream because she knows that the next
moment she is going to prick herself with a needle as she puts on
her scarf.
138 Before Time began

In quantum physics, such things exist: an effect of anticipation

as well as one of cause. An electron hit by a photon emits a further
photon as the result of being hit. But for a complete description of
the process one needs a further contribution, in which the electron
emits the second photon before being hit by the first, in anticipation
of the coming interaction. And for the two-body process of electron
and photon, both the "causal" and the "acausal" contribution are
crucial.

r
time
'
,
,�
�< '
''

I
,
,

''
electron photon electron , photon
,
'
'

causal (effect) acausal (anticipation)

The symmetry of cause and effect in quantum physics

The direction of time appears only as an emergent effect in the

macroscopic world of many particles, of many degrees of freedom.
Only then does the second law of thermodynamics take hold. The
gas in Joule's experiment never returns to its initial volume, even
though the trajectories of two atoms in the box can go forward as
well as backward. For them, there is not yet a direction of time-­
that appears only as the joint effort of many particles.
So there exists an order of events, and time marks subsequent
points along this direction. But one can also imagine other quan­
tities fulfilling that purpose in the chronology of the universe, for
example, temperature. The primeval world started in an extremely
hot stage, and the cooling connected to the expansion also deter­
mines specific points of evolution. As the result of the binding of
quarks into hadrons, the physical vacuum appeared some ten
 The End of Time 139

microseconds after the Big Bang, or at about 1012 degrees Kelvin.

The binding of nucleons to form atoms occurred some 380,000 years
later, at a temperature of about 3000 degrees Kelvin. If we label the
axis of happenings by the time, our scale runs from zero (the Big
Bang) to-well, to where, to eternity? And if instead we choose tem­
perature, the scale starts around infinity and ends, we guess, around
zero. But while the time seems to flow smoothly, the temperature
could in principle fluctuate. We can try to use the observable evo­
lution to determine tendencies that we then "extrapolate into the
future." But as the well-known Austro-American physicist Viktor
Weiskopf once said, "Predictions are a tricky thing, particularly if
they involve the future." Weiskopf knew what he was talking about,
since he was one of the first directors of the European Council for
Nuclear Research (CERN) in Geneva, Switzerland, when that center
was just starting. So what is in store for our universe?
To investigate this question (to solve it would be too optimistic),
we have to look a little more closely at the temporal evolution of the
universe and of the mechanisms causing its expansion. The expan­
sion is caused by the remaining dark space energy; gravity provides
an opposing effect. Shortly after the Big Bang, the volume of our
future universe was so small that gravity could dominate the expan­
sion force of the dark energy present at that stage: the rate of expan­
sion decreased with time. Note that it is the rate that decreased, not
the expansion as such. If in a certain year the volume tripled, then
in the next year it only doubled.
At some time, however, the volume had become so large that
dark energy took over. The slowing down of the expansion stopped,
and from that point on the opposite behavior started: if in a given
year the volume doubled, then in the next it tripled. This switch
occurred around a hundred years after the Big Bang, and the new
pattern continues until today. At the time of last scattering, less
than 1% of the overall energy of the universe was dark energy,
whereas today it is 75%. And if nothing happens, this development
will continue.
The behavior we just sketched leaves only
140 Before Time began

Three possibilities

for the chronology of the universe. The evolution up to now and

the different possible future forms are illustrated schematically in
the following picture.

time

"==::::;;::�;:::---� inflation
Big Bang
.... ...
visible universe

The expansion of the universe

After the termination of the Big Bang inflation, the expansion rate
initially decreased, because gravity overcame the expansion effect of
the dark energy: there is simply not enough space. Nevertheless, the
expansion continued and as soon as space had become large enough;
the amount of dark energy became sufficient to increase the expan­
sion rate again. In this stage of ever-growing expansion rate we find
ourselves (still? ) today, according to the latest measurements car­
ried out at the end of the nineties of the last century.
Future development will be decided by the dark energy. If its
density persists unchanged, the expansion of the visible universe will
continue undiminished (pattern a) . It remains possible, however,
 The End of Time 141

that the strength of the dark energy will decrease just enough to
create a balance between it and gravity. Then the expansion would
continue a constant rate (pattern b).And finally, it is possible that
the effect of the dark energy is somehow diminished, reduced in the
course of time, so that in the long run gravity wins-then it would
cause the world to contract again (pattern c).What consequences do
these quite different future perspectives lead to?
If dark energy remains victorious in the long run, then expansion
will continue, either at an increasing rate (a) or uniformly (b).That
means for the future universe, there is neither a spatial nor a tem­
poral limit. Distant galaxies become ever more distant and finally
disappear beyond the Hubble horizon of what for us is within reach
of sight. For us, their lights are turned off. Our own galaxy is com­
pact enough to keep it in shape by gravity-the amount of dark
energy contained within its volume is too small to tear it apart. But
"our" stars eventually also stop shining, because they will have used
up all their fuel. Possibly some new stars are formed, but eventually
also they lose their power: we arrive at

The ultimate nightfall.

The universe becomes ever larger, but also ever colder and ever
emptier.The temperature of the cosmic background radiation drops
to zero, since its wavelength is dilated more and more. Finally, our
world, once filled with billions of galaxies, contains nothing more
but our own small world, our Milky Way, in which more and more
dead stars are circling.All that was farther away has been shifted into
an unattainable distance. If the strength of the dark energy remains
what it is today, our galaxy is saved, because gravity is sufficient to
keep it bound. If, however, the density of dark energy should increase
sufficiently with time, then even that is no longer assured: now
everything could be torn apart, even matter and atoms. In the final
stage, lone nucleons and electrons would then be flying around an
otherwise empty and ever-larger space.The world would then have
died a thermal death--it would have become a uniform and ever more
142 Before Time began

dilute gas. The formation of structure we described in Chapter 6 was

largely based on gravity, and since its role is now over, there is no
more structure. Time is not over, but it has become meaningless:
there are no more events that can define a sequence. The world has
once more become timeless.
In the other case, the presently still increasing expansion would
start to become less, eventually stopping altogether, and finally the
universe would start to contract. To have that happen, the nature of
the dark energy would of course have to change; it would somehow
have to deteriorate, so that it could be overcome by gravity. Then
everything would contract, but this would not lead back to the ori­
ginal starting state; the transition from false to true ground state
can never be reversed. Accidental collisions, no matter how strong,
can never bring the ball back from the valley to come to rest at the
top of the hill. The end is now decided by gravity. The result is a
gigantic black hole, from which nothing can ever escape and which
holographically stores on its surface all the information of the pre­
vious world. The amount of information determines the extent of
the surface.
In this case, our space as well as our time would have been a
limited show. It may be a consolation that other such bubbles appear
continuously-the soup remains boiling-even though we are no
longer around to explain the next universe. In a way, such a picture
is esthetically quite pleasing after all; our own finite life allows us to
imagine a universe that also has a finite life. And it as well is aging,
from a young, hot gas to an old, cold black hole. And the finite life
we witness is only that of our own universe--there are infinitely
many others that appear, exist, and vanish, just as our descendants
continue, live and vanish, with their descendants carrying on to
make evolution endless.
At this point, we briefly mention an alternative approach that has
appeared in various contexts over time. Its first appearance is per­
haps the concept of samsara, reincarnation, an aspect of all Indian
religions. It has also been pursued by ancient Greek and subsequent
Western philosophers. Could it not be that also the--our-universe
 The End of Time 143

will be reborn in the end? Such cyclic cosmic chronologies have

been attempted at various times and have for quite some time failed
because of the second law of thermodynamics. More recently, scen­
arios of this nature have been put forward by Paul Steinhardt and
collaborators in Princeton. They require that the Big Bang picture
presented here, with a transition from false to true normal state and
the resulting inflation, has to be given up. Instead, our universe is
created in a collision of structures in a higher dimensional world,
and rebounds of such collisions could produce a cyclic chronology.
It would in any case also require a decreasing expansion rate of our
universe, for which so far cosmologists have found no indications;
but perhaps the last word has not yet been spoken.
Even if our present universe should finally end as a black hole,
that does not mean the ultimate end. Only in the framework of clas­
sical physics can nothing escape from a black hole. Some fifty years
ago, the well-known English astrophysicist Stephen Hawking found
a quantum-mechanical way out. Empty space, the "nothing," is only
on the average empty. There are always short-time fluctuations, in
which a particle-antiparticle pair comes up out of the depths of
nothing, enters reality, and very shortly afterward annihilates and
thus disappears again. If this process happens at the surface of a black
hole, there is the possibility that the black hole grabs one of the pair
and draws it into its interior. That means the other particle now is
"saved": it no longer has a partner to annihilate with. Energy con­
servation requires that someone has to pay for this process; initially
there was a black hole of a certain mass, and afterwards a black hole
plus the leftover particle. The mass of the black hole must have been
reduced by the mass of this particle. As observed from the outside
(even though there is no one left to observe ) , it appears as if the
black hole is radiating: it emits so-called Hawking radiation. And
because black holes emit more and more such radiation as their
mass decreases, they will evaporate in the long run-in the very
long run, however. The Big Bang was some 1011 years ago; estimates
of a very speculative nature indicate that it will take more than 10100
years before our universe is finally evaporated, with a black hole as
144 Before Time began

the last intermediate step. After the evaporation, only electrons,

positrons, photons, and neutrinos will remain in the form of an
immensely dilute ideal gas in an otherwise empty space.
So, barring reincarnation, there seems to be no escape from the
second law of thermodynamics: in the end, there exists neither
structure nor order nor time. In biblical words: All are from dust, and to
dust all return. Our lives and that of our universe fit in-between.
 APPENDIX I

How Many Configurations of Balls Are There?

In this appendix, we want to illustrate the counting of configurations, of

microstates, using balls in boxes with compartments. Let's begin with a
simple case, two balls and a box with four compartments. For the first ball,
there are four possibilities, for the second there remain three: in total 12
different configurations. To obtain this result, we have assumed that the
balls are distinguishable, say, one red, one blue. If that is not the ca�e. we
counted too many: we have to count configurations arising from the inter­
change of two balls only once. So for identical balls, we get only six con­
figurations, as shown in Figure.
We now turn to four balls in nine compartments. For the first ball, there
are nine possibilities, for the second eight, and so on, so in total we get

9 x 8 x7x 6 = 3024

configurations. But here we also want to consider identical balls, so that we

have to divide by 4 X 3 X 2 = 24. That gives us 126 different configurations
for four identical balls in nine compartments.
For nine balls in nine compartments, there is of course only one pos­
sibility. But doubling the number of compartments now leads to

18Xl7X...X10
----- = 48,620
9X8X...X2

configurations. If we take the balls to be gas atoms and use as a starting

point a Joule experiment in which nine balls are in a box with nine com­
partments, then opening the divider (doubling the number of compart­
ments) gives with equal a priori probabilities for all configurations the
chance of finding once more all balls in the "left" (nine-compartment)
part of the box 1 :48,620.
For the general case, we consider n balls in a box of xn compartments,
where x � 1 is supposed to be a whole number. The number of possible con­
figurations then becomes
146 Appendix 1

different identical

identical

Two balls in four compartments

X( -1)X( -2)X ...X( -n)

xn xn xn xn
�--'-�-'---'�--'��---'-�---'- =

nX(n-l)X(n- 2)X...Xl

=
x
[
n (1 -( l / xn ))X(1 -(2/ xn ))X...X(1 -((n-1)/ xn ))
(1-(1/ n))x (1-(2/ n))x ...X( l-((n-1)/ n))
] .

For sufficiently large x and n, with x >>n >> 1, the factor in square brackets
approaches unity, so that the number of possible configurations becomes
"
x , as quoted in Chapter 6.
 APPENDIX 2

Orbits and Dark Matter

In this appendix, we want to elaborate a bit the relationship between grav­

ity and centrifugal force, as they apply in particular to planetary orbits. In
the second part, we address the role of gravity in maximizing the entropy
of a system.
The law of gravity, as formulated by Isaac Newton, describes the attrac­
tive force between two masses M and m, separated by a spatial distance R.
The law says
Mm
F=G , (1)
2
R

where G= 6.7 X 10-11 m3fkB s1 is the universal constant of gravity. The law is
a special case of the first law of mechanics,

F= ma, (2)
giving the acceleration that a mass m experiences due to the action of a force
F. On the surface of the Earth, we can measure the acceleration caused by
gravity. As already shown by Galileo Galilei, it is the same for all masses, 9.8
m/s1. If we know the radius of the Earth (6.4 X 106m), we can use the last two
equations to get

(3)

for the weight of the Earth. So this is in fact a way to weigh it.
The orbits of the planets around the sun are determined by the inter­
play of the sun's gravity and the opposite centrifugal force,

fnl'2

K =R (4)

caused by the circular motion of the planets. Here we denote with m the
mass of the planet and with R the radius of its orbit around the sun, which
for simplicity we take to be circular; v denotes the velocity of the planetary
motion. Newton's first big success was that with this equation, he could
148 Appendix2

determine how long the moon would take to circulate once around the
Earth. From equations (1) and (4) we obtain for its velocity
CM
v2 =--, (5)
R
8
where M denotes the mass of the Earth and R = 4 X 10 km the distance
between Earth and moon. With 2nR for the orbit, again assumed to be cir­
cular, we find
2 3
(2JT) R
2
T = , (6)
CM

for the passage time T : 30 days. And the same equation (5) determines
the orbits of the planets around the sun, if we take M to be the mass of
2
the sun. In this way, we obtain Kepler's rule v �GM/R, which, as we saw,
is indeed followed by all planets of the solar system. For this, we need the
mass of the sun; it can be determined, once we know the passage time for
11
the Earth (one year) and the distance Earth-sun (1.5 X 10 m) . The result
30
is Msun = 2 X 10 kg, so that the sun is half a million times more massive
than the Earth.
If now the stars at the edge of galaxies show orbital velocities independ­
ent of their distance to the center of gravity of the system-that is, if they
remain constant for increasing R-then equation (5) necessarily leads to
the conclusion that the mass must increase linearly with R. The visible
ma�s, however, remains almost constant for stars beyond the edge of the
galaxy; hence we need the mentioned invisible dark matter to obtain a con­
3
stant orbital passage. The density of dark matter , Mdm/ R then decreases as
2
R- , given that the dark mattl'r mass i n c re ase s linearly in R.
 APPENDIX 3

Cosmological Constant and Dark Energy

In this appendix, we want to show in a little more detail how Einstein

arrived at his cosmological constant, and why today we interpret this as
dark energy. For readers without any physics background, things will be a
bit rough; but I hope that they will understand the conclusions even if they
have to jump over the mathematical argumentation.
We begin with the original equations proposed by Einstein,

(1)

where G is Newton's gravitational constant and c the velocity of light. The

indices µ and V indicate space and time and hence take on the values 0
(time) and 1,2,3 (space). The left side of the equation specifies the curvature
of spacetime in terms of the metric tensor Rµv. The above relation fixes
the spacetime structure as determined by the energy-momentum tensor
T µ,v, defined in terms of the energy density p and the resulting pressure p.
The equations thus give us 8µv a� the solution of an interplay of curvature
operators and spacetime content. If the space is empty, i.e., for Tµv = 0,
we have as the simplest solution the flat Minkowski space, with a diagonal
matrix 8µv = 1, -1, -1, -1, i.e., there is no curvature. Other, more general
solutions allow for the presence of black holes surrounded by empty space.
The details are not really so necessary for an understanding of these equa­
tions; the crucial feature is that the contents of space (the right side of the
equation) determine the curvature of spacetime (the left side).
One result of these equations is that due to its contents, the size of the
universe can change in time. A mea�ure of this change is the so-called
scale factor a(t), which fixes the scale for the determination of the size. As
solution of the Einstein equations, one obtains two relations for this scale
factor, relations that were first derived by the Russian physi cist Alexander
Friedmann. The first of them,
2H =(�)2 = 87r(; p-.!._, (2)
a 3 a2
150 Appendix3

determines the time rate of change v = da/dt = ci of the scale factor. With the
H= / ci a for the Hubble constant H this immediately leads to Hubble's law,

v = ci = Ha, (3)

according to which distant galaxies recede faster the farther away they are.
In equation (2), the energy density in space is given by p while k specifies
the form of space (see Chapter 5). For flat space we have k = 0, for spherical
space k= +1, and for hyperbolic k= -1. As we noted in Chapter 5, the data
of the cosmic microwave radiation indicate a flat space; but for the time
being we'll keep all possibilities.
The second Friedmann equation determines the rate of change of the
scale velocity ii= Ii

ii 4nG
-=-- ( p+3 p ). (4)
a 3
This la�t equation showed that Einstein's quest for a static universe was
ruled out by his equations in their original form. Even if for a spherical
s pac e structure (k =+1) equation (2) would only momentarily allow ci = O ;
equation (4) shows that that soon will change again.
At this point Einstein noted that the mathematical structure of his
equatio ns allowed a modification: one could add a term on the left side

(5)

with A denotinp, a uniwrsal, positive quantity constant in both space and

time: the cosmolo11ical ron.1tant. This modification result e d in correspondingly
modified Friedmann equations:

H2 =(�)2a
=
sm;
3
p
-!_+ A
a2 3'
(6)

and
ii = 1m; + A
- --
(p 3p)+ -.
(7)
a 3 3
At first sight, that seemed to satisfy Einstein's wish for a static universe:
for a spherical structure (k =+l) there were common values of p, p and A
leading to ci =Oas well as to a = O
 Appendix3 151

But the joy did not last long. It soon became clear that any small tem­
poral change of the energy density p would destroy the static nature again.
Einstein's universe was effectively as stable as the ball on the hill: any small
perturbation will cause it to roll down. In the same way, the static universe
was not a stable situation: any density fluctuation would cause it to expand
or contract.
As this point Hubble's discovery entered the scene: the universe was
expanding; it wa5 not static. Einstein noted with regret that he had missed
the chance of predicting such an expansion. It is said that he called the
introduction of the cosmological constant the "biggest blunder" of his life.
We know today that that was not the case---as usual he was simply ahead
of his time. For A= 0 one can indeed get an expanding universe, but the
rate of expansion decreases with time. To agree with the latest supernova
data, the acceleration ii has to be positive, and that requires a sufficiently
large and positive A.
Today's new scenario of multiverse and inflation can be fitted without
problem into the given framework. One only had to move A from one side
of the equation to the other,

(8)

Now the curvature of space and its evolution are determined not only by
the externally introduced energy density in T, but in addition by the space
energy density A, the dark energy. And if we now measure the accelera­
tion of the spatial expansion, as was done the supernova experiments men­
tioned earlier, then we can fix the value of A. That is how the values given
in Chapter 7 were obtained.
 Some references for further reading

This list is in no way exhaustive---it is meant simply to indicate some mate­

rial of help for further information. It also does not aim to cite original
scientific works; however, these are in general referred to in the books or
surveys mentioned.

Martin Bojowald, Once before Time: A Whole Story of the Universe. New York:
Knopf, 2010.
Brian Clegg, Before the BiB Ban8: The Prehistory of Our Universe. New York: St.
Martin's Press, 2009.
Alan Guth, The Inflationary Universe: The Quest for a New Theory of Cosmic OriBins.
Reading, MA: Addison-Wesley, 1997.
Brian Greene, The Hidden Reality. Parallel Universes and the Deep Laws of the Cosmos.
New York: Knopf, 2011.
David Layzer, Cosmo8enesis-- The Growth of Order in the llniverse. New York:
Oxford University Press, 1990.
Andrei Linde, The self-reproducing intlationary universe. Scientific American
Special Edition Cosmos, Spring 1998, 9(1), 98-104.
Roger Penrose, Cycles of Time: An Extraordinary New View of the Universe. London:
Bodley Head, 2010.
Helmut Satz, Ultimate Horizons-Probing the Limits of the Universe. Berlin: Springer
Verlag, 2013.
Paul J. Steinhardt, The inflation debate: is the theory at the heart of mod­
ern cosmology deeply flawed? Scientific American, April 2011, 304(4), 36-43.
Erik Verlinde, On the origin of gravity and the laws of Newton. Journal of
HiBh Enerfl)' Physics April 2011, 29. This is the original reference; for a gen­
eral discussion, see e.g. Dennis Overbye, A scientist takes on gravity. The
New York Times, July 12, 2010.
Alexander Vilenkin, Many Worlds 1n One: The Search for Other Universes, New
York: Hill and Wang, 2006.
 Person Index

B
Bingen, Hildegard von 2 Ising, Ernst 51,52,55,58,65
Boltzmann, Ludwig 102,105
Bose, Satyendra Nath 22
Joule, James Prescott 101, 102,103, 106,
Brout, Robert 63
107,110,111,112,114,115, 138,145
c
K
Carnot, Sadi 107
Kelvin, Lord 109
Carroll, Lewis 123,137
Kepler, Johannes 125,126,128, 129, 148
Chadwick, James 121
Khayyam, Omar 121
Clausius, Rudolf 107
Kibble, Tom 63
Copernicus, Nicolas 13
Curie, Irene 50,51 L
Curie, Marie 50,51 Laplace, Pierre-Simon de 123,127
Curie, Pierre 50,51 Layzer, David 113

D Lenz, Wilhelm 51
Linde, Andrei 9, 10, 76
Democritus 35,46, 73
Lucretius 17,19,37
Dirac, Paul Adrien 18,26
Doppler, Christian 81,120 M
Magellan, Fernando 53
E
Mandelbrot, Benoit 12
Eddington, Arthur 107,109
Mather, John 80
Einstein, Albert 7,8,21,22,89,93,100,
Michell, John 123,127
116,117,132,133,134,149,150,151
Empedocles 99 N
Englert, Francois 63,64 Newton, Isaac 124,125,126,132,147,149

F 0
Fermi, Enrico 21,27,121 Onsager, Lars 51,54
Friedmann, Alexander 149,150
p
G Pauli, Wolfgang 20,27, 121,122
Galilei, Galileo 7,147 Penzias, Arno 13, 14, 77, 79
Gibbs, John Willard 102 Perlmutter, Saul 135
Guralnik, Gerald 63
Guth, Alan 9,10, 76 R
Riess, Adam 135
H
Russell, Bertrand 137
Hagen, Carl 63 Rutherford, Ernest 121, 122
Hawking, Stephen 3,143
Helmholtz, Hermann von 109 s
Higgs, Peter 63,64 Salomon 52

Hubble, Edwin 13,133,141,150,151 Schmidt, Brian 135

156 Person Index

Shaposhnikov, Mikhail 26 w
Sierpinski, Waclaw 12 Weiskopf, Victor 139
Smoot, George 80 Wheeler, John 133
Sobral de Almada Negreiros 59 Wilson, Kenneth 48
Steinhardt, Paul 143 Wilson, Robert 13, 14, 77, 79

v z
Verlinde, Eric 25 Zwicki, Fritz 127
 Subject Index

A expansion 3,6--18,35,37,45,63,67-71,78,

annihilation 41,44,59,66-68,73,74 83,85,88,90,100,101,110,113-118,

antimatter 25-29,40,47,48,58-61,67 133-143,151

atom 14,19,26,31,32,35,41,45,46,49,59,
F
63,69,72,74,77,79,97,98,107,108,
false normal state 11
110,121,138-145
fermion 21-25,28-33,58--60,66,71-76
B form of space 88,89,150

baryogenesis 59,73 forms of matter 99,100

baryons 27,28 fractal 12

black hole 90,94,95,119,122-127,

G
130-133,142,149
galaxy formation 118,134
boojum 123
gluons 23,31,32,37-41,63,67,71,136
boson 22-24,31,32,44,58--67,71-76
grand unification theory (CUT) 28,29,
c 41,42,45,58,59,6Q-63,66-70,73,75

centrifugal force 117,125,127,132,147 gravitational wave detector 94

charge conservation 26 gravitational waves 92-95

charge polarization 30 gravity 8,9,25,28-30,76,83-85,89, 90,95,

cluster formation 117,118 98,115,118-124,142,147,148

continuous transition 57 ground state 5,17,18,25,33,40,62,70,

cosmic background radiation 14-16,35, 75,142

77-92,134,141
H
cosmological principle 11
hadrons 37-41,44,65--68,71-74,136,138
critical behavior 54,72
hadrosynthesis 40,73
cyclic chronology 143
Hawking radiation 143
Cygnus X-1 124
Hexameron 1,2
D Higgs boson 71,74

dark energy 8,9,35-40,62,63,73,89-93, Hubble constant 150

132-135,139, 140-142,149,151
dark matter 89,90,127-136,147,148
Inflation 10, II, 15-18,71,75,76,83,84,
decoupling time 14,79,85
92, 94, 9:'i, 140, 143, t:'il
disorder 55,96, 101,108,110,114
lsingmodel :'il,.�2.:'i4,:'i:'i,65
Doppler shift 81,120
K
E
Kepler's rule 125,126,128,129,148
Einstein equations 133,149
electroweak interaction 62,66,75 L
entropy 105-118,147 leptons 27-33,37,41,42,44,58--63,66,
evaporation 5,49,56, 57,144 68-75
exclusion principle 20,21 light polarization 91-93
158 Subject Index

M quarks 19,20,23,26-32,36--42,45,58-75,

magnetization 50-57,60 136-138

mesons 27,36,37,39,65,67 quintessence 100

metric tensor 149

R
multiverse 4,5,10-13,17,70,73,76,
relaxation 110,111,113,114
132,151
Rigveda 4
N
s
normal state 5,9,11,13-16,71,143
second law of thermodynamics 106-109,
nucleons 19,20,21,26-28,31,32,35-45,
113,117,137,138,143,144
58,59,64,65,68,70,73,74,95,
snark 123
139,141
space curvature 87--89,116
nucleosysnthesis 43-45,69,70,72,73
space energy 8,13,18,35, 62,90,132,
0 134-136
order 5�59,65,98-101,103,108, 109--115 spatial extension 33,83,117
spiral galaxy 119
p strong interaction 36,37,71
percolation 52-54,56 superheating 5,49
phase transition 5,49--52,72 supermassive black hole 124,130
photons 14,22,31,32,41--45,63,66-74, supersymmetry 23, 33
77,79--86,92,118,144
Planck era 76
T

planetary orbits 125,126,147 thermodynamic arrow 100

proton decay 60,62, 70

v

Q vakasha 100

quantum chromodynamics (QCD) 19,

w
36,65,74
WIMP 130, 131
quantum fluctuations 83, 84, 118, 136
quark confinement 19,20 y
quark plasma 20,37, 38, 45 Yin and Yang 25