The End of Time

The End of Time

The Next Revolution in Physics

Julian Barbour
 Oxford New York
Athens Auckland Bangkok Bogotá Buenos Aires
Cape Town Chennai Dar es Salaam Delhi Florence Hong Kong
Istanbul
Karachi Kolkata Kuala Lumpur Madrid Melbourne Mexico City
Mumbai
Nairobi Paris São Paulo Shanghai Singapore Taipei Tokyo Toronto
Warsaw

and associated companies in

Berlin Ibadan

Copyright © 1999 by Julian Barbour

First published by Oxford University Press, Inc., 1999

First issued as an Oxford University Press paperback, 2001

198 Madison Avenue, New York, New York 10016

Oxford is a registered trademark of Oxford University Press

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,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication Data

Barbour, Julian B.
The end of time: the next revolution in our
understanding of the universe/by Julian Barbour.
p. cm.
Includes bibliographical references and index.
ISBN-13 978-0-19-511729-5; 978-0-19-514592-2 (pbk.)

1. Space and time. 2. Relativity (Physics).

3. Quantum theory. I. Title.
QC173.59.S65 B374 2000
530.11 21—dc21 99-044319

9 10
Printed in the United States of America
on acid-free paper
CONTENTS

The Story in a Nutshell

Preface

Acknowledgements

PART 1 THE BIG PICTURE IN SIMPLE TERMS

CHAPTER 1 The Main Puzzles

The Next Revolution in Physics

The Ultimate Things

Getting to Grips with Elusive Time

The Properties of Experienced Time

Newton’s Concepts

Laws and Initial Conditions

Why is the Universe so Special?

CHAPTER 2 Time Capsules

The Physical World and Consciousness

Time Without Time

 Time Capsules

Examples of Time Capsules

CHAPTER 3 A Timeless World

First Outline

The Crisis of Time

The Ultimate Arena

Is Motion Real?

The Big Picture

PART 2 THE INVISIBLE FRAMEWORK AND THE

ULTIMATE ARENA

CHAPTER 4 Alternative Frameworks

Absolute or Relative Motion?

An Alternative Arena

CHAPTER 5 Newton’s Evidence

The Aims of Machian Mechanics

Apparent Failure

Space and Spin

Energy

CHAPTER 6 The Two Great Clocks in the Sky

 Where is Time?

The First Great Clock

The Inertial Clock

The Second Great Clock

CHAPTER 7 Paths in Platonia

Nature and Exploration

Developing Machian Ideas

Exploring Platonia

PART 3 THE DEEP STRUCTURE OF GENERAL

RELATIVITY

CHAPTER 8 The Bolt from the Blue

Historical Accidents

Background to the Crisis

Einstein and Simultaneity

The Forgotten Aspects of Time

CHAPTER 9 Minkowski the Magician

The New Arena

From Three to Four Dimensions

Are There Nows in Relativity?

CHAPTER 10 The Discovery of General Relativity

Funny Geometry

Einstein’s Way to General Relativity

The Main Advances

The Final Hurdle

General Relativity and Time

CHAPTER 11 General Relativity: The Timeless Picture

The Golden Age of General Relativity

Platonia for Relativity

Best Matching in the New Platonia

Catching up with Einstein

A Summary and the Dilemma

PART 4 QUANTUM MECHANICS AND QUANTUM

COSMOLOGY

CHAPTER 12 The Discovery of Quantum Mechanics

CHAPTER 13 The Lesser Mysteries

Introduction

The Wave Function

Interpreting the Wave Function

 States Within States

The Copenhagen Interpretation

Heisenberg’s Uncertainty Principle

The Enigmatic Gem

CHAPTER 14 The Greater Mysteries

Schrödinger’s Vast Arena

Correlations and Entanglement

The EPR Paradox

Bell’s Inequalities

The Many-Worlds Interpretation

A Dualistic Picture

CHAPTER 15 The Rules of Creation

The End of Change

Creation and the Schrödinger Equation

Quantum Mechanics Hovering in Nothing

CHAPTER 16 ‘That Damned Equation’

History and Quantum Cosmology

A Simple-Minded Approach

‘That Damned Equation’

 PART 5 HISTORY IN THE TIMELESS UNIVERSE

CHAPTER 17 The Philosophy of Timelessness

CHAPTER 18 Static Dynamics and Time Capsules

Dynamics Without Dynamics

Why do we Think the Universe is Expanding?

The Idea of Time Capsules: The King sher

CHAPTER 19 Latent Histories and Wave Packets

Smooth Waves and Choppy Seas

History Without History

Airy Nothing and a Local Habitation

Schrödinger’s Heroic Failure

CHAPTER 20 The Creation of Records

History and Records

The Creation of Records: First Mechanism

The Prerequisites of History

The Improbability of History

The Creation of Records: Second Mechanism

CHAPTER 21 The Many-Instants Interpretation

Many Histories in One Universe

 Bell’s ‘Many-Worlds’ Interpretation

The Many-Instants Interpretation

CHAPTER 22 The Emergence of Time and its Arrow

Causality in Quantum Cosmology

Soccer in the Matterhorn

Timeless Descriptions of Dynamics

A Quantum Origin of the Universe?

Vision of a Timeless Universe

A Weil-Ordered Cosmos?

EPILOGUE Life Without Time

Notes

Further Reading

Bibliography

Index

LIST OF DISPLAY ITEMS

BOXES
 1 The Great Revolutions in Physics

2 The Two Big Mysteries

3 Possible Platonias

4 Centre of Mass

5 The Galilean Relativity Principle

6 The Equation of Time

7 Tait’s Inertial Clock

8 Intrinsic Di erence and Best Matching

9 Relativity in One Diagram and 211 Words

10 The Impossible Becomes Possible

11 The Two-Slit Experiment

12 Entangled States

13 How Creation Works

14 The Semiclassical Approach

15 Static Wave Packets

ILLUSTRATION

The Ariel in a Snowstorm

FIGURES

1 Absolute Space and Time

2 Motion as an Illusion

3 Triangle Land

4 Triangle Land with Frontiers

5 Platonia

6 Platonia with Mist Distribution

7 Triangle Land with Similar Triangles

8 Shape Space

9 Three-Body History in Shape Space

10 Another History in Shape Space

11 Centre of Mass

12 E ect of Galilean Relativity

13 ‘Spaghetti Diagrams’ in Absolute Space

14 Nine Histories in Shape Space

15 A Spiral Galaxy

16 Saturn and its Rings

17 Potential Energy

18 Galileo’s Diagram of Parabolic Motion

19 Coordinates in Tait’s Problem

20 A Solution of Tait’s Problem

21 Trial Placing of Two Triangles

22 Illustration of Interference Fringes

23 Magnetic Lines of Force

24 The Pond Argument

25 Einstein’s De nition of Simultaneity

26 Mutual Contraction of Rods

27 Space-Time Diagram

28 Space-Time with Light Cones

29 Three Kinds of Space-Time

30 Three-Spaces in Space-Time

31 Space-Time as a Tapestry

32 Distribution of Hits Behind One Slit

33 Expected Distribution for Two Slits

34 Actual Distribution

35 Behaviour of Wave Function

36 Wave Function of Momentum State

37 Superposition of Waves

38 Superposition of ‘Spiky’ Waves

39 Two-Dimensional Con guration Space

40 Con guration Space with Probability Density

41 Collapse of Wave Function

42 Entangled States

43 Measurement Based on an Entangled State

44 Two Nearly Sinusoidal Wave Patterns

45 A Regular Wave Pattern

46 Explanation of Fermat’s Principle

47 E ect of Wave Interference

48 Wave Patterns with Spikes

49 A Moving Wave Packet

50 Creation of an Alpha-Particle Track

51 Multiple Tracks of Elementary Particles

52 Division of Platonia

53 Schematic Representation of Platonia

54 Chronology of the Universe

55 From Big Bang to Big Crunch

THE STORY IN A NUTSHELL

Two views of the world clashed at the dawn of thought. In the great
debate between the earliest Greek philosophers, Heraclitus argued
for perpetual change, but Parmenides maintained there was neither
time nor motion. Over the ages, few thinkers have taken Parmenides
seriously, but I shall argue that Heraclitan ux, depicted nowhere
more dramatically than in Turner’s painting below, may well be
nothing but a well-founded illusion. I shall take you to a prospect of
the end of time. In fact, you see it in Turner’s painting, which is
static and has not changed since he painted it. It is an illusion of
ux. Modern physics is beginning to suggest that all the motions of
the whole universe are a similar illusion – that in this respect Nature
is an even more consummate artist than Turner. This is the story of
my book.
 Snow Storm – Steamboat o a Harbour’s Mouth Making Signals in
Shallow Water, and Going by the Lead. The Author was in this Storm on
the Night the Ariel left Harwich (1842). The 67-year-old Turner
claimed that he had made the sailors bind him to the Ariel’s mast so
that he should be forced to experience the full fury of the storm.
PREFACE

We must philosophize about these things di erently.

Johannes Kepler

On a beautiful October afternoon in 1963 I travelled to the Bavarian

Alps with a student friend, Jürgen. We planned to spend the night in
a hut and climb to the peak of the Watzmann at dawn next day. On
the train, I read an article about Paul Dirac’s attempt to unify
Einstein’s general theory of relativity with quantum theory. A single
sentence in it was to transform my life: ‘This result has led me to
doubt how fundamental the four-dimensional requirement in
physics is.’ In other words, Dirac was doubting that most wonderful
creation of twentieth-century physics: the fusion of space and time
into space-time.

I never climbed the Watzmann. When Jürgen’s alarm rang an

hour before dawn, I awoke with a splitting headache. I still
remember vividly the brilliant stars of Orion and the other winter
constellations, high in the sky before the October dawn. But stars or
no stars, I could not face the climb with that headache. Jürgen set
o alone, but I took two aspirin and went back to my bunk. Waking
an hour or two later, I fell to thinking about Dirac’s words. Might
the notion of space-time have been a mistake? This prompted an
even more fundamental question: what is time? Before Jürgen had
returned, I was – and still am – the prisoner of this question.

Richard Feynman once quipped, ‘Time is what happens when

nothing else does.’ My conclusion, reached within a few days, was
the exact opposite: time is nothing but change. I spent hours and
hours pacing through the Englischer Garten in Munich while
persuading myself of this fact. Physics must be recast on a new
foundation in which change is the measure of time, not time the
measure of change. After a week or two I had become so gripped by
the issue of time that I decided to go to Cambridge, where Dirac was
the Lucasian Professor, as Newton had once been, to try to explain
my ideas to him. I could have saved myself the trouble. Although I
had studied mathematics in Cambridge from 1958 to 1961, I had
never been to Dirac’s lectures and did not know he was a man of
few words, seldom engaging in general discussions even with his
most distinguished colleagues. I did speak to him brie y on the
telephone, on which he introduced himself with ‘This is Mr Dirac’,
but, very reasonably I am sure, he ended the conversation quite
quickly.

If the trip to Cambridge failed in that way, it had for me a most

fortunate side e ect. While over in England, I went back to the
village in north Oxfordshire where I had grown up. There I found
my younger brother wondering if he could nd the money to
become a farmer. New College in Oxford had suddenly decided to
auction, in separate lots, a farm they owned in the village. The
auction was set for 21 November. We had both received some
money from my father as a way of avoiding death duties. When it
became clear, just twenty-four hours before the auction, that my
brother might manage to get together, with loans, enough money to
buy the land but not the farmhouse and buildings, I decided on the
spur of the moment to bid for them myself and rent them to my
brother for a few years until he could make other arrangements. At
dawn on the morning of the sale, my father woke me and my
mother woke my brother. They had been unable to sleep, lying
awake all night worrying about our plans. We must give up the idea.
However, our bank manager encouraged us to go ahead. A few
hours later – and about twenty-eight hours before President
Kennedy was assassinated – I was the proud owner of College Farm,
as it was and still is called. Built at the end of the Commonwealth in
1659 and standing next to a ne medieval church, it is one of the
best-preserved yeoman farmhouses in the country.

I have told you this story of my serendipitous purchase because

it signi cantly a ected the kind of scientist I became. After I
returned to Munich, and my brother started to plough his elds, I
decided to give up the Ph.D. in astrophysics I had begun and turn to
fundamental physics, above all time. I did a Ph.D. on Einstein’s
theory of gravitation in Cologne and then started to think about a
university position in Britain. But even then there was pressure to
‘publish or perish’. If you could not turn out one or two research
papers each year (now, crazily, one is expected to produce four or
ve) and do all the teaching and administrative duties, I was
warned, you could not look forward to much of a career. But I might
want to spend years thinking about basic issues before publishing
anything. As luck would have it, I had learned Russian as a hobby
while in Munich and had earned some money by translating Russian
scienti c journals. Once you get into such work, it goes quite fast,
especially if you can dictate it. So I decided to earn my living that
way, and work away at the question of time as and when I could. In
1969 my German wife and I moved into College Farm with our two
small children, who were soon followed by two more. For twenty-
eight years, until I felt the size of my pension fund allowed me to
stop, I turned out translations at the rate of two and a half million
words a year. I think that together they would ll about twenty
metres of library shelves.

It was a great way to bring up a family, but an unconventional

way to do physics. For years I never met anyone else at conferences
who was not at a university or research institute. Now you do meet
a few. James Lovelock, the creator of the Gaia theory of life on
Earth, is a great advocate for going independent. I think it worked
for me. I greatly valued the feeling that I could do just what I
wanted when I wanted. Publication of papers led to fruitful
collaboration with other physicists and trips to many parts of the
world. I had the luxury of being able to work on topics other
physicists felt they could not risk, either because nothing might
come of the work or because their reputations would su er. But
they still liked to talk about them, and I made several good friends
in this way. And all the while, I did seem to make some progress on
the enigma of time. Key ideas came every ve or six years, the most
radical in 1991. In fact, 35 years on from that failed attempt on the
Watzmann, I now believe that time does not exist at all, and that
motion itself is pure illusion. What is more, I believe there is quite
strong support in physics for this view. I have a vision and I want to
tell you about it.

You may wonder how I can preface a belief that time does not
exist with a bit of personal history. How can history be if there is no
time? That is the great question, and my answer comes at the end of
the book. Most of the book is about what evidence physics can o er
for and against the existence of time. However, in the rst part I try
to explain, in the simplest terms possible, the main issues, and to
relate them to your direct experience of time. I want to try to make
sure, if you have bought or borrowed this book, that you do not put
it down in despair, unable to understand what I am driving at. I
hope also that this introduction will encourage you to read on to the
details. Many are fascinating in their own right. Because temporal
concepts are so deeply lodged in our experience and language, I
shall often write as if time existed in the way most people think it
does. The same applies to motion. Please do not think I am being
inconsistent – I should have to use many more words to express
everything in a timeless fashion.
 I have tried to make the text self-contained and accessible to any
reader fascinated by time. If you nd some parts harder then others,
please do not worry if you have to give up on them. Several non-
scientists who read a much more technical rst draft found they
could simply skim the harder parts and still pick up much of the
message. For this reason, the more technical material that is not
completely central to the story is generally put in boxes – take that
as a sign not to worry if you have di culty digesting it (though I
hope you will at least try it). Also, various digressions, of potential
interest to all readers, and genuinely technical material for
cognoscenti are to be found in the notes at the end. I suggest you
look at them after you have read each chapter. To help readers with
little or no scienti c background, the most important technical
terms appear in the Index so that you can readily locate
explanations of them in the text if necessary. Books for further
reading are also recommended.

I dedicated my rst book to my wife and our children. I dedicate

The End of Time to my indomitable mother, just ninety-six and still
hearing the larks clearly and singing as lustily in her church choir. I
dedicate it equally to the memory of my father, who died three
years after my wife and I moved into College Farm. My father,
whom I missed very much, had a most useful saying that I should
like to share with you: ‘Never believe anything anyone ever tells you
without checking again and again.’ That has saved me from many a
disaster. A very good friend of mine, Michael Purser, once remarked
that if my mother was the irresistable force, my father was surely
the immovable object. Whatever the truth, I should not be here but
for them. Being here is the supreme gift.

J.B.
South Newington, March 1999

Note This printing of the book di ers from the initial hardback in
the correction of some minor errors and misprints, additions to the
bibliography and books recommended for further reading, and slight
rearrangement of the Notes to take into account new results
obtained with Niall Ó Murchadha after the book had been written.
This recent work should, if it stands up to critical examination,
strengthen my arguments that time does not exist. See especially p.
358.
ACKNOWLEDGEMENTS

Several people who have helped me greatly are mentioned in the

text and notes, where it seemed more appropriate to express my
gratitude to them. All of them also helped by reading some or all of
an early draft and making comments. I am also grateful to several
others (listed here in no particular order) who did the same: Dr
Ti any Stern, Michael Pawley, David Rizzo, Mark Smith, Dr Fotini
Markopoulou, Gretchen Mills Kubasiak (with particularly detailed
and helpful comments), Oliver Pooley, Dr Joy Christian, Cyril
Aydon, Dr John Purser, Jason Semitecolos, Todd Heywood, John
Wheeler (this is not J.A. Wheeler, though he did read the later draft,
for which I am most grateful), Christopher Richards, Michael Ives,
Elizabeth Davis and Ian Phelps. Joyce Aydon, Mark Smith and Tina
Smith helped greatly with the preparation of the text. I should like
to thank too Steve Farrar and his editor Tim Kelsey, who went to
great trouble to report my ideas accurately in an article (entitled
‘Time’s assassin’!) in the Sunday Times in October 1998.

I am especially indebted to my friend Dierck Liebscher of the

Astrophysikalisches Institut Potsdam, who prepared all the
computer-generated diagrams (and also made helpful comments on
the text).
 Both my editors (Peter Tallack for the UK edition, Kirk Jensen
for the North American edition) have done very well what editors of
a book like this should do: be supportive but insist that it is for the
popular market, not an academic text. It is not for me to judge how
readable the nal result is, but to the extent that it is, my readers
must be grateful to them, as I am. I am also grateful to my copy-
editor, John Woodru , for numerous stylistic improvements and his
thorough work. Lee Smolin, who appears often in the main text,
needs to be mentioned especially here too, since he made the most
valuable suggestion that I write the introductory chapters that
comprise Part 1. Without these, the book in its rst draft was much
tougher.

My wife, Verena, and our children have been wonderfully

supportive.

I also want to thank here my literary agent Katinka Matson and

her partner John Brockman, founder of Brockman, Inc., not only for
nding me quite the best publishers and editors I could hope for but
also for a remark of John’s that encouraged me to write the kind of
book this has become. According to John, ‘Roger Penrose has found
the right way to write popular science today. He’s really writing for
his colleagues, but he is letting the public look over his shoulder.’
For myself, I have certainly tried to write primarily for the general
reader, but, in a reversal of John’s aphorism, I shall be more than
happy if my colleagues look over my shoulder. This is a serious
book, and it draws its inspiration from the way Penrose’s The
Emperor’s New Mind engages with intensity – passion, even – both
the interested public and working scientists. That is what gives his
book its cutting edge and thereby makes it more absorbing for the
non-specialist. Richard Dawkins’s The Sel sh Gene is another
example that comes to mind.

I have left to the end one other important person – you, the
reader. As you will know from the Preface, I have tried throughout
my life to fund my own research and would like to continue to do
so. Every copy of this book that is bought (and borrowed from a
library) helps me in this way. Thank you, and I do hope you get
some pleasure from this book. I have enjoyed writing it. I hope to
continue popularizing the study of time and will post details on my
Website (www.julianbarbour.com) together with any signi cant
developments of which I become aware in the study of time.
PART 1
The Big Picture in Simple Terms

As explained in the Preface, I start with three chapters in which I

have attempted to present my main ideas with the minimum of
technical details. The main aim is to introduce a de nite way of
thinking about instants of time without having to suppose that they
belong to something that ows relentlessly forward. I regard
instants of time as real things, identifying them with possible
instantaneous arrangements of all the things in the universe. They
are con gurations of the universe. In themselves, these
con gurations are perfectly static and timeless. But how and why
can something static and timeless be experienced as intensely
dynamic and temporal?

That is what I hope to explain in simple terms in these rst three

chapters.
CHAPTER1
The Main Puzzles

THE NEXT REVOLUTION IN PHYSICS

Nothing is more mysterious and elusive than time. It seems to be the

most powerful force in the universe, carrying us inexorably from
birth to death. But what exactly is it? St Augustine, who died in AD
430, summed up the problem thus: ‘If nobody asks me, I know what
time is, but if I am asked then I am at a loss what to say.’ All agree
that time is associated with change, growth and decay, but is it more
than this? Questions abound. Does time move forward, bringing into
being an ever-changing present? Does the past still exist? Where is
the past? Is the future already predetermined, sitting here waiting for
us though we know not what it is? All these questions will be
addressed in this book, but the biggest remains the one St Augustine
could not answer: what is time?

Curiously, physicists have tended not to ask this question,

preferring to leave it to philosophers. The reason is probably the
colossal and dominating in uence of Isaac Newton and Albert
Einstein. They shaped the way physicists think about space, time and
motion. Each created a representation of the world of unsurpassed
clarity. But having seen their way to a structure of things, they did
not bother unduly about its foundations. This creates potential for
confusion. Without question, their theories contain wonderful truths,
but they both take time as given. It is a building block on a par with
space, a primary substance. In fact, Einstein fused it with three-
dimensional space to make four-dimensional space-time. This was
one of the great revolutions of physics (Box 1).

BOX 1 The Great Revolutions of Physics

1543: The Copernican Revolution. In On the Revolutions of the

Celestial Spheres, Nicolaus Copernicus proposed that the Earth moves
around the centre of the universe. The modern meaning of revolution
derives from his title. He established the form of the solar system.
Curiously, the Sun plays little part in his scheme; he merely placed it
near the centre of the universe. About sixty years later Johannes
Kepler showed that the Sun is the true centre of the solar system, and
with Galileo Galilei he prepared the way for the next revolution.

1687: The Newtonian Revolution. In The Mathematical

Principles of Natural Philosophy, Newton formulated his three famous
laws of motion and the theory of universal gravitation. He showed
that all bodies – terrestrial and celestial – obey the same laws, and
thus set up the rst scheme capable of describing the entire universe
as a uni ed whole. Newton created the science of mechanics, now
often called dynamics, which ushered in the modern scienti c age.
He claimed that all motions take place in an in nite, immovable,
absolute space and that time too is absolute and ‘ ows uniformly
without relation to anything external’.

1905: The Special Theory of Relativity. In a relatively short

paper on electro-magnetism, Einstein showed that simultaneity
cannot be de ned absolutely at spatially separated points, and that
space and time are inextricably linked together. What appears as
space and what appears as time depends on the motion of the
observer. He made startling predictions about the behaviour of
measuring rods and clocks, and found his famous equation £ = mc2.
In 1908 Hermann Minkowski formalized the notion of space-time as
a rigid, indissoluble, four-dimensional arena of world events.

1915: The General Theory of Relativity. The special theory of

relativity describes a world without gravitation. After an eight-year
gestation, Einstein nally formulated his general theory of relativity
in which the rigid arena of Minkowski’s space-time is made exible,
responding to the presence of matter in it. Gravity is given a
brilliantly original interpretation as an e ect of the curving of space-
time. The theory showed that time can have a beginning (the Big
Bang) and that the universe can expand or contract. Although to a
remarkable degree it was a creation of pure thought, many
predictions of this theory have now been very well con rmed. It
describes the large-scale properties of matter and the universe as a
whole.

1925/6: Quantum Mechanics. This gets its name because it

shows that some mechanical quantities are found in nature only in
multiples of discrete units called quanta. This is a distinctive
di erence from the theories of Newton and Einstein, which are now
called classical (as opposed to quantum) theories. The rst quantum
e ects were discovered and described on an ad hoc basis by Max
Planck (1900), Einstein (1905) and Niels Bohr (1913), while a
consistent quantum theory was found in two di erent but equivalent
forms: matrix mechanics, by Werner Heisenberg (1925), and wave
mechanics, by Erwin Schrödinger (1926). Paul Dirac also made
outstanding contributions. Quantum mechanics describes the
properties of light, especially lasers, and the microscopic world of
atoms and molecules. It is the bedrock of all modern electronic
technology, but its results are ba ingly counter-intuitive and raise
profound issues about the nature of reality. It is also puzzling that
theories of completely di erent structures are used to describe the
macroscopic universe (classical general relativity) and microscopic
atoms (quantum mechanics).

Revolutions are what make physics such a fascinating science. Every

now and then a totally new perspective is opened up. But it is not
that we close the shutters on one window, open them on another,
and nd ourselves looking out in wonder on a brand-new landscape.
The old insights are retained within the new picture. A better
metaphor of physics is mountaineering: the higher we climb, the
more comprehensive the view. Each new vantage point yields a
better understanding of the interconnection of things. What is more,
gradual accumulation of understanding is punctuated by sudden and
startling enlargements of the horizon, as when we reach the brow of
a hill and see things never conceived of in the ascent. Once we have
found our bearings in the new landscape, our path to the most
recently attained summit is laid bare and takes its honourable place
in the new world.

Today, physicists con dently, indeed impatiently, await the next

revolution. But what will it be? In 1979, when, like Newton and
Dirac before him, Stephen Hawking became the Lucasian Professor at
Cambridge, he announced in his inaugural address the imminent end
of physics. Within twenty years physicists would possess a theory of
everything, created by a double uni cation: of all the forces of
nature, and of Einstein’s general theory of relativity with quantum
mechanics. Physicists would then know all the inner secrets of
existence, and it would merely remain to work out the consequences.

Neither uni cation has yet happened, though one or both

certainly could. (Hawking has recently said that his prediction still
stands but that ‘the twenty years starts now’.) For myself, I doubt
that would spell the end of physics. But uni cation of general
relativity and quantum mechanics may well spell the end of time. By
this, I mean that it will cease to have a role in the foundations of
physics. We shall come to see that time does not exist. Though still
only a prospect on the horizon, this, I think, could well be the next
revolution. What a denouement if it is!

I believe that the basic elements of this potential revolution – the

reasons for it and its likely outcome – can already be discerned. In
fact, as we shall shortly see, clear hints that time may not exist, and
that quantum gravity – the uni cation of general relativity and
quantum mechanics – will yield a static picture of the quantum
universe, started to emerge about thirty years ago, but made
remarkably little impact. This is one of my reasons for writing this
book: these things should be better known. They are only just
beginning to be mentioned in books for the general reader, and even
most working physicists know little or nothing about them.

No doubt many people will dismiss the suggestion that time may
not exist as nonsense. I am not denying the powerful phenomenon
we call time. But is it what it seems to be? After all, the Earth seems
to be at. I believe the true phenomenon is so di erent that,
presented to you as I think it is without any mention of the word
‘time’, it would not occur to you to call it that.

If time is removed from the foundations of physics, we shall not

all suddenly feel that the ow of time has ceased. On the contrary,
new timeless principles will explain why we do feel that time ows.
The pattern of the rst great revolution will be repeated. Copernicus,
Galileo and Kepler taught us that the Earth moves and rotates while
the heavens stand still, but this did not change by one iota our direct
perception that the heavens do move and that the Earth does not
budge. Our grasp of the interconnection of things was, however,
eventually changed out of recognition in ways that were impossible
to foresee. Now I think we must, in an ironic twist to the Copernican
revolution, go further, to a deeper reality in which nothing at all,
neither heavens nor Earth, moves. Stillness reigns.

People often ask me what are the implications of the non-

existence of time. What will it mean for everyday life? I think we
cannot say. Copernicus had no inkling of what Newton (let alone
Einstein) would nd, though it all owed from his revolution. But we
can be certain that our ideas about time, causality and origins will be
transformed. At the personal level, thinking about these things has
persuaded me that we should cherish the present. That certainly
exists, and is perhaps even more wonderful than we realize. Carpe
diem – seize the day. I expand on this in the Epilogue.

THE ULTIMATE THINGS

This book revolves around three questions: What is time? What is

change? What is the plan of the universe? The only way to answer
them is to examine the structure of our most successful theories. We
must fathom the architecture of nature. What part, if any, is played
by time in these theories? Can we identify the ultimate arena of the
world?

These questions were forced upon physicists by the work I

mentioned in the Preface. It is one of the two big (and almost
certainly intimately connected) mysteries of modern physics (Box 2).
Both are aspects of an as yet unbridged chasm between classical and
quantum physics.
BOX 2 The Two Big Mysteries

As explained in Box 1, physicists currently describe the world by

means of two very di erent theories. Large things are described by
classical physics, small things by quantum physics. There are two
problems with this picture.

First, general relativity, Einstein’s theory of gravity, seems to be

incompatible with the principles of quantum mechanics in a way that
Newtonian dynamics and the theory of electromagnetism, developed
by Michael Faraday and James Clerk Maxwell in the nineteenth
century, are not. For these theories, it proved possible to transform
them, by a process known as quantization, from classical into
quantum theories. Attempts to apply the same process to general
relativity and create quantum gravity failed. It was this technical
work, by Dirac and others, which brought to the fore all the
problems about time with which this book is concerned.

The second mystery is the relationship between quantum and

classical physics. It seems that quantum physics is more fundamental
and ought to apply to large objects, even the universe. There ought
to be a quantum theory of the universe: quantum cosmology. But
quantum physics does not yet exist in such a form. And its present
form is very mysterious. Part of it seems to describe the actual
behaviour of atoms, molecules and radiation, but another part
consists of rather strange rules that act at the interface between the
microscopic and macroscopic worlds. Indeed, the very existence of a
seemingly unique universe is a great puzzle within the framework of
quantum mechanics. This is very unsatisfactory, since physicists have
a deep faith in the unity of nature. Because general relativity is
simultaneously a theory of gravity and the large-scale structure of
the universe, the creation of quantum cosmology will certainly
require the solution of the only slightly narrower problem of
quantum gravity.

One of the themes of my book is that this chasm has arisen because
physicists have deep-rooted but false ideas about the nature of space,
time and things. Preconceptions obscure the true nature of the world.
Physicists are using too many concepts. They assume that there are
many things, and that these things move in a great invisible
framework of space and time.

A radical alternative put forward by Newton’s rival Leibniz

provides my central idea. The world is to be understood, not in the
dualistic terms of atoms (things of one kind) that move in the
framework and container of space and time (another quite di erent
kind of thing), but in terms of more fundamental entities that fuse
space and matter into the single notion of a possible arrangement, or
con guration, of the entire universe. Such con gurations, which can
be fabulously richly structured, are the ultimate things. There are
in nitely many of them; they are all di erent instances of a common
principle of construction; and they are all, in my view, the di erent
instants of time. In fact, many people who have written about time
have conceived of instants of time in a somewhat similar way, and
have called them ‘nows’. Since I make the concept more precise and
put it at the heart of my theory of time, I shall call them Nows. The
world is made of Nows.

Space and time in their previous role as the stage of the world are
redundant. There is no container. The world does not contain things,
it is things. These things are Nows that, so to speak, hover in
nothing. Newtonian physics, Einstein’s relativity and quantum
mechanics will all be seen to do di erent things with the Nows. They
arrange them in di erent ways. What is more, the rules that govern
the universe as a whole leave imprints on what we nd around us.
These local imprints, which physicists take as the fundamental laws
of nature, reveal few hints of their origin in a deeper scheme of
things. The attempt to understand the universe as a whole by
‘stringing together’ these local imprints without a grasp of their
origin must give a false picture. It will be the at Earth writ large.
My aim is to show how the local imprints can arise from a deeper
reality, how a theory of time emerges from timelessness. The task is
not to study time, but to show how nature creates the impression of
time.

It is an ambitious task. How can a static universe appear so

dynamic? How is it possible to watch the ashing colours of the
king sher in ight and say there is no motion? If you read to the
end, you will nd that I do propose an answer. I make no claim that
it is de nitely right – choices must be made, and many physicists
would not make mine. If all were clear, I should not have promised a
but the theory of time. In order not to interrupt the ow of the text, I
make few references to the problems in my timeless description of
the world. Instead, I have collected together all those of which I am
aware in the Notes. Although, as will be evident throughout the
book, I do believe rather strongly in the theory I propose, there is a
sense in which even clear disproof of my theory would be exciting
for me. The problems of time are very deep. Clear proof that I am
wrong would certainly mark a signi cant advance in our
understanding of time. In a way, I cannot lose! Whatever the
outcome, I shall be more than happy if this book gives you a novel
way of thinking about time, exposes you to some of the mysteries of
the universe, and encourages even one reader to embark, as I did 35
years ago, on a study of time.

For the study of time is not just that – it is the study of

everything.

GETTING TO GRIPS WITH ELUSIVE TIME

The hardest thing of all is to nd a black cat in a dark room,

especially if there is no cat.
Confucius

We must begin by trying to agree what time is. The problems start
already, as St Augustine found. Nearly everybody would agree that
time is experienced as something linear. It seems to move forward
relentlessly, through instants strung out continuously on a line. We
ride on an everchanging Now like passengers on a train. Each point
on the line is a new instant. But is time moving forward – and if so
through what – or are we moving forward through time? It is all very
puzzling, and philosophers have got into interminable arguments. I
shall not attempt to sort them out, since I do not think it would get
us anywhere. The trouble with time is its invisibility. We shall never
agree unless we can talk about something we can see and grasp.

I think it is more fruitful to try to agree on what an instant of

time is like. I suggest it is like a ‘three-dimensional snapshot’. In any
instant, we see objects in de nite positions. Snapshots con rm our
impression; artists were painting pictures that look like snapshots
long before cameras were invented. This does seem to be a natural
way to think about the experience of an instant. We also have
evidence from the other senses. I feel an itch at the same time as
seeing a moving object in a certain position. All the things I see,
hear, smell and taste are knit together in a whole. ‘Knitting together’
seems to me the de ning property of an instant. It gives it a unity.

The three-dimensional snapshots I have in mind could be

constructed if many di erent people took ordinary two-dimensional
snapshots of a scene at the same instant. Comparison of the
information in them makes it possible to build up a three-
dimensional picture of the world in that instant. That is what I mean
by a Now. It is very remarkable that such completely di erent two-
dimensional pictures can be reconciled in a three-dimensional
representation. The possibility of this ordering is what leads us to say
that things exist in three-dimensional space. It leads to an even
deeper ‘knitting together’ over and above the directly experienced
sense of being aware of many di erent things at once (it is this that
enables us to know instantly that we are seeing, say, six distinct
objects without counting them individually). I regard space as a
‘glue’, or a set of rules, that binds things together. It is a plurality
within a deep unity, and it makes a Now.

You may object that no experience is instantaneous, just as

snapshots require nite exposures. True, but we can still liken
instants to snapshots. It is the best idealization I know. It allows us to
begin to get our hands on time, which is otherwise for ever slipping
through our ngers. As instants, rather than an invisible river, time
becomes concrete. We can pore over photographs, looking for
evidence in them like military intelligence analysts studying satellite
pictures. We can imagine ‘photographing’ our successive experiences,
obtaining innumerable snapshots. Using them, we can identify the
most important properties of experienced time.

THE PROPERTIES OF EXPERIENCED TIME

Suppose that the snapshots are taken when we are witnessing lots of
things happening, say people streaming past us in a street, and that
the snapshots (either two-dimensional, as directly experienced, or
‘three-dimensional’, as explained above), once taken, are jumbled up
in a heap. A di erent person, given the heap, could relatively easily,
by examining the details in the snapshots, arrange them in the order
they were experienced. A movie can be reassembled from its
individual frames. My notion of time depends crucially on the details
that the ‘snapshots’ carry. It requires the richly structured world we
do experience.

This imaginary exercise brings out the most important property of

experienced time: its instants can all be laid out in a row. They come
in a linear sequence. This is a very strong impression. It is created
not by invisible time, but by concrete things.

It is harder to pin down other properties. I have already

mentioned the di culty of saying precisely what the powerful
impression of moving forward in time consists of. We also have the
intuition of length of time, or duration. Indeed, seconds, minutes,
hours dominate our age, though you may not know how these
precise notions have arisen. That is an important issue. Finally, there
is the remarkably strong sense that time has a direction. A line traced
in the sand does not by itself de ne a direction. If time is a line, it is
a special one.

The evidence for time’s direction is in the ‘snapshots’. Many

contain memories of other snapshots. We can do a test on time. We
can stop at one of our experienced instants laid out in a line, and see
that it contains a memory. We locate the remembered instant
somewhere in the line. That de nes a direction – from it to the
memory of it. We can do this with other pairs of instants. They
always de ne the same direction. Many other phenomena de ne a
direction. Co ee cools down unless we put it in the microwave; it
never heats up. Cups shatter when we drop them; shards never
reassemble themselves and leap back up onto the table as a whole
cup. All these phenomena, like memories, de ne a direction in time,
and they all point the same way. Time has an arrow.

Thus, experienced time is linear, it can be measured and it has an

arrow. These are not properties of an invisible river: they belong to
concrete instants. Everything we know about time is garnered from
them. Time is inferred from things.

NEWTON’S CONCEPTS

In 1687, Newton created precise notions of space, time and motion.

Despite major revisions, much of his scheme remains intact. It is still
close to the way many people, including scientists, think about time.

Newton’s time is absolute. It ows with perfect uniformity for

ever and nothing in the world a ects its ow. Space, too, is absolute.
Newton conceived of space as a limitless container. It stretches from
in nity to in nity like a translucent block of glass, through which,
nevertheless, objects can move unhindered. Space is a huge arena;
time is a clock in the grandstand. Both are more fundamental than
things. Newton could imagine an empty world but not a world
without space and time. Many philosophers have agreed with him.
So does the proverbial man in the pub, convinced that space goes on
for ever and that ‘there must have been time before the Big Bang’.

At any instant, all the things in the Newtonian world are at

de nite positions. His absolute space performs two distinct roles. As
in the discussion above, it binds, or holds, things together, in one
instant. But it also places them in a container. Imagine taking two-
dimensional snapshots of a table in a room. Paint out the background
room, and you could still reconstruct the form of the three-
dimensional table, but you would not know where to place it.
Newton insisted that the things in the world in any instant have a
de nite place, and he posited absolute space as a kind of room to
provide that place. His xed container persists through time. We
could take real snapshots of the things in the world (Figure 1).
Ideally, these snapshots should be three-dimensional, like space, and
show all things relative to each other and their positions in absolute
space, just as snapshots of a soccer match show the players, ball and
referee on the pitch with its markings. The grandstand clock records
the time.

According to Newton, all bodies move through absolute space in

accordance with de nite laws of motion which govern the speed and
direction of the bodies in that space as measured by absolute time.
The laws are such that if the motions of the bodies are known at
some instant, the laws determine all the future movements. All the
world’s history can be determined from two snapshots taken in quick
succession. (If you know where something is at two closely spaced
instants, you can tell its speed and direction. Two such snapshots
thus encode the future.)
 Figure 1 As explained in the text, Newton conceived of space as a
container, or arena, and time as a uniform ow. The di culty is that
both are invisible. This diagram attempts to represent the way he
thought about space and time. The blank white of the page is a two-
dimensional substitute for the invisible three-dimensional space, and
the e ect of the ow of time is mimicked by supposing that it
triggers light ashes at closely spaced equal intervals of time. These
 ashes illuminate the objects in absolute space at the corresponding
instants of time just as strobe lighting illuminates dancers in a
darkened room. In this computer-generated perspective view, the
vertices of a triangle represent the positions of three mass points as
they move through absolute space. The triangles formed by the
points at successive instants are shown.

Newton’s picture is close to everyday experience. We do not see

absolute space and time, but we do see something quite like them –
the rigid Earth, which de nes positions, and the Sun, whose motion
is a kind of clock. Newton’s revolution was the establishment of strict
laws that hold in such a framework.

LAWS AND INITIAL CONDITIONS

These laws have a curious property. They determine motions only if

certain initial conditions are combined with them. Newton believed
that God ‘set up’ (created) the universe at some time in the past by
placing objects in absolute space with de nite motions; after that,
the laws of motion took over. The statement that Newton’s is a
clockwork universe is a bit misleading. Clocks have one
predetermined motion: the pendulum of the grandfather clock simply
goes backwards and forwards. The Newtonian universe is much more
remarkable, being capable of many motions. However, once an
initial condition has been chosen, everything follows.
 Thus, there are two disparate elements in the scienti c account of
the universe: eternal laws, and a freely speci able initial condition.
Einstein’s relativity and major astronomical discoveries have merely
added to this dual scheme the exciting novelty of a universe
exploding into being about fteen billion years ago. The initial
condition was set at the Big Bang.

Some people question this dual scheme. Is it an immutable

feature? Might we not nd laws that stand alone, without initial
conditions? These questions are particularly relevant because
Newton’s laws (and also Einstein’s theories of relativity, which
replaced them) have a property that seems quite at variance with the
way we feel the universe works – that the past determines the future.
We do not think that causality works from the future to the past.
Scientists always consider initial conditions. But Newton’s and
Einstein’s laws work equally well in both directions. The truth is that
the string of triangles in Figure 1 is determined by Newton’s laws
acting in both directions by any two neighbouring triangles
anywhere along the string. You can persuade yourself of this by
looking at the gure again. It is impossible to say in which direction
time ows. The caption speaks of ‘strobe lighting’ illuminating the
triangles at equal time intervals, but does not say which is
illuminated rst. Scientists could examine the triangles until the
crack of doom but could never nd which came rst. This is related
to one of the biggest puzzles in science.

WHY IS THE UNIVERSE SO SPECIAL?

The universe we see around us today is special: it is very highly
ordered. For example, light streams away in a very regular ow from
billions upon billions of stars throughout the universe. These stars
are themselves collected together in galaxies, of which there are just
a few basic types. Here on Earth we nd very complex molecules and
very complicated life forms that could not possibly exist were it not
for the steady stream of sunlight that constantly bathes our planet.
However, the vast majority of conceivable initial conditions there
could have been at the Big Bang would have led to universes much
less interesting – indeed, positively dull – compared with ours. Only
an exceptional initial condition could have led to the present order.
That is the puzzle. Modern science is in the remarkable position of
possessing beautiful and very well tested laws without really being
able to explain the universe. In the dual scheme of laws and initial
conditions, the great burden of explaining why the universe is as it is
falls to the initial conditions. Science can as yet give no explanation
of why those conditions were as they must have been to explain the
presently observed universe. The universe looks like a uke.

There are two remarkable things about the order in the universe:
the amount of it and the way it degrades. One of the greatest
discoveries of science, made about a hundred and fty years ago,
was the second law of thermodynamics. Studies of the e ciency
with which steam engines turn heat into mechanically useful motion
led to the concept of entropy. As originally discovered, this is a
measure of how much useful work can be got out of hot gas, say. It is
here that the arrow of time, which we know from direct experience,
enters physics. Almost all processes observed in the universe have a
directionality. In an isolated system, temperature di erences are
always equalized. This means, for example, that you cannot extract
energy from a cooler gas to make a hotter gas even hotter and chu
along in your steam engine even faster. More strictly, if you did, you
would degrade more energy than you gain and nish up worse o .

I have already mentioned the unidirectional process of a cup

breaking. Another is mixing cream with co ee. It is virtually
impossible to reverse these processes. This is beautifully illustrated
by running a lm backwards: you see things that are impossible in
the real world. This unidirectionality, or arrow, is precisely re ected
in the fact that the entropy of any isolated system left to itself always
increases (or perhaps stays constant).

It was recognized in the late nineteenth century that this

unidirectionality of observed processes was in sharp con ict with the
fact that Newton’s laws should work equally well in either time
direction. Why do natural processes always run one way, while the
laws of physics say they could run equally well either way? For four
decades, from 1866 until his suicide on 5 September 1906 in the
picturesque Adriatic resort of Duino, the Austrian physicist Ludwig
Boltzmann attempted to resolve this con ict. He introduced a
theoretical de nition of entropy as the probability of a state. He
rmly believed in atoms – the existence of which remained
controversial until the early years of the twentieth century –
conceived of as tiny particles rushing around at great speed in
accordance with Newtonian laws. Heat was assumed to be a measure
of the speed of atoms: the faster the atoms, the hotter the substance.
By the second half of the nineteenth century, physicists had a good
idea of the immense number of atoms (assuming that they existed)
there must be even in a grain of sand, and Boltzmann, among others,
saw that statistical arguments must be used to describe how atoms
behave.

He asked how probable a state should be. Imagine a grid of 100

holes into which you drop 1000 marbles at random. It is hugely
improbable that they will all nish up in one hole. I am not going to
give numbers, but it is simple to work out the probability that all
will land in one hole or, say, in four adjacent holes. In fact, one can
list every possible distribution of the marbles in the grid, and then
see in how many of these distributions all the marbles fall in one
hole, in four adjacent holes, eight adjacent holes, and so on. If each
distribution is assumed to be equally probable, the number of ways a
particular outcome can happen becomes the relative probability of
that outcome, or state. Boltzmann had the inspired idea that, applied
to atoms, this probability (which must also take into account the
velocities of the atoms) is a measure of the entropy that had been
found through study of the thermodynamics of steam engines.

There is no need to worry about the technical details. The

important thing is that states with low entropy are inherently
improbable. Boltzmann’s idea was brilliantly successful, and much of
modern chemistry, for example, would be unthinkable without it.
However, his attempt to explain the more fundamental issues
associated with the unidirectionality of physical processes was only
partly successful.

He wanted to show that, matching the behaviour of macroscopic

entropy, his microscopic entropy would necessarily increase solely by
virtue of Newton’s laws. This seems plausible. If a large number of
atoms are in some unlikely state, say all in a small region, so that
they have a low entropy, it seems clear that they will pass to a more
probable state with higher entropy. However, it was soon noted that
there are exactly as many dynamically possible motions of the atoms
that go from states of low probability to states of high probability as
vice versa. This is a straight consequence of the fact that Newton’s
laws have the same form for the two directions of time. Newton’s
laws alone cannot explain the arrow of time.

Only two ways have ever been found to explain the arrow: either
the universe was created in a highly unlikely special state, and its
initial order has been ‘degrading’ ever since, or it has existed for
ever, and at some time in the recent past it entered by chance an
exceedingly improbable state of very low entropy, from which it is
now emerging. The second possibility is entirely compatible with the
laws of physics. For example, if a collection of atoms (which obey
Newton’s laws) is con ned in a box and completely isolated, it will,
over a su ciently long period of time, visit (or rather come
arbitrarily near) all the states that it can in principle ever reach, even
those that are highly ordered and statistically very unlikely.
However, the intervals of time between returns to states of very low
entropy are stupendously long (vastly longer than the presently
assumed age of the universe), and neither explanation is attractive.

The fact is that mechanical laws of motion allow an almost

incomprehensibly large number of di erent possible situations.
Interesting structure and order arise only in the tiniest fraction of
them. Scientists feel they should not invoke miracles to explain the
order we see, but that leaves only statistical arguments, which give
bleak answers (only dull situations can be expected), or the so-called
anthropic principle that if the world were not in a highly structured
but extremely unlikely state, we should not exist and be here to
observe it.

One of my reasons for writing this book is that timeless physics

opens up new ways of thinking about structure and entropy. It may
be easier to explain the arrow of time if there is no time!
CHAPTER2
Time Capsules

THE PHYSICAL WORLD AND CONSCIOUSNESS

The discussion in Chapter 1 prompts the question of how our sense

of the passage of time arises. Before we can begin to answer this, we
have to think about another mystery – consciousness itself. How does
brute inanimate matter become conscious, or rather self-conscious?

No one has any idea. Consciousness and matter are as di erent as

chalk and cheese. Nothing in the material world gives a clue as to
how parts of it (our brains) become conscious. However, there is
increasing evidence that certain mental states and activities are
correlated with certain physical states in di erent speci c regions of
the brain. This makes it natural to assume, as was done long ago,
that there is psychophysical parallelism: conscious states somehow
re ect physical states in the brain.

Put in its crudest form, a brain scientist who knew the state of
our brain would know our conscious state at that instant. The brain
state allows us to reconstruct the conscious state, just as musical
notes on paper can be transformed by an orchestra into music we can
hear. By the ‘state’ of a system, say a collection of atoms, scientists
usually mean the positions of all its parts and the motions of those
parts at some particular instant. It is widely assumed that conscious
states, in which, after all, we are aware of motion directly, are at the
least correlated with (correspond to) brain states that involve not
only instantaneous positions but also motions and, more generally,
change (associated with ow of electric currents or chemicals, for
example). This is a natural assumption. Our awareness of motion and
change is vivid and often exciting: think of watching gymnastics, or
the 100-metre sprint nal in the Olympic Games. We suppose that
the impression of motion must be created by some motion or change
in the brain.

However, if the physical processes in the brain are controlled by

laws like Newton’s, such an assumption runs up against the problem
that they distinguish no direction of time. Figure 1, with its
impossibility of saying in which direction time ows, makes this
clear. It is no help to go from its three particles to billions of them.
Observed e ects should have a real cause. The chain from cause to
e ects may be quite long and take surprising forms, but a cause there
must be. It is unsatisfactory to suppose that we have a direct
awareness of an invisible ow of time. Our sense of the passage of
time and, even more basically, of seeing motion and knowing its
direction, ought to have a cause we can get our hands on.

The lack of time direction in the bare laws of motion led

Boltzmann to a remarkable suggestion (quoted in the Notes). As we
have seen, Newtonian systems can enter highly ordered phases.
These are exceptionally rare periods separated by ‘deserts’ of
monotony. Nevertheless, every now and then a system will enter one.
Its entropy will go down, reach a minimum, and then start to
increase.

We should not think of this happening in a de nite direction of

time. Instead, we should picture the states of the system strung out in
a line, as in Figure 1, which we could ‘walk along’ in either direction.
Every now and then, with immense stretches between them, we will
come upon regions in which the entropy decreases and the order
increases. Then the entropy will start to increase again. Someone
‘walking’ in the opposite direction would have the same experience.
Now, such a line of states can represent the entire universe,
including human beings. Since we are very complicated and exhibit
much order, we can be present only in the exceptional regions of low
entropy.

Boltzmann’s suggestion, startling when rst encountered, was

that conscious beings could exist on either side of a point of lowest
entropy, and that the beings on both sides would regard that point as
being in their past. Time would seem to increase in both directions
from it. In this view, time itself neither ows nor has a direction; it is
at most a line. It is only the instantaneous con gurations of matter,
strung out like washing on the line, that very occasionally suggest
that time has a direction associated with it. The direction is in the
washing, not the line. What is more, depending on the position in the
line, the ‘arrow’ will point in opposite directions.
 This, then, gives a genuine cause for our awareness of motion and
the passing of time. The conscious mind, in any instant, is actually
aware of a short segment of the ‘line of time’, along which there is an
entropy gradient. Time seems to ow in the direction of increasing
entropy. Interestingly, consciousness and understanding are always
tied to a short time span, which was called the specious present by
the philosopher and psychologist William James (brother of novelist
Henry). The specious present is closely related to the phenomenon of
short-term memory and our ability to grasp and understand
sentences, lines of poems and snatches of melody. It has a duration
of up to about three seconds.

The key element in Boltzmann’s idea is comparison of structures.

There needs to be qualitative change in the brain patterns along a
segment of the ‘line of time’. If the brain pattern in each instant is
likened to a card, then the patterns become a pack of cards, and our
conscious experience of time ow arises (somehow) from the change
of pattern across the pack. Though we may not understand the
mechanism, the e ect does have a cause.

To summarize: Newtonian time is an abstract line with direction

– from past to future. Boltzmann keeps the line but not the direction.
That belongs to the ‘washing’. But do we need the line?

TIME WITHOUT TIME

Perhaps not. The brain often fools us. When we rst look at certain
drawings, they appear to represent one thing. After a while, the
image ickers and we see something di erent. The reason is well
understood: the brain processes information before we get it. We do
not see things as they are but as the brain interprets them for us.
There are very understandable reasons for this, but the fact remains
that we are often fooled by such ‘deceptions’.

Could all motion be a similar deception? Suppose we could freeze

the atoms in our brains at some instant. We might be watching
gymnastics. What would brain specialists nd in the frozen pattern of
the atoms? They will surely nd that the pattern encodes the
positions of the gymnasts at that instant. But it may also encode the
positions of the gymnasts at preceding instants. Indeed, it is virtually
certain that it will, because the brain cannot process data
instantaneously, and it is known that the processing involves
transmission of data backwards and forwards in the brain.
Information about the positions of the gymnasts over a certain span
of time is therefore present in the brain in any one instant.

I suggest that the brain in any instant always contains, as it were,

several stills of a movie. They correspond to di erent positions of
objects we think we see moving. The idea is that it is this collection
of ‘stills’, all present in any one instant, that stands in psychophysical
parallel with the motion we actually see. The brain ‘plays the movie
for us’, rather as an orchestra plays the notes on the score. I am not
going to attempt to elaborate on how this might be done; all I want
to do is get the basic idea across. There are two parts to it. First, each
instantaneous brain pattern contains information about several
successive positions of the objects we see moving in the world. These
successive positions need correspond only to a smallish fraction of a
second. Second, the appearance of motion is created by the
instantaneous brain pattern out of the simultaneous presence of
several di erent ‘images’ of the gymnasts contained within it (Figure
2). This happens independently of the earlier and later brain states.

Figure 2 My explanation of how it might be possible to ‘see’ motion

when none is there is illustrated in this chronophotograph of a
sideways jump. My assumption is that the pattern of the atoms in our
brain encodes, at any instant, about six or seven images of the
gymnast. The standard ‘temporal’ explanation is that the gymnast
passes through all these positions in a fraction of a second. My idea
is that when we think we are seeing actual motion, the brain is
interpreting all the simultaneously encoded images and, so to speak,
playing them as a movie.

This proposal is not so very di erent from Boltzmann’s idea that

the sense of motion is created from several qualitatively di erent
patterns arranged along the ‘line of time’. Instead, I am suggesting
that it is created by the brain from the juxtaposition of several
subpatterns within one pattern. The arrow of time is not in the
washing line, it is not in several pieces of washing, it is in each piece.
If we could preserve one of these brain patterns in aspic, it would be
perpetually conscious of seeing the gymnasts in motion. If you nd
this idea a bit startling, I am glad because I nd it does bring home
the ‘freezing of motion’ that I think we have to contemplate. In fact,
since brain function and consciousness are elds in which I have no
expertise, I would like you to regard this suggestion in the rst place
as a means of getting across an idea, the main application of which I
see in physics.

To that end, I want to introduce the notion of special Nows, or

time capsules, as I call them.

TIME CAPSULES

By a time capsule, I mean any xed pattern that creates or encodes

the appearance of motion, change or history. It is easiest to explain
the idea by examples, for example the Ariel in the storm in Turner’s
painting. Although they are all static in themselves, pictures often
suggest that something has happened or is happening – with a
vengeance in this painting. But in reality it simply is. I know no
better example of something static that gives the impression of
motion.

In pictures, the impression is deliberately created. Much more

signi cant for my purposes are time capsules that arise naturally and
have to be interpreted, by the examination of records they seem to
contain. Records, or apparent records, play a vital role in my idea
that time is an illusion. I use records primarily in the sense of, for
example, fossils, which occur naturally and are interpreted by us as
relics of things that actually existed. Less directly, all geological
formations, rock strata in particular, are now invariably interpreted
by geologists as constituting a record (to be interpreted) of past
geological processes. Finally, there are records that people create
deliberately: doctors’ notes, minutes of committee meetings,
astronomical observations, photographs, descriptions of the initial
and nal conditions of controlled experiments, and so on. All such
things, and many more, I call records. My position is that the things
we call records are real enough, and so is their structure. They are
the genuine cause of our belief in time. Our only mistake is the
interpretation: time capsules have a cause, but time is no part of it.

Let me now attempt a more formal de nition. Any static

con guration that appears to contain mutually consistent records of
processes that took place in a past in accordance with certain laws
may be called a time capsule. From my point of view, it is
unfortunate that the dictionary de nition (in Webster’s) of a time
capsule is ‘a container holding historical records or objects
representative of current culture that is deposited (as in a
cornerstone) for preservation until discovery by some future age’. I
do not mean that. But we have all had the experience of walking into
a house untouched by historical development for decades or
centuries and declaring it to be a perfect time capsule. This, I
believe, happens to us in each instant of time we experience. The
only di erence is that we experience our current time capsule, not
someone else’s. And we are mistaken in the way we interpret the
experience.

It is important for me that, as I point out in the next section, the

phenomenon of time capsules is very widespread in the physical
world, and is not restricted to our mental states and experiences. In
addition to my caveat at the end of the previous section, I should
emphasize that I am not claiming consciousness plays some
remarkable novel or extraphysical role in the world. Unlike Roger
Penrose in his best-seller The Emperor’s New Mind, I am not
suggesting that there is any ‘new physics’ associated with mental
states. There may be, but that is not part of my time-capsule idea.
However, I do believe we have to think carefully about the role of
consciousness in the picture that we form of the world.

First, all knowledge and theorizing comes to us through the

conscious state. If we want to form an overall picture of things, we
cannot avoid allotting a place to consciousness. It is necessary for
completeness: we have to consider ‘where we stand’. This is closely
related to a second factor. Viewed as a physical system, the brain is
organized to an extraordinary degree. It is vastly more complicated
and intricate than the air we breathe or the star clusters we see
through telescopes. There may not be any locations anywhere in the
universe that are more subtly and delicately organized than human
brains. There is not merely the brain structure as such, but also the
distillation of accumulated human experience and culture that we
carry in our brains. But this very organization may be giving us a
distorted picture of the world. If you stand, like Turner bound to the
Ariel’s mast, in the tornado’s maelstrom, you might well suspect that
the universe is just one great whirlpool.

The lesson we learned from Copernicus, Kepler and Galileo is

here very relevant. They persuaded us, against what seemed to be
overwhelming evidence to the contrary, that the Earth moves. They
taught us to see motion where none appears. The notion of time
capsules may help us to reverse that process – to see perfect stillness
as the reality behind the turbulence we experience.

Stand, as 1 have with a daughter, and look at Jupiter against the

winter stars. Every clear frosty night we stood on the utterly
motionless Earth – as it appeared to our senses – and watched
through the winter as Jupiter, high in the sky, tracked night by night
eastwards against the background of the stars. But then Jupiter
slowed down, came to a stop, and went backwards in the retrograde
motion that so puzzled the ancients. Then this motion stopped, and
the eastwards motion recommenced. In all this Jupiter moved, not
us. We could see it with our eyes. Seeing is believing. But what did
Copernicus say? We must be careful not to attribute to the heavens
(Jupiter) what is truly in the Earth-bound observer. I could persuade
my daughter that the motion of the Earth, not of Jupiter, gives rise to
the retrograde motion. To interpret events, we must know where we
stand and understand how that a ects what we witness. But we
observe the universe from the middle of a most intricate processing
device, the human brain. How does that a ect our interpretation of
what we see?

EXAMPLES OF TIME CAPSULES

As a rst example, we can stay within the brain but consider long-
term memory. A game we sometimes play at Christmas brings out
the importance of mutually consistent records held in structures.
Fifty events in recent world history are written down on separate
cards without dates attached. Players are divided into teams and
given the cards jumbled up. The challenge is to put them in the
correct chronological order. The only resource each team has to
attack the task is their collective long-term memories, which every
good realist (myself included) will surely agree are somehow or
other ‘hard-wired’ into their brains. How each team fares depends on
the consistency of its members’ recollections – the records in their
brains.
 This example shows clearly that all we know about the past is
actually contained in present records. The past becomes more real
and palpable, the greater the consistency of the records. But what is
the past? Strictly, it is never anything more than we can infer from
present records. The word ‘record’ prejudges the issue. If we came to
suspect that the past is a conjecture, we might replace ‘records’ by
some more neutral expression like ‘structures that seem to tell a
consistent story’.

The relevance of this remark is brought home by the sad

examples of brain damage that takes away the ability to form new
memories but leaves the existing long-term memory intact. One
patient, still alive, retains good memory and a sense of himself as he
was before an operation forty years ago, but the rest is blank. It is
possible to have meaningful discussions about what are for him
current events even though they are all those years away, but the
next day he has no recollection of the discussion. The mature brain is
a time capsule. History resides in its structure.

After our own brains, the most beautiful example of a time

capsule that we know intimately is the Earth – the whole Earth.
Above all, I am thinking of the geological and fossil records in all
their multifarious forms. What an incredible richness of structure is
there, and how amazingly consistent is the story it tells. I nd it
suggestive that it was the geologists – not the astronomers or
physicists – who rst started to suggest an enormous age for the
Earth. They were the discoverers of deep time, which did start as
conjecture. And it was all read o from rocks, most of which are still
with us now, virtually unchanged from the form they had when the
geologists reached those conclusions. The story of the antiquity of
the Earth and of its creation from supernova debris – the Stardust
from which we believe we ourselves are made – is a story of patient
inference built upon patient inference based upon marks and
structures in rocks. On this rock – the Earth in all its glory – the
geologists have built the history of the world, the universe even.

What is especially striking about the Earth is the way in which it

contains time capsules nested within time capsules, like a Russian
doll. Individual biological cells (properly interpreted) are time
capsules from which biologists read genetic time. Organs within the
body are again time capsules, and contain traces of the history and
morphogenesis of our bodies. The body itself is a time capsule.
History is written in a face, which carries a date – the approximate
date of our birth. We can all tell the rough age of a person from a
glance at their face. Wherever we look, we nd mutually consistent
time capsules – in grains of sand, in ripe cherries, in books in
libraries. This consistent meshing of stories even extends far from the
Earth and into the outermost reaches of the universe. The
abundances of the chemical elements and isotopes in the gas of stars
and the waters of the oceans tell the story of the stars and a Big Bang
that created the lightest elements. It all ts together so well.

For me, two facts above all stand out from this miracle of nature.
If we discount the direct perception of motion in consciousness, all
this fantastic abundance of evidence for time and history is coded in
static con gurational form, in structures that persist. This is the rst
fact, and it is ironic. The evidence for time is literally written in
rocks. This is why I believe the secret of time is to be unravelled
through the notion of time capsules. It is also the reason why I seek
to reduce the other hard and persistent evidence for time and motion
– our direct awareness of them in consciousness – to a time-capsule
structure in our brains. If I can make such a structure responsible for
our short-term memory – the phenomenon of the specious present –
and for the actual seeing of motion, then all appearances of time will
have been reduced to a common basis: special structure in individual
Nows.

The second fact that needs to be taken on board is the sheer

creativity of Nature. How does Nature create this rich, rich structure
that speaks to us so insistently and consistently of time? How could it
and we come to be if there is no time? The appearance of time is a
deep reality, even without the motion we see and the passage of time
we sense in consciousness. It is written all over the rocks. Any
plausible account of the universe must, rst and foremost, explain
the existence of the structures we see and the semantic freight (i.e.
the seemingly meaningful story) that they carry.

If we can explain how they arise, time capsules o er the prospect

of a much more radical explanation of the properties of time than
Boltzmann’s account of the origin of its arrow. To explain the
appearance of an arrow, he still had to assume a succession of
instants strung out along a ‘line of time’. I have already suggested
that the line may be redundant. The inference that it exists can
emerge from a single Now. The instant is not in time – time is in the
instant.
CHAPTER 3
A Timeless World

FIRST OUTLINE

Now I want to start on the attempt to show you that, at least as a

logical possibility, the appearance of time can arise from utter
timelessness. I shall do this by comparing two imaginary exercises. I
begin by presenting you with two bags, labelled Current Theory and
Timeless Theory. When you open them up, you nd that each bag is
lled with cardboard triangles, all jumbled up. Now, triangles come
in all shapes and sizes. The rst thing you notice is that the rst bag
contains far fewer triangles than the second. Closer examination
reveals that the two collections are very di erent. Let me begin by
describing the contents of Current Theory.

First, you notice that it contains triangles of all di erent sizes.

There is a smallest triangle, very tiny; then another very like it, but a
little larger and with a slightly di erent shape; and so on. In fact,
you soon realize that you can lay out all the triangles in a sequence.
The order in which they should go is clear because each successive
triangle di ers only slightly from its predecessor. Their increasing
size makes the ordering especially easy. Of course, a real bag can
contain only nitely many triangles, but I shall suppose that there
are in nitely many and that the sequence is endless, the triangles
getting ever larger.

Such a sequence of triangles is like the sequence of experienced

instants that I suggested ‘photographing’. It is also like the succession
of Newtonian instants from the moment God decided to create the
universe, or the succession of states of the universe expanding out of
the Big Bang, represented by the smallest triangle. In fact, the
contents of Current Theory correspond to the simplest Newtonian
universe that can begin to model the complexity of the actual
universe: three mass points moving in absolute space and time, as in
Figure 1. Initially very close to each other, they move apart so
rapidly that gravity cannot pull them back, and they y o to
in nity.

According to Newton, the three mass points are, at all instants, at

certain positions in absolute space and form certain triangles. The
triangles tell us how the points are placed relative to one another,
but not where they are in absolute space. It is such triangles,
represented in cardboard, that I imagine have been put into the
Current Theory bag. Since we cannot experience absolute space and
time directly, I have tried to match the model more closely to our
actual experience. The sequence of triangles corresponds to one
possible history. There could be many such histories that match the
dual scheme of laws and initial conditions. But we nd only one in
the Current Theory bag.
 Next, we examine the Timeless Theory bag. There are two big
di erences. First, it contains vastly more triangles (it could, in fact,
contain all conceivable triangles). More signi cantly, there are so
many of them that it is quite impossible to arrange them in a
continuous sequence. Second, the triangles are present in multiple
copies. That is, we might, after a very extensive search, nd ten
identical copies of one particular triangle, two of another, and ten
million of yet another. That is really the complete story. It is all that
most people would notice.

I think you will agree that the Current Theory bag does match
experience quite closely. The triangles stand for each of the instants
you experience, and they follow one another continuously, just as the
instants do. By giving them to you in a bag and getting you to lay
them out in a sequence, I am giving you a ‘God’s eye’ view of history.
All its instants are, as it were, spread out in eternity as if you
surveyed them from a mountain-top. In fact, this way of thinking
about time has long been a commonplace among Christian
theologians and some philosophers, and has prompted them to claim
that time does not exist but that its instants all exist together and at
once in eternity. My claim is much stronger. I am saying that reality,
if we could see all of it, is not at all like the contents of the Current
Theory bag with its single sequence of states. It is like the contents of
the Timeless Theory bag, in which in principle all conceivable states
can be present. Nothing in it resembles our experience of history as a
unique sequence of states: that experience is usually explained by
assuming that there is a unique sequence of states. I deny that there
is such a sequence, and propose a di erent explanation for the
experience that prompts us to believe in it. The only thing the bags
have in common with our direct experience of time is the parallel
between individual triangles as models of individual instants of time.

Actually, the bags share another property – their contents satisfy

a law. Given the sequence of triangles of the rst bag, clever
mathematicians could deduce that they correspond to the triangles
formed by three gravitationally interacting bodies. They could even
reconstruct the bodies’ positions in absolute space, and the amount
of time that elapses between any two of the triangles in the
sequence. With the second bag, mathematicians would discover that
the numbers in which the di erent triangles occur are not random –
chosen by chance – but satisfy a law. The numbers vary from triangle
to triangle in an ordered fashion. But at rst glance at least, this law
seems to have no connection with the law that creates the unique
sequence of triangles in the rst bag. Also, there is nothing like the
dual scheme of law and initial condition that creates the sequence of
the rst bag. In a sense that I shall not yet try to explain, there is just
a law, with nothing like an initial condition that has to be added to
it.

How is the appearance of time ever going to emerge from the

contents of the Timeless Theory bag as just described? Bare triangles
lying in a jumbled heap certainly cannot make that miracle happen.
Triangles have a structure that is much too simple. This is why I said
that rich structure ordered in a special way is an essential element if
a notion of time is to emerge. If, when we open the Timeless Theory
bag, we nd it contains, not triangles, but vastly richer structures,
some of which are time capsules in the sense I have de ned, my task
does not seem quite so hopeless. By de nition, time capsules suggest
time. But nding just a few time capsules in a vast heap of otherwise
nondescript structures will not get me very far.

This is where the assumption that all the structures found in the
bag come in multiple copies, and that the numbers of these copies,
which can vary very widely, are determined by a de nite timeless
rule, becomes crucial. Imagine that all the structures for which the
numbers of identical copies in the bag are large are time capsules,
while there are few copies of structures that are not time capsules.
Since the overwhelming majority of possible structures that can exist
are certainly not time capsules, any rule that does ll the bag with
time capsules will be remarkably selective, creative one might say. If,
in addition, you can nd evidence that the universe is governed by a
timeless law whose e ect is to discriminate between structures and
which actually selects time capsules with surprising accuracy, then
you might begin to take such ideas more seriously. You might begin
to see a way in which the Timeless Theory could still explain our
experience of time, and could perhaps be superior to the Current
Theory.

However, you will probably dismiss such a possibility as the

wildest fantasy. Why should Nature go to such contrived lengths
simply to create an impression of time and fool poor mortals? To
counter this natural reaction, let me give a little more detail about
those hints of the non-existence of time that I mentioned in Chapter
1. This may at least persuade you that some dramatic change could
be in the o ng.

THE CRISIS OF TIME

Physics is regarded as the most fundamental science. It is an attempt

to create a picture of reality as we should see it if we could,
somehow, step out of ourselves. For this reason it is rather abstract.
In addition, it often deals with conditions far removed from everyday
human experience – deep inside the atom, where quantum theory
holds sway, and in the far ung reaches of space, where Einstein’s
general relativity reigns. The ideas I want to tell you about have
come from attempts during the last forty years to unite these two
realms (Box 2). They have produced a crisis. The very working of the
universe is at stake: it does not seem to be possible, in any natural
and convincing way, to give a common description of them in which
anything like time occurs.

Frustratingly little progress has been made. However, in 1967 a

possible picture did emerge from a paper by the American Bryce
DeWitt. He found an equation that, if his reasoning is sound,
describes the whole universe – both atoms and galaxies – in a uni ed
manner. Because John Wheeler, the American physicist who coined
the term ‘black hole’, played a major part in its discovery, this
equation is called the Wheeler-DeWitt equation. It is controversial in
at least three respects. First, many experts believe that the very
derivation of the equation is awed – that it was obtained by an
invalid procedure. Second, the equation is not yet even properly
de ned, as there are still many technical di culties to be overcome.
In fact, it is more properly regarded as a conjecture: a tentative
proposal for an equation that is not yet proved. And third, the
experts argue interminably over what meaning it might have and
whether it can ever be promoted to the status of a bona de
equation. Ironically, DeWitt himself thinks that it is probably not the
right way to go about things, and he generally refers to it as ‘that
damned equation’. Many physicists feel that a di erent route,
through so-called superstring theory, which it is hoped will establish
a deep unity between all the forces of nature, is the correct way
forward. That many of the best physicists have concentrated on
superstring theory is probably the main reason why the ‘crisis of
time’ brought to light by the Wheeler-DeWitt equation has not
attracted more attention. However, there is no doubt that the
equation re ects and uni es deep properties of both quantum theory
and general relativity. Quite a sizeable minority of experts take the
equation seriously. In particular, much of the work done by Stephen
Hawking in the last twenty years or so has been based on it, though
he has his own special approach to the problem of time that it raises.

For now, all I want to say about the Wheeler-DeWitt equation is

that if one takes it seriously and looks for its simplest interpretation,
the picture of the universe that emerges is like the contents of the
Timeless Theory bag. For a long time, physicists shied away in
distrust from its apparently timeless nature, but during the last
fteen years or so a small but growing number of physicists, myself
included, have begun to entertain the idea that time truly does not
exist. This also applies to motion: the suggestion is that it too is pure
illusion. If we could see the universe as it is, we should see that it is
static. Nothing moves, nothing changes. These are large claims, and
the bulk of my book will discuss the arguments from physics
(presented as simply as I can) that lead me and others to such
conclusions. At the end, I shall outline, through the notion of time
capsules, a theory of how a static universe can nevertheless appear to
teem with motion and change.

Now I want to give you a better feel for what a timeless universe
could be like. What we need rst is a proper way to think about
Nows.

THE ULTIMATE ARENA

One issue that runs through this book is this: what is the ultimate
arena of the universe? Is it formed by space and time (space-time), or
something else? This is the issue raised by Dirac’s sentence I quoted
in the Preface: ‘This result has led me to doubt how fundamental the
four-dimensional requirement in physics is.’ I believe that the
ultimate arena is not space-time. I can already begin to give you an
idea of what might come in its place.

I illustrated the Newtonian scheme by a model universe of just

three particles. Its arena is absolute space and time. The Newtonian
way of thinking concentrates on the individual particles: what counts
are their positions in space and time. However, Newton’s space and
time are invisible. Could we do without them? If so, what can we put
in their place? An obvious possibility is just to consider the triangles
formed by the three particles, each triangle representing one possible
relative arrangement of the particles. These are the models of Nows I
asked you to contemplate earlier. We can model the totality of Nows
for this universe by the totality of triangles. It will be very helpful to
start thinking about this totality of triangles, which is actually an
in nite collection, as if it were a country, or a landscape.

If you go to any point in a real landscape, you get a view. Except

for special and arti cial landscapes, the view is di erent from each
point. If you wanted to meet someone, you could give them a
snapshot taken from your preferred meeting point. Your friend could
then identify it. Thus, points in a real country can be identi ed by
pictures. In a somewhat similar way, I should like you to imagine
Triangle Land. Each point in Triangle Land stands for a triangle,
which is a real thing you can see or imagine. However, whereas you
view a landscape by standing at a point and looking around you,
Triangle Land is more like a surface that seems featureless until you
touch a point on it. When you do this, a picture lights up on a screen
in front of you. Each point you touch gives a di erent picture. In
Triangle Land, which is actually three-dimensional, the pictures you
see are triangles. A convenient way of representing Triangle Land is
portrayed in Figures 3 and 4.
 I have gone to some trouble to describe Triangle Land because it
can be used to model the totality of possible Nows. Like real
countries, and unlike absolute space, which extends to in nity in all
directions, it has frontiers. There are the sheets, ribs and apex of
Figure 4. They are there by logical necessity. If Nows were as simple
as triangles, the pyramid in Figure 4 could be seen as a model of
eternity, for one notion of eternity is surely that it is simply all the
Nows that can be, laid out before us so that we can survey them all.
Figure 3 The seven triangles represent several possible arrangements
of a model universe of three particles A, B, C. Each triangle is a
possible Now. Each Now is associated with a point (black diamond)
in the ‘room’ formed by the three grid axes AB, BC, CA, which meet
at the corner of the ‘room’ farthest from you. The black diamond that
represents a given triangle ABC is situated where the distance to the
‘ oor’ is the length of the side AB (measured along the vertical axis),
and the distances to the two ‘walls’ are equal to the other two sides,
BC and CA. The dash-dotted lines show the grid coordinates. In this
way, each model Now is associated with a unique point in the
‘room’. As explained in the text, if you ‘touched’ one of the black
diamonds, the corresponding triangle would light up. However, not
every point in the ‘room’ corresponds to a possible triangle – see
Figure 4.

A three-particle model universe is, of course, unrealistic, but it

conveys the idea. In a universe of four particles, the Nows are
tetrahedrons. Whatever the number of particles, they form some
structure, a con guration. Plastic balls joined by struts to form a rigid
structure are often used to model molecules, including
macromolecules such as DNA, which are ‘megamolecules’. You can
move such a structure around without changing its shape. For any
chosen number of balls, many di erent structures can be formed.
That is how I should like you to think about the instants of time.
Each Now is a structure.
 Figure 4 This shows the same ‘room’ and axes as in Figure 3, but
without the walls shaded. Something more important is illustrated
here. In any triangle, no one side can be longer than the sum of the
other two. Therefore, points in the ‘room’ in Figure 3 for which one
coordinate is larger than the sum of the other two do not correspond
to possible triangles. All triangles must have coordinates inside the
‘sheets’ spanned between the three ‘ribs’ that run (towards you) at
 45° between the three pairs of axes AB, BC (up to the left), AB, CA
(up to the right) and BC, CA (along the ‘ oor’, almost towards you).
Points outside the sheets do not correspond to possible triangles.
However, points on the sheets, the ribs and the apex of the pyramid
formed by them correspond to special triangles. If vertex A in the
thin triangle at the bottom right of Figure 3 is moved until it lies on
BC, the triangle becomes a line, which is still just a triangle, because
BC is now equal to (but not greater than) the sum of CA and AB.
Such a triangle is represented by a point on one of the ‘sheets’ in
Figure 4. If point A is then moved, say, towards 8, the point
representing the corresponding triangle in Figure 4 moves along the
‘sheet’ to the corresponding ‘rib’, which represents the even more
special ‘triangles’ for which two points coincide. Finally, the apex,
where the three ribs meet in the far corner of the ‘room’, corresponds
to the unique and most special case in which all three particles
coincide. Thus, Triangle Land has a ‘shape’ which arises from the
rules that triangles must satisfy. The unique point at which the three
particles coincide I call Alpha.

For each de nite collection of structures – triangles, tetrahedrons,

molecules, megamolecules – there is a corresponding ‘country’ whose
points correspond to them. The points are the possible
con gurations. Each con guration is a possible thing; it is also a
possible Now. Unfortunately it is impossible to form any sort of
picture of even Tetrahedron Land: unlike Triangle Land, which has
three dimensions, it has six dimensions. For megamolecules, one
needs a huge number of dimensions. In Tetrahedron Land you could
‘move about’ in its six dimensions. As in my earlier example, the way
to think about its individual points is that if you were to touch any
one of them, a picture of the tetrahedron to which it corresponds
would ‘light up’. In any Megamolecule Land, with its vast number of
dimensions, ‘touching a point’ would cause the corresponding
megamolecule to ‘light up’. The more complicated the structures, the
greater the number of dimensions of the ‘land’ that represents them.
However, the structures that ‘light up’ are themselves always three-
dimensional.

You do not need to try to imagine these much larger spaces –

Triangle Land will do. I hope you do not nd it a dull structure or
too hard to grasp. It is, in fact, an example of a very basic notion in
physics called a con guration space that is normally regarded as too
abstract to attempt to explain in books for non-scientists. But I
cannot begin to get across to you my vision of a timeless universe
without this concept. If you can get your mind round this concept –
and I do encourage you to try – you will certainly understand a lot of
my book. The notion of con guration space opens up a wonderfully
clear way to picture, all at once, everything that can possibly be.

It will also give us new notions of time and history, stripping

away and revealing as redundant the Newtonian superstructure. The
observable history of a three-particle universe, when the invisible
absolute space and time are abstracted away, is just a continuous
sequence of triangles. Suppose we are given such a history. We can
then mark, or plot, the points in Triangle Land that correspond to the
triangles. We shall obtain a curve that winds around within the
pyramid in Figure 4. In this new picture, history is not something
that happens in time but a path through a landscape. A path is just a
continuous track of points in a land. In this book I use the word path
very often in the generalized sense of a continuous series of
con gurations taken by some system (consisting, usually, of material
points). Understood in this sense, paths are possible histories. There
is no time in this picture.

Paths highlight the dilemma brought to light by Boltzmann’s

work. On any path, you can call the point where you stand Now. But
you can walk along a path in either direction. There is nothing in the
notion of a path that can somehow make it a one-way street. You can
also see that the notion of a moving present may be redundant. You
might try to represent it by a spot of light moving along the path,
making each successive point on the path into the present Now, and
therefore more real than the ‘past’, through which the spot has
already passed, and the ‘future’, which the spot has not yet reached.
But if, as I have suggested, all our conscious experiences have their
origin in real structure within the Nows, we can do without the
ction of the moving present. The sense we have that time has
advanced to the present Now is simply our awareness of being in
that Now. Di erent Nows give rise to di erent experiences, and
hence to the impression that the time in them is di erent.
 I need a name for the land of Nows. Plato, who lived about a
century after Heraclitus and Parmenides, taught that the only real
things are forms or ideas: perfect paradigms, existing in a timeless
realm. In our mortal existence we catch only eeting glimpses of
these ideal forms. Now each point – each thing – in these ‘countries’ I
have asked you to imagine could be regarded as a Platonic form.
Triangles certainly are. I shall call the corresponding ‘country’
Platonia. The name re ects its mathematical perfection and timeless
landscape. Nothing changes in Platonia. Its points are all the instants
of time, all the Nows; they are simply there, given once and for all.

Platonia is vast. Size alone is insu cient to convey its vastness.

Triangle Land already has three dimensions, and stretches out to
in nity from its apex and frontiers. That re ects the already huge
number of ways in which three objects can be arranged in space. As
the number of objects is increased, the number of ways in which they
can be arranged increases incredibly fast. The numbers one
encounters in astronomy are as nothing compared with the number
of possible arrangements of large numbers of objects. The instants of
time are numberless. And each is di erent.

There is a saying about time, apparently rst expressed in a piece

of gra ti and much loved by John Wheeler, that seems apt here:
‘Time is nature’s way of preventing everything from happening all at
once.’ In a timeless world, verbs of becoming like ‘happen’ have no
place. But if Nows are both concrete and distinct, it is a logical
contradiction to suppose that they could ‘happen at once’, i.e. be
superimposed on one another. I believe that the aphorism expresses a
profound truth.

Developing the ‘Platonic’ theme, I conjecture that the actual

universe in which we nd ourselves corresponds to some Platonia.
We have not yet fully grasped the structure of its points, its Nows.
Perhaps we never shall, but I assume that in any instant what we
experience, including the appearance of motion, is a transmuted
representation of a part of one such Now. This is not far removed
from Plato’s original idea that we mortals are like beings con ned
from birth to a cave, and that all we ever comprehend of the outside
world and the real beings in it are the shadows they cast on the wall
of our cave as they pass its entrance. I also think that Plato was right
when he said that Being (one of his forms, one of my instants of
time) is real, but that Becoming is an illusion. However, I go further
than Plato in attributing the illusion of Becoming to something that
is real – a special time-capsule structure of Nows. The illusion of
Becoming has its basis in real structure in special Being.

Platonia is the arena that I think must replace space and time.
Why this should be so, how it can be done, and what physics in
Platonia is like is the meat of the book. But it is already possible to
see how di erently creation and a supposed beginning of time
appear in Platonia. Most people are ba ed that time could begin.
How many times do we hear the question, ‘But what happened
before the Big Bang?’ The question reveals the depth to which the
notion of an eternally owing time is ingrained in the psyche. This is
why I call the instants of time ‘things’, so as to break the spell, and
why I have chosen the name Platonia for our home. It is also why I
use paths as the image of history. In itself, there are no paths in
Platonia, just as there were no paths on Earth before animals made
them. The points of Platonia – the Nows – are worlds unto
themselves. No thread of time joins them up. We must think of
Newtonian-type dynamics as something that ‘paints a path’ onto the
timeless landscape of Platonia.

Once the instinctive notion of time is expunged, it is easy to see

that history, as a path in Platonia, can certainly start or end. The
path to Land’s End does terminate there: only the sea lies beyond.
Triangle Land has a point like Land’s End: it is the apex of the
pyramid, which in Figure 4 I called Alpha. Beyond it is nothing, not
even sea. Looking for time before the Big Bang is like looking for
Cornwall in the Irish Sea. If we think that time exists and increases
or decreases along a path in Triangle Land that terminates at that
apex, then we can see that time will certainly begin or end at that
point. I think this is how we should think about the Big Bang. It is
not in the past, it is at a kind of Land’s End.

All Platonias seem by necessity to possess a distinguished point

like the apex of Triangle Land. This is why I call it Alpha. It is
suggestive that Platonia has an Alpha but no Omega: there is no limit
to the size or complexity of things that can exist. Triangle Land
opens out from Alpha to in nity, as do all Platonias. To underline
this fact, Figure 5 is my own attempt to give a somewhat more
artistic and simultaneously realistic representation of the actual
Platonia of our universe, which of necessity is vastly more richly
structured than Triangle Land.

Now we must begin to consider how the notion of Platonia will

change the way we think about such seemingly simple things as
motion. How can it emerge from a scheme without a vestige of time?
Is motion really a pure illusion? If we were in London yesterday and
New York today, we must have moved. Motion must exist. Let me
persuade you that it does not.

IS MOTION REAL?

We had a cat called Lucy, who was a phenomenal hunter. She could
catch swifts in ight, leaping two metres into the air. She was seen in
the act twice, and must have caught other victims since several times
we found just the outermost wing feathers of swifts by the back door.
Faced with facts like this, isn’t it ridiculous to claim there is no
motion?

The argument seems decisive because we instinctively feel that

Lucy has (or, rather had, since sadly she was killed by a car) some
unchanging identity. But is the cat that leaps the cat that lands?
Except for the changes in her body shape, we do not notice any
di erence. However, if we could look closely we might begin to have
doubts. The number of atoms in even the tiniest thing we can see is
huge, and they are in a constant state of ux. Because large numbers
play a vital role in my arguments, I shall give two illustrations. Have
you ever tried to form a picture of the number of atoms in a pea?

Figure 5 Triangle Land is like an inverted pyramid, with frontiers

formed by special triangles as explained in Figure 4. Platonias
corresponding to con gurations of more than three particles have
not only frontiers but also analogous internal topographic features.
This illustration, based on the parachute of a salsify seed (shown life-
size on left) from my wife’s garden, is an attempt to give some idea
of the rich structure of the frontiers of Platonia. No attempt is made
to represent the even richer internal structure. Platonia’s Alpha is
where the ribs converge. Because Platonia has no Omega, the salsify
ribs should extend out from Alpha for ever. (The wind carries the
 actual seeds rather e ciently into our neighbours’ gardens, where
the progeny ourish, but they are not always welcome, although
salsify is an excellent vegetable.)

Imagine a row of dots a millimetre apart and a metre long. That

will be one thousand dots (103). (Actually, it will be 1001, but let us
forget the last 1.) One thousand such rows next to one another, also
a millimetre apart, gives a square metre of dots, one million (106) in
total. The number of dots in one or two squares like that is about the
number of pounds or dollars ordinary mortals like me can hope to
earn in a lifetime. Now stack one thousand such squares into a cube
a metre high. That is already a billion (109). So it is surprisingly easy
to visualize a billion. Five such cubes are about the world’s human
population. Yet we are nowhere remotely near the number of atoms
in a pea.

We shall keep trying. We make another cube of these cubes. One

thousand of them stretched out a kilometre long takes us up to a
trillion (1012). A square kilometre of them will be 1015 (about the
number of cells in the human body), and if we pile them a kilometre
high we get to 1018. We still have a long way to go. Make another
row of one thousand of these kilometre cubes, and we get to 1021.
Finally, make that into a square, one thousand kilometres by one
thousand kilometres and a kilometre high – it would comfortably
cover the entire British Isles to that height. At last we are there: the
number of dots we now have (1024) is around the number of atoms
in a pea. To get the number in a child’s body, we should have to go
up to a cube a thousand kilometres high. It hardly bears thinking
about.

Equally remarkable is the order and organized activity in our

bodies. Consider this extract from Richard Dawkins’s The Sel sh
Gene:

The haemoglobin of our blood is a typical protein molecule.

It is built up from chains of smaller molecules, amino acids,
each containing a few dozen atoms arranged in a precise
pattern. In the haemoglobin molecule there are 574 amino
acid molecules. These are arranged in four chains, which
twist around each other to form a globular three-dimensional
structure of bewildering complexity. A model of a
haemoglobin molecule looks rather like a dense thornbush.
But unlike a real thornbush it is not a haphazard
approximate pattern but a de nite invariant structure,
identically repeated, with not a twig nor a twist out of place,
over six thousand million million million times in an average
human body. The precise thornbush shape of a protein
molecule such as haemoglobin is stable in the sense that two
chains consisting of the same sequences of amino acids will
tend, like two springs, to come to rest in exactly the same
three-dimensional coiled pattern. Haemoglobin thornbushes
are springing into their ‘preferred’ shape in your body at a
rate of about four hundred million million per second and
others are being destroyed at the same rate.
 If, as I think they must be, things are properly considered in
Platonia, Lucy never did leap to catch the swifts. The fact is, there
never was one cat Lucy – there were (or rather are, since Lucy is in
Platonia for eternity, as we all are) billions upon billions upon
billions of Lucys. This is already true for the Lucys in one leap and
descent. Microscopically, her 1026 atoms were rearranged to such an
extent that only the stability of her gross features enables us to call
her one cat. What is more, compared with her haemoglobin
molecules the features by which we identi ed her – the sharp eyes,
the sleek coat, the wicked claws – were gross. Because we do not and
cannot look closely at these Lucys, we think they are one. And all
these Lucys are themselves embedded in the vast individual Nows of
the universe. Uncountable Nows in Platonia contain something we
should call Lucy, all in perfect Platonic stillness. It is because we
abstract and ‘detach’ one Lucy from her Nows that we think a cat
leapt. Cats don’t leap in Platonia. They just are.

You might argue that even if cats do not have a permanent

identity, their atoms do. But this presupposes that atoms are like
billiard balls with distinguishing marks and permanent identities.
They aren’t. Two atoms of the same kind are indistinguishable. One
cannot ‘put labels on them’ and recognize them individually later.
Moreover, at the deeper, subatomic level the atoms themselves are in
a perpetual state of ux. We think things persist in time because
structures persist, and we mistake the structure for substance. But
looking for enduring substance is like looking for time. It slips
through your ngers. One cannot step into the same river twice.
 Zeno of Elea, who belonged to the same philosophical school as
Parmenides, formulated a famous paradox designed to show that
motion is impossible. After an arrow shot at a target has got halfway
there, it still has half the distance to go. When it has gone half that
distance, it still has half of that way to go. This goes on for ever. The
arrow can never reach the target, so motion is impossible. In normal
physics, with a notion of time, Zeno’s paradox is readily resolved.
However, in my timeless view the paradox is resurrected, but the
arrow never reaches the target for a more basic reason: the arrow in
the bow is not the arrow in the target.

There are two parts to my claim that time does not exist. I start
from the philosophical conviction that the only true things are
complete possible con gurations of the universe, unchanging Nows.
Unchanging things do not travel in time from Now to Now. Material
things, we included, are simply parts of Nows. This philosophical
standpoint must be matched by a physical theory that seems natural
within it. The evidence that such a physical theory exists and seems
to describe the universe forms the other part of my claim. This
section has merely made the philosophy, the notion of being, clear.
The physics, the guts of the story, is still to come.

THE BIG PICTURE

Before Newton was born, René Descartes raised a nightmarish

prospect. How do I know, he asked, whether anything exists? Is some
malignant demon conjuring up my thoughts and experiences?
Perhaps there isn’t any world. How can we be sure of anything?
Descartes famously argued that we can at least be certain of our own
existence. Cogito ergo sum: I think, therefore I am. In fact, this did not
get him very far, and his main argument for a real world was that
God would not deceive us on such a fundamental matter.

Modern science has a better answer to the solipsists – those who,

like Descartes in that extreme moment of doubt, deny existence
outside their own thoughts. The starting point is that we do observe
a great variety of phenomena. We can then ask whether we can
postulate a world and laws that lead to the phenomena. If this is so,
it does not explain how or why the world is there, but it does provide
grounds for taking its existence more seriously.

You may think that time capsules and a brain preserved in aspic
aware of seeing motion are getting dangerously close to solipsism
and the machinations of a demon. Without anticipating the rest of
the book, an outline may still be helpful. There are only two rules of
the game: there must be an external world subject to laws and a
correspondence between it and experiences.

Apart from the fact that Newton placed the material objects of
the universe in an arena, my things are his things. They are Nows,
the relative con gurations of the universe. Newton’s Nows form a
string, brought into being by an act of creation at one end, called the
past. It is usually assumed that our experiences in some instant
re ect the structure in a short segment of the string at a point along
it. It is a segment, rather than one Now, because we see things not
only in positions but in motion. However, a single Now contains only
positional information. It seems that we need at least two Nows to
have information about changes of position.

Newtonian history, as modi ed by Big Bang cosmology, translates

into a path in Platonia. It begins at a certain point with a creation
event, after which the laws of nature determine the path. Many paths
satisfy the same laws, but the laws by themselves do not tell us why
one path is chosen by the creation event in preference to others.

The alternative picture, suggested by quantum mechanics and

proposed in this book, is quite di erent. There are no paths with
unique starting points conceived as creation events. Indeed, there are
no paths at all. Instead, the di erent points of Platonia, each of
which represents a di erent possible con guration of the universe,
are present – as potentialities at least – in di erent quantities. This
matches what we found in the Timeless Theory bag: many di erent
triangles present in di erent quantities. It will be helpful to represent
this in a more graphic way. Imagine that Platonia is covered by a
mist. Its intensity does not vary in time – it is static – but it does vary
from position to position. Its intensity at each given point is a
measure of how many con gurations (as in the previous example,
with triangles in the Timeless Bag) corresponding to that point are
present. All these con gurations, present in di erent quantities, you
should imagine for the moment as being collected together in a
‘heap’ or ‘bag’.
 So, Platonia is covered with mist. Its intensity cannot change in
time (there is no time), but it does vary from point to point. In some
places it is much more intense than in others. A timeless law,
complete in itself, determines where the mist collects. The law is a
kind of competition for the mist between the Nows. Those that
‘resonate’ well with each other get more mist. The outcome is a
distribution of mist intensity. This, as I have just explained, is simply
another description of the Timeless Theory bag – for mist intensity
read numbers of triangle copies. But the Nows of this Platonia are
much more complex than triangles.

This opens up possibilities. Triangles tell no stories, they are too

simple. But if the Nows are de ned by, say, the arrangements of
three large bodies and of many thousands of small bodies, things are
di erent. For example, the three large bodies could form the tenth
triangle from the right in Figure 1. The remaining small bodies could
be arranged in such a way that they literally create the pattern of the
rst nine triangles from the right of the sequence. This may seem
contrived, but it is possible. It is a Now in a greatly enlarged
Platonia. Shown such a Now, what could we make of it? One
interpretation is that the small bodies record what the large bodies
have done: the Now is a time capsule, a picture of a Newtonian
history. As soon as a su cient number of bodies are present, the
possibilities for creating time capsules are immense.

I believe the sole reason we believe in time is because we only

ever experience the universe through the medium of a time capsule.
My assumptions are:

(1) All experience we have in some instant derives from the structure
in one Now.

(2) For Nows capable of self-awareness (by containing brains, etc.)

the probability of being experienced is proportional to their mist
intensity.

(3) The Nows at which the mist has a high intensity are time
capsules (they will also possess other speci c properties).

Thus, the one law of the universe that determines the mist intensity
over Platonia is timeless. The Nows and the distribution of the mist
are both static. The appearance of time arises solely because the mist
is concentrated on time capsules, and a Now that is a time capsule is
therefore much more likely to be experienced than a Now that is not.
(Please remember that this is only an outline: the detailed arguments
are still to come.)

Of the three assumptions, the second is the most problematic. The

rst and third may seem strange and implausible, but they can be
made de nite. If correct, their signi cance and meaning are clear-
cut. Both could be shown to be false, but this is good, since a theory
that cannot be disproved is a bad theory. The best theories make rm
predictions that can be tested. The main di culty with the second
assumption is in saying what it means. We encounter, in a modi ed
form, the di culty that Descartes raised. It is acute.
 In a Newtonian scheme, the connection between theory and
experience is unambiguous. There is a path through Platonia, and all
the Nows on it are realized: sentient beings within any Nows on the
path do experience those Nows. In the alternative scheme, the
distribution of the mist over Platonia – its intensity at each Now – is
as de nite as the line of the Newtonian path. The di culty, which is
deeply rooted in quantum mechanics, is how to interpret the
intensity of the mist. When we get to grips with quantum mechanics,
I shall explain my reasons for assuming that the mist intensity at a
Now measures its probability of being experienced. Perhaps some
cosmic lottery is the best way to explain this.

Each Now has a mist intensity. Suppose that all the Nows
participate in a lottery, receiving numbers of tickets proportional to
their mist intensities. Nows where the mist is intense get tickets
galore, others very few. By assumption (1), conscious experience is
always in one Now. If a Now has a special structure, it is capable of
self-awareness. But is it actually self-aware? Structure in itself, no
matter how intricate and ordered, cannot explain how it can be self-
aware. Consciousness is the ultimate mystery.

Perhaps it is a mystery that makes some sense of the mist that

covers Platonia. If there is a cosmic lottery, clearly the Nows with the
most tickets will have the best chance. If a ticket belonging to a Now
capable of self-awareness is drawn, this can, so to speak, ‘bring to
life’ the Now. It is aware. The consciousness potentially present in
Nows structured the right way is actual in those that are drawn. Two
questions about this cosmic lottery may well be asked: when are the
tickets drawn, and how many are drawn?

The rst question is easily answered: it has no meaning. Think of

the brain preserved in aspic, or the unfortunate brain-damaged
patient who believes that Harold Macmillan is Prime Minister and
Dwight Eisenhower is President. The structure capable of making a
Now self-aware is eternal and timeless. Structure is all that counts.
Self-awareness does not happen at a certain time and last for some
fraction of a second. Yesterday seems to come before today because
today contains records (memories) of yesterday. Nothing in the
known facts is changed by imagining them hung on a ‘line of time’ –
or even reversing their positions on that line. The instant is not in
time, time is in the instant. We do not have to worry when the draw
is made, only whether our number comes up.

The question of how many tickets are drawn is a tough one. If

only one is drawn, your present Now, which does exist, must be the
one and only instant realized and experienced. All your memories
are then illusions in the sense that you never experienced them. That
seems very hard to believe. What is more, memories are legion. If
you believe you did actually experience them all, then lots of Nows
have been drawn. From this it is a small step to saying that all Nows
in Platonia are drawn. In quantum mechanics, this is called the
many-worlds hypothesis. But then the theory seems to become
vacuous: everything that can be is, no predictions appear to be made.
The root of the problem is the assumption, neat and clean in itself,
that each experienced instant is always tied to a single Now and that
the distribution of the mist over Platonia is determined by a law
indi erent to the workings of the cosmic lottery. Whether or not
particular Nows are drawn has no e ect on the mist intensity. The
rules of the scheme make it quite impossible to say how many, if
any, of your memories are real. All we know is that the present Now
is real. You can see how Descartes’s dilemma is revived in such a
scheme. I suspect that it is a problem we just have to live with.

The theory is still testable because only Nows with high mist
intensity (and therefore high probability) are likely to be
experienced, and such Nows have characteristic properties: above all,
they are time capsules. We can therefore test our own experiences
and see if they verify the predictions of the theory. This is something
that in principle can be settled by mathematics and observations. For
if physicists can determine or guess the structure of Platonia and
formulate the law that determines how the mist is distributed over it,
then it is simply a matter of calculation to nd out where in Platonia
the mist is most intense. If the mist is indeed concentrated on
structures that are time capsules, the theory will make a very strong
prediction – any Now that is experienced will contain structures that
seem to be records of a past of that Now. It will also contain other
characteristic structures.

The huge number of things that can coexist simultaneously in one

Now is signi cant here. It means that many independent tests can be
made on a single time capsule to see whether the predictions are
con rmed. The laws of nature are usually tested by repeating
experiments in time. If the same initial state gives the same outcome,
the law is con rmed. However, for an object as richly structured as
the Earth (which in any instant belongs to one of the Nows in
Platonia), repeating experiments in time can be replaced by
repeating them in space. As it happens, even con rming a theory by
repeating experiments in time as normally understood boils down to
comparing records in one Now. The precondition of all science is the
existence of time capsules. All the Nows we experience are time
capsules. The question is whether we can explain why this is so from
rst principles: can the strong impression of time emerge from
timelessness? It is a logical possibility, but the real test must await
mathematical advances. Unfortunately, they are not likely to be easy.

Strange as a timeless theory may seem, it has the potential to be

very powerful. Boltzmann’s work highlighted two di culties
inherent in any theory of time – initial conditions must be imposed
arbitrarily; and dull, unstructured situations are far more probable
than the interesting structured things we nd all around us.
Interestingly structured Nows are an extreme rarity among all the
Nows that can be. If the mist does pick out time capsules in Platonia,
it must be very selective. Since all possible structures are present in
Platonia, the vast majority of Nows do not contain any structures at
all that could be called records. Even then, the apparent records will
be mutually consistent in only a tiny fraction of what is already a
tiny fraction. Only our habitual exposure to the time capsules we
experience blinds us to the magnitude of the phenomenon that needs
to be explained. Stars in real space give us only an inkling of how
thinly time capsules are spread. Any scheme that does select them
will be very powerful. But more than that, it will be more fully
rational than classical physics, with its need to invoke a very special
initial condition, can ever be. Once the law that governs the
distribution of the mist over Platonia has been speci ed, nothing
more remains to be done. The mist gathers where it does for only
two reasons: the structure of the law and the structure of Platonia.

So where is the mist likely to gather? The mathematics needed to

answer this question will certainly be di cult, but there are some
hints (which I shall elaborate in the nal chapters). They suggest that
mist is likely to be distributed along thin, gossamer-like laments
that bifurcate and form a tree-like structure (Figure 6).

A tendency to bifurcation is deeply rooted in quantum mechanics.

In principle, it could happen in both directions along a lament.
However, the Nows we experience all seem to have arisen from a
unique past. There seems to be no branching in that direction.
Within quantum mechanics, as presently formulated in space and
time, this fact is not impossible, but it is as puzzling as the low
entropy that so exercised Boltzmann. It does seem improbable. I
suspect that everything will look di erent if we learn to think about
quantum mechanics in Platonia. For one thing, the arena has a very
di erent shape. This is why I was keen to show you at this early
stage the diagrams of Triangle Land (Figures 3 and 4) and my
representation of Platonia (Figure 5). It opens out in one direction
from nothing. I suspect that the branching laments of mist in Figure
6 arise because they re ect this overall, ower-like structure of
Platonia. If that is so, the great asymmetries of our existence – past
and future, birth and death – arise from a deep asymmetry in being
itself. The land of possible things has one absolute end, where it
abuts onto mere nothing, but it is unbounded the other way, for
there is no limit to the richness of being.

Who knows what experiences are possible in the oases of richly

structured Nows strung out along the trade routes that cross the
deserts of Platonia? The plurality of experience is remarkable and
suggestive. In any instant, we are aware of many things at once.
Through memories we are, as it were, present simultaneously in
many di erent Nows in Platonia. Richness of structure permits this.
One grand structure contains substructures that are ‘pictures’ –
simpli ed representations that capture the essential features – of
other structures. Our memories are pictures of other Nows within
this Now, rather like snapshots in an album. Each Now is separate
and a world unto itself, but the richly structured Nows ‘know’ about
one another because they literally contain one another in certain
essential respects. As consciousness surveys many things at once in
one Now, it is simultaneously present, at least in part, in other Nows.
This awareness of many things in one could well exist in a much
more pronounced form in other places in Platonia.
 Figure 6 The conjectured lamentary distribution of mist in
Platonia. The instant you experience now is marked NOW. To its left
lie Nows of which you have memories in NOW. There is no
bifurcation in this direction, matching our conviction that we have a
unique past. In the other direction there is a branching into di erent
alternative ‘futures’ of NOW. In all of them, you think you have
advanced into the future by the same amount from NOW. These
di erent laments are ‘parallel worlds’ that seem to have a common
past, to which NOW belongs. Note that the laments have a nite
width, unlike a Newtonian string of successive instants. All around
NOW, along the lament and to either side of it, are other Nows with
slightly di erent versions of yourself. All such Nows are ‘other
worlds’ in which there exist somewhat di erent but still recognizable
 versions of yourself. In other laments are worlds you would not
recognize at all.

The picture of ourselves dividing into parallel Nows may be

unsettling, but the phenomenon itself is familiar. We are used to
being in di erent Nows and being slightly di erent in all of them –
that is simply the e ect of time as it is usually conceived. The
account of Lucy’s leaps emphasized that the di erences in ourselves
between Nows are far greater than we realize within consciousness.
Huge numbers of microscopically di erent Nows could give identical
conscious experience. As we shall see, quantum mechanics forces us
to consider Nows everywhere, not just those on one path. It unsettles
by division, seeming to threaten dissolution and personal integrity.
But it simultaneously binds us into the far mightier whole of
everything that can be, doing so much more decisively than any
Newtonian scheme can do. For the Nows that are likely to be
experienced are the ones that are most sensitive to the whole of
Platonia.

I think this is su cient introduction. I could go on to talk about

free will, the future, our place in the universe, religion, and so on. If
the theory is correct, it must change the way we think about these
things. However, without some real understanding of the arguments
for a timeless universe, I feel further discussion would lack a solid
basis. I therefore postpone these issues to later in the book, especially
the epilogue. My aim so far has been to outline the scheme and to
show that it is truly timeless and at least logically possible.
PART 2
The Invisible Framework and the Ultimate
Arena

Newton introduced two ‘great invisibles’ as the arena of physics:

absolute space and time. In Part 2 we shall see why they have
appeared for so long to be better suited to acting as the frame of the
world than Platonia. It is all to do with an issue that physicists and
philosophers have been arguing about for centuries: is motion
absolute or relative? Newton’s position has seemed to be so strong
that many people still believe it cannot be overthrown. But it can.
The demonstration of the relatively simple solution in Newtonian
physics will prepare us for the almost miraculous way in which
things work out in Einstein’s theory (Part 3). They give the strongest
suggestion that quantum cosmology – and hence our universe – is
timeless. That we come to in Parts 4 and 5. Chapter 4 is a brief
historical introduction, and sets the scene for the remainder of Part
2 – and for much of the rest of the book.
CHAPTER 4
Alternative Frameworks

ABSOLUTE OR RELATIVE MOTION?

Both Copernicus and Kepler believed that the universe, with the
solar system at its centre, was bounded by a huge and distant rigid
shell on which the luminous stars were xed. They did not speculate
what lay beyond – perhaps it was simply nothing. They de ned all
motions relative to the shell, which thus constituted an
unambiguous framework. Many factors, above all Galileo’s
telescopic observations in 1609 and the revival of interest in the
Greeks’ idea of atoms that move in the void, destroyed the old
cosmology. New ideas crystallized in a book that Descartes wrote in
1632. He was the rst person to put forward clearly an idea which,
half a century later, Newton would make into the most basic law of
nature: if nothing exerts a force on them, all bodies travel through
space for ever in a straight line at a uniform speed. This is the law of
inertia. Descartes never published his book because in 1633 the
Inquisition condemned Galileo for teaching that the Earth moves.
The Copernican system was central to Descartes’s ideas, and to
avoid Galileo’s fate he suppressed his book.
 He did publish his ideas in 1644, in his in uential Principles of
Philosophy, but with a very curious theory of relative motion as an
insurance policy. He argued that a body can have motion only
relative to some other body, chosen as a reference. Since any other
body could play the role of reference, any one body could be
regarded as having many di erent motions. However, he did allow a
body to have one true ‘philosophical motion’, which was its motion
relative to the matter immediately adjacent to it. (Descartes believed
there was matter everywhere, so any body did always have matter
adjacent to it.) This idea let him o the Inquisition’s hook, since he
claimed that the Earth was carried around the Sun in a huge vortex,
as in a whirlpool. Since the Earth did not move relative to the
immediately adjacent matter of the vortex, he argued that it did not
move!

However, he then formulated the law of inertia, just as in 1632.

When, sometime around 1670, long after Descartes’s death in 1650,
Newton came to study his work, he immediately saw the aw. To
say that a body moves in a straight line presupposes a xed frame of
reference, which Descartes had denied. Since Newton could see the
great potential of the law of inertia, to exploit it he came up with
the concept of an immovable space in which all motion takes place.
He was very scornful of Descartes’s inconsistency, and when he
published his own laws in 1687 he decided to make it a big issue,
without, however, mentioning Descartes by name. He introduced
the notion of absolute space and, with it, absolute time.
 Newton granted that space and time are invisible and that one
could directly observe only relative motions, not the absolute
motions in invisible space. He claimed that the absolute motions
could nevertheless be deduced from the relative motions. He never
gave a full demonstration of this, only an argument designed to
show that motion could not be relative. He was making a very
serious point, but at the same time he wanted to make a fool of
Descartes. This had strange and remarkable consequences.

Descartes had sought to show that all the phenomena of nature

could be explained mechanically by the motion of innumerable,
tiny, invisible particles. Vital to his scheme was the centrifugal force
felt as tension in a string that retains a swung object. The object
seems to be trying to escape, to ee from the centre of rotation. In
Newtonian terms, it is actually trying to shoot o along the tangent
to the circle, but that is still a motion that would take it away from
the centre and create the tension. Descartes claimed that light was
pressure transmitted from the Sun to the Earth by centrifugal
tension set up in the vortex that he pictured swirling around the
Sun. Because centrifugal force was so important to Descartes,
Newton used it to show that motion cannot be relative. Newton’s
intention was to hoist Descartes by his own petard.

Newton imagined a bucket lled with water and suspended by a

rope from the ceiling. The bucket is turned round many times,
twisting the rope, and is then held still until the water settles. When
the bucket is released, the rope unwinds, twisting the bucket.
Initially the surface of the water remains at, but slowly the motion
of the bucket is transmitted to the water, which starts to spin, feels a
centrifugal force and starts to rise up the side of the bucket. After a
while, the water and bucket spin together without relative motion,
and the water surface reaches its greatest curvature.

Newton asked what it was that caused the water’s surface to

curve. Was it the water’s motion relative to the side of the bucket
(Descartes’s claimed true philosophical motion relative to the
immediately adjacent matter) or motion relative to absolute space?
Surely the latter, since when the relative motion is greatest, at the
start, there is no curvature of the water’s surface, but when the
relative motion has stopped (and the water and bucket spin
together) the curvature is greatest. This was Newton’s main
argument for absolute space. It was strong and it ridiculed
Descartes.

In Newton’s lifetime, his notion of absolute space, to which he

gave such prominence, attracted strong criticism. If space were
invisible, how could you say an object moves in a straight line
through a space you cannot see? Newton never satisfactorily
answered this question. Many people felt, as Descartes did, that
motion must be relative to other matter, though not necessarily
adjacent matter. Bishop Berkeley argued that, as in Copernican
astronomy, motion must ultimately be relative to the distant stars,
but he failed to get to grips with the problem that the stars too must
be assumed to move in many di erent ways and thus could not
de ne a single xed framework, as Copernicus and Kepler had
believed.

Newton’s most famous critic was the great German

mathematician and philosopher Wilhelm Gottfried Leibniz, who had
been involved in a very unpleasant dispute with Newton about
which of them had rst discovered the calculus, the revolutionary
new form of mathematics that made so many things in science much
easier, including the development of mechanics. In 1715, Leibniz
began a famous correspondence on Newton’s ideas with Samuel
Clarke, who was advised by Newton. The Leibniz-Clarke
Correspondence has become a classic philosophy text. Many
undergraduates study it, and philosophers of science often discuss it.

The exchange had an inconclusive outcome. It is generally

agreed that Leibniz advanced e ective philosophical arguments, but
he never addressed the detailed issues in mechanics. Typically, he
argued like this. Suppose that absolute space does exist and is like
Newton claimed, with every point of space identical to every other.
Now consider the dilemma God would have faced when he created
the world. Since all places in absolute space are identical, God
would face an impossible choice. Where would he put the matter?
God, being supremely good and rational, must always have a
genuine reason for doing something – Leibniz called this the
‘principle of su cient reason’ (I have already appealed to this when
I discussed brain function and consciousness, by requiring an
observable e ect to have an observable cause) – and because
absolute space o ered no distinguished locations, God would never
be able to decide where to put the matter. Absolute time, on the
assumption that it existed, presented the same di culty. Newton
had said that all its instants were identical. But then what reason
could God have for deciding to create the world at some instant
rather than another? Again, he would lack a su cient reason. For
reasons like these, not all of them so theological, Leibniz argued that
absolute space and time could not exist.

A century and a half passed before the issue became a hot topic
again. This raises an important issue: how could mechanics have
dubious foundations and yet ourish? That it ourished
nevertheless was due to fortunate circumstances that are very
relevant to the theme of this book. First, although the stars do move,
they are so far away that they provide an e ectively rigid
framework for de ning motions as observed from the Earth. It was
found that in this framework Newton’s laws do hold. It is hard to
overestimate the importance of this fortunate e ective xity of the
distant stars. It presented Newton with a wonderful backdrop and
convenient framework. Had the astronomers been able to observe
only the Sun, Moon and planets but not the stars (had they been
obscured by interstellar dust), Newton could never have established
his laws. Thus, scientists were able to accept Newton’s absolute
space as the true foundation of mechanics, using the stars as a
substitute for the real thing – that is, a true absolute frame of
reference. They also found that Newton’s uniformly owing time
must march in step with the Earth’s rotation, since when that was
used to measure time (in astronomical observations spanning
centuries, and even millennia) Newton’s laws were found to hold.
Once again, a substitute for the ‘real thing’ was at hand. One did not
have to worry about the foundations. Fortunate circumstances like
these are undoubtedly the reason why it is only recently that
physicists have been forced to address the issue of the true nature of
time.

The person who above all brought the issue of foundations back
to the fore was the Austrian physicist Ernst Mach, whose brilliant
studies in the nineteenth century of supersonic projectiles and their
sonic boom are the reason why the Mach numbers are named after
him. Mach was interested in many subjects, especially the nature
and methods of science. His philosophical standpoint had points in
common with Bishop Berkeley, but even more with the ideas of the
great eighteenth-century Scottish empiricist David Hume. Mach
insisted that science must deal with genuinely observable things,
and this made him deeply suspicious of the concepts of invisible
absolute space and time. In 1883 he published a famous history of
mechanics containing a trenchant and celebrated critique of these
concepts. One suggestion he made was particularly in uential.

It arose as a curious consequence of the covert way Newton had

attacked Descartes. Considering Newton’s bucket argument, Mach
concluded that, if motion is relative, it was ridiculous to suppose
that the thin wall of the bucket was of any relevance. Mach had no
idea that Newton was attacking Descartes’s notion of the one true
philosophical motion, just as Newton had not seen that Descartes
had invented it only to avoid the wrath of the Inquisition. Newton
had used the bucket argument to show that relative motion could
not generate centrifugal force, but Mach argued that the relative
motions that count are the ones relative to the bulk of the matter in
the universe, not the puny bucket. And where is the bulk of the
matter in the universe? In the stars.

This led Mach to the revolutionary suggestion that it is not space

but all the matter in the universe, exerting a genuine physical e ect,
that creates centrifugal force. Since this is just a manifestation of
inertial motion, which Newton claimed took place in absolute space,
Mach’s proposal boiled down to the idea that the law of inertia is
indeed, as Bishop Berkeley believed, a motion relative to the stars,
not space. Mach’s important novelty was that there must be proper
physical laws that govern the way distant matter controls the
motions around us. Each body in the universe must be exerting an
e ect that depends on its mass and distance. The law of inertia will
turn out to be a motion relative to some average of all the masses in
the universe. For this basic idea, Einstein coined the expression
Mach’s principle, by which it is now universally known (though
attempts at precise de nition vary quite widely).

Mach’s idea suggests that the Newtonian way of thinking about

the workings of the universe, which is still deep-rooted, is
fundamentally wrong. The Newtonian scheme describes an
‘atomized’ universe. The most fundamental thing is the containing
framework of space and time: that exists before anything else.
Matter exists as atoms, tiny unchanging masses that move in space
and time, which govern their motion. Except when close enough to
interact, the atoms move with complete indi erence to one another,
each following a straight and lonely path through the in nite
reaches of absolute space. The Machian idea takes the power from
space and time and gives it to the actual contents of the universe,
which all dance in their motions relative to one another. It is an
organic, holistic view that knits the universe together. Very
characteristic is this remark of Mach in his The Science of Mechanics
(pp.287-8):

Nature does not begin with elements, as we are obliged to

begin with them. It is certainly fortunate for us that we can,
from time to time, turn aside our eyes from the
overpowering unity of the All and allow them to rest on
individual details. But we should not omit, ultimately to
complete and correct our views by a thorough consideration
of the things which for the time being we have left out of
consideration.

Mach himself made only tentative suggestions for a new relative

mechanics, but his remarks caught the imagination of many people,
above all Einstein, who said that Hume and Mach were the
philosophers who had in uenced him most deeply. Einstein spent
many years trying to create a theory that would embody Mach’s
principle, and initially believed that he had succeeded in his general
theory of relativity. That is why he gave it that name. However,
after a few years he came to have doubts. Eventually he concluded
that Mach’s idea had been made obsolete by developments in
physics, especially the theory of electro-magnetism developed by
Faraday and Maxwell, which had introduced new concepts not
present in Newton’s scheme.

Throughout the twentieth century, physicists and philosophers

discussed Mach’s principle at great length, without coming to any
conclusion. It is my belief that the problem lies in Einstein’s highly
original but indirect approach. Mach had not made a really clear
proposal, and Einstein never really stopped and asked himself just
what should be achieved by Mach’s principle. I shall consider this in
Part 3, but I need to anticipate a small part of the story in order to
justify Part 2. Einstein’s theory is rather complicated and achieves
several things at once. It is not easy to separate the parts and see the
‘Machian’ structure. In my opinion, general relativity is actually as
Machian as it could be. What is more, it is the Machian structure
that has such dramatic consequences when one tries to reconcile the
theory with quantum mechanics. If, as I believe, the quantum
universe is timeless, it is so because of the Machian structure of
general relativity. To explain the core issues, I need a simpli ed
model that captures the essentials. This Part 2 will provide. It will
also provide a direct link between the great early debate about the
foundations of mechanics and the present crisis of quantum
cosmology. Two key issues are still the same: what is motion, and
what is time? It will also enable me to explain the main work in
physics with which I have been involved, and make it easier for you
to see why I have come to doubt the existence of time.

Science advances in curious ways, and scientists are often

curiously unconcerned with foundations. Descartes was one of the
greatest philosophers, yet in that rst book in 1632 he never gave a
moment’s thought to the de nition of motion. We are so used to
living on the solid Earth that it seems unproblematic to say that a
body moves in a straight line. If the Inquisition had not condemned
Galileo, Descartes would never have argued for the relativity of
motion. But for the inconsistency of his system, Newton would not
have made an issue out of absolute space and time. He would not
have devised the bucket argument, Mach might never have had his
novel idea, and Einstein would not have been inspired to his
greatest creation.

Had the Inquisition condemned Galileo a few months later,

Descartes would have published his ideas in their original form –
and general relativity might never have been found.

AN ALTERNATIVE ARENA

I would like to say a bit more about my own personal development,

which as the book progresses will help you to understand why I am
so deeply convinced of the need to have a new concept of time. In
the very rst days after my trip to the Bavarian Alps, while thinking
hard about time, I came across Mach’s book. Like so many others, I
was captivated by his idea about inertia. His comments on time also
encouraged me greatly: ‘It is utterly beyond our power’, he said, ‘to
measure the changes of things by time. Quite the contrary, time is
an abstraction, at which we arrive by means of the changes of
things.’ This was just the conclusion I had reached. A year or so
later, after I had decided to study the foundations of physics, I
started to read the papers Einstein had written when he was
creating general relativity. Comparing them with what Mach had
written, I came to the conclusion that Einstein had simply not set
about the problem in the right way: he had not attacked it directly.
It seemed to me necessary to go back to rst principles.

It was six or seven years before I came to form really clear ideas.
I eventually concluded that what was needed above all was a new
arena in which to describe the universe. I arrived at the notion of
Platonia (or, as I originally called it, the relative con guration space
of the universe). The argument was quite simple. First, it is a fact
that we orient ourselves in real life by objects we actually see, not
by invisible space (see the Notes on the previous chapter). Things
are the signposts that tell us where we are. There is also the
fortunate fact that we live on the nearly rigid Earth. We can orient
ourselves by means of just a few objects xed on its surface, say
church spires when hiking in the English countryside. Always there,
the Earth provides a natural background. Motion seems to take
place in a framework. But imagine what life would be like if we
lived on a jelly sh!

The fact is that we live in a very special location. Only the tiniest
fraction of matter in the solar system, let alone the universe, is in
solid form. Imagine that we lived in an environment much more
typical of the universe – in space. To simplify things, let there be
only a nite number of objects, all in motion relative to one
another. At any instant there are certain distances between these
other objects and us. There is nothing else. In these circumstances,
what would be the natural way to answer what is always a
fundamental question: where are we? We have no other means of
saying where we are except in terms of our distances to other
objects. What is more, it would be arti cial to choose just a few of
them to locate ourselves. Why these rather than those? It would be
much more natural to specify our distances to all objects. They
de ne our position. This conclusion is very natural once we become
aware that nothing is xed. Everything moves relative to everything
else.

Taking this further, thinking about the position and motion of

one object is arti cial. We are part of Mach’s All, and any motion
we call our own is just part of a change in the complete universe.
What is the reality of the universe? It is that in any instant the
objects in it have some relative arrangement. If just three objects
exist, they form a triangle. In one instant the universe forms one
triangle, in a di erent instant another. What is to be gained by
supposing that either triangle is placed in invisible space? The
proper way to think about motion is that the universe as a whole
moves from one ‘place’ to another ‘place’, where ‘place’ means a
relative arrangement, or con guration, of the complete universe.

An arena is the totality of places where one can go in some

game. But who is playing the game and where? In Newton’s game,
individual objects play in absolute space. In Mach’s game, there is
only one player – the universe. It does not move in absolute space, it
moves from one con guration to another. The totality of these
places is its relative con guration space: Platonia. As the universe
moves, it therefore traces out a path in Platonia. This captures,
without any redundant structure, the idea of history. History is the
passage of the universe through a unique sequence of states. In its
history, the universe traces a path through Platonia.

However, such language makes it sound as though time exists. I

may have inadvertently conjured up an image in your mind of the
universe as a lone hiker walking the fells in northern Platonia.
Properly understood, the Machian programme is much more radical.
For no Sun rises or sets over that landscape to mark the walker’s
progress. The Sun, like the moving parts of any clock, is part of the
universe. It is part of the walker. Of course, to say that time has
passed, we must have some evidence for that. Something must
move. That is the most primitive fact of all. In the Newtonian
picture, as in Feynman’s quip, time can pass without anything
happening. If we deny that, the grandstand clock must go. There is
nothing outside the universe to time it as it goes from one place to
another in Platonia – only some internal change can do that. But
just as all markers are on an equal footing for de ning position, so
are all changes for the purposes of timing. We must reckon time by
the totality of changes. But changes are just what takes the universe
from one place in Platonia to another. Any and all changes do that.
We must not think of the history of the universe in terms of some
walker on a path who can move along it at di erent speeds. The
history of the universe is the path. Each point on the path is a
con guration of the universe. For a three-body universe, each
con guration is a triangle. The path is just the triangles – nothing
more, nothing less.

With time gone, motion is gone. If you saw a jumbled heap of

triangles, it would not enter your head that anything moved, or that
one triangle changed into another. When Newton’s superstructure is
removed, Newtonian history is like that jumbled heap of triangles,
except that it is a special heap. If you picked up each triangle – I call
that picking up an instant of time – and marked its position in
Triangle Land, you would nd that the marks of the triangles form a
continuous curve.

This was the decisive picture that crystallized in my mind about

1971. At that stage I had no thought of applications to quantum
mechanics, and no inkling that it might lead to the replacement of
one clearly delineated path through Platonia by a mist that hovers
over the same timeless landscape. We had a blackboard in our
kitchen in College Farm, and I wrote at the top it it: The history of
the universe is a continuous curve in its relative con guration
space.’ My wife, perhaps understandably, was rather sceptical about
the progress I was making. After all, fourteen words were not much
to show for seven years of thought. But the clear formulation of the
concept of Platonia was the important thing. It shifts attention from
the parts of the universe to the universe itself. It shows that time is
not needed as an extra element, the Great Timekeeper outside the
universe. The universe keeps track of itself. In one instant it is
where it is, in another it is somewhere else. That is what a di erent
instant of time is: it is just a di erent place in Platonia. Instants of
time and positions of the objects within the universe are all
subsumed into the single notion of place in Platonia. If the place is
di erent, the time is di erent. If the place is the same, time has not
changed. This change of viewpoint is made possible only because
the universe is treated as a single whole and time is reduced to
change.

I think the reason why I take the possibility of a completely

timeless universe more seriously than almost all other physicists is
this background that came from thinking about Mach’s principle. As
we shall see, Platonia is the natural arena for the realization of that
idea. Many years after I had rst recognized that Platonia would
provide the basis for the solution to the Machian problem, I began
to see that it had deep relevance in the quantum domain too. The
problems of the origin of inertia and of quantum cosmology form a
seamless whole.
CHAPTER5
Newton’s Evidence

THE AIMS OF MACHIAN MECHANICS

Merely changing the framework in which one conceives of the

universe does nothing, but it is still very illuminating to look at some
fundamental facts of mechanics in the alternative arenas of absolute
space and Platonia. This exercise brings out the strengths of
Newton’s position, and at the same time shows what a Machian
approach must achieve. The following discussion is based on
penetrating remarks made in 1902 by the great French
mathematician Henri Poincaré. More clearly than Mach, he
demonstrated what is required of a theory of relative motion.
Unfortunately, his remarks were overshadowed by Einstein’s
discovery of relativity and did not attract the attention they deserved
– and still deserve.

You may nd that this chapter requires more re ection than all
the others. You certainly do not need to grasp it all, but I hope that
you will be able to change from a way of thinking to which we have
been conditioned by the fact that we evolved on the stable surface of
the Earth to a more abstract way of thinking that would have been
forced upon us had we evolved from creatures that roamed in space
between objects moving through it in all directions. We have to learn
how to nd our bearings when the solid reassuring framework of the
Earth is not there. This is the kind of mental preparation you need to
understand the ideas Poincaré developed. In this respect, he was
smarter than Einstein.

Poincaré simply asked, rather more precisely than anyone before

him, what information is needed to predict the future. Another
French mathematician, Pierre Laplace, had already imagined a divine
intelligence that at one instant knows the positions and motions of
all bodies in the universe. Using Newton’s laws, the divinity can then
calculate all past and future motions – it can see, in its mind’s eye,
all of history laid out for the minutest inspection. As an alternative to
the standard representation in Newton’s absolute space, it will help
to see this miracle performed in Triangle Land, the simplest Platonia.
This will reveal a curious defect in Newtonian mechanics.
Figure 7 Here Alpha, the apex of Triangle Land, is at the bottom.
The axes that were shown in Figures 3 and 4 have been removed
since they would detract from the essence of this and the following
gure. Two of the ribs of Triangle Land run upwards to the right and
left. They are marked BC = 0 and CA = 0 to indicate that the
triangles corresponding to points on these ribs have ‘collapsed’
because their sides BC and CA, respectively, are zero. The third rib
recedes into the gure (this is the rib that in Figure 4 runs along the
‘ oor’). The shaded planes cut the faces of the pyramid in the lines
on the ‘sheets’ in Figure 4. As shown here, all points on any straight
line from Alpha through Triangle Land represent di erent triangles,
but they di er only in size. In fact, all the points on any one of the
shaded sheets represent triangles that have the same perimeter. If we
are interested only in the shapes of the triangles, these are
represented by the points on just one of the planes (i.e. the di erent
points on any one plane represent di erently shaped triangles; this
possibility is depicted more fully in Figure 8).
 Figure 8 This shows one of the shaded planes in Figure 7. It can be
called Shape Space, because each point in it represents a di erent
possible shape of a triangle. The point at the centre represents an
equilateral triangle (all three sides equal). Points on the three dashed
lines correspond to isosceles triangle (two sides equal). All other
possible triangles are scalene (all three sides unequal). Points on the
three curved lines correspond to right-angled triangles (the central
triangle at the top is the Pythagorean 3, 4, 5 triangle). All points
inside the curved lines correspond to acute triangles (all angles less
than 90°); all points outside the curved lines correspond to obtuse
triangles (one angle greater than 90°). When I asked my friend
Dierck Liebscher to create this diagram, I had no idea that it would
turn out to be so beautiful. I do like the curved lines of the right-
angled triangles! Let me remind you that all the points on the
straight edges of Shape Space correspond to triangles that have
become ‘ at’ because all three corners of each of these triangles are
collinear (on one line). Each of the three vertices of Shape Space
correspond to con gurations in which two particles coincide, while
the third is some distance from them.

You may like to refresh your memory by returning to Figures 3

and 4 before you examine Figures 7 and 8. Figures 3, 4, 7, 8 and 9
are very important. I am rather concerned that younger readers
(those under forty, or even fty!) may have some di culty with
them, since fundamental geometry is not taught nearly as thoroughly
at school as it used to be. However, if you can spend a bit of time on
these gures and begin to understand what they mean, you will
certainly get a great deal more out of this book. In fact, you will also
be absorbing some of the deepest and most fruitful concepts in
mathematics and theoretical physics. Don’t worry – it can be done.
Once the clutter of technical detail is removed, all the great ideas in
mathematics and physics are in essence very simple and intuitive.
But you need patience to absorb them. When Newton was asked how
he had come to make his great discoveries, he answered: ‘By thinking
about these things for a long time.’ Try lying in bed or a nice warm
bath and thinking about Triangle Land!

Figure 8 brings out the rich topography of any Platonia (Shape

Space is a possible Platonia). Wherever you go, you nd something
di erent. Each point is a di erent ‘world’ – and a di erent instant of
time. There are even characteristically di erent regions (of acute and
obtuse triangles), like provinces or counties, as well as internal and
external frontiers (the right-angled and isosceles triangles). Any
Platonia is quite unlike Newton’s absolute space, all points of which
are identical. As I remark in the Notes to Chapter 4, there is
something unreal about that property of absolute space. Real things
have genuine attributes that distinguish them from other real things.
Platonia is a land of real things. I nd Figure 8 very suggestive.
Leibniz always said that it is necessary to consider all possible worlds
and nd some reason why one rather than another occurs or is
actually created. In Figure 8, we do see all the possible worlds of
triangle shapes laid out before us. Box 3 contains a short digression
on possible kinds of Platonias.
BOX 3 Possible Platonias

One of the big unsolved problems of physics is the origin of distance

and whether it is absolute. Since we need a measuring rod to
measure any distance, this suggests rather strongly that distance is
relative (to the chosen rod). If we tried to double every distance in
the universe, the length of the rod would be doubled too, and
nothing actually observable would be changed. For reasons like this,
physicists have a hunch that absolute scale should have no objective
meaning. However, this is not con rmed by existing theories and
experimental facts, which do suggest that distance is in a well-
de ned sense absolute. The hope remains that a physics completely
without scale will be found. If so, Shape Space gives an idea of what
the corresponding Platonia will be like. There is still a uniquely
distinguished point in it – the central point in Figure 8. It takes the
place of Alpha in Triangle Land. If you ‘touched’ the central point,
the equilateral triangle would ‘light up’. This is the most symmetrical
con guration the three-body universe can have. Symmetry is
beautiful in one way but bland in another. The boundaries of Shape
Space are somewhat unappealing too, because they represent
improper triangles that atten into a line (mathematicians call such
con gurations, collinear in this case, degenerate). The vertices of
Shape Space correspond to collinear con gurations in which one
particle is in nitely far from the other two. The interesting structures
in this case lie between the bland centre and the degeneracy of the
frontiers.
 In modern theoretical cosmology, distance is absolute and the
universe expands. For reasons that are not yet understood, it
simultaneously becomes more richly structured. In a cosmology
without both time and scale, this would correspond, in a realistic
scale-free Platonia, to going from the bland centre to the more
interestingly structured ‘instants of time’ situated between it and the
frontiers. That is where the mist I introduced in Chapter 3 must
collect most thickly – have the highest intensity – at time capsules
structured so that they seem to record evolution from the symmetric
centre. This would be a cosmology of pure structure, an appealing
thought. The scalene and obtuse triangles that inhabit the ‘favoured
belt’ in Shape Space remind me of the line in Gerard Manley
Hopkins’s poem ‘Pied Beauty’, in which he praises ‘All things
counter, original, spare, strange’. In such a scheme, the bland centre
and degenerate frontiers still have a vital role to play in the scheme
of things. This is because a kind of resonance between all the instants
of time determines where the mist settles. Any acoustician will
recognize the importance of the walls and centre of a building in
determining its harmonies. Platonia, shown here as Shape Space, is a
‘heavenly vault’ in which the music of the spheres is played.
However, I should emphasize that the more realistic Shape Spaces
corresponding to universes with more than three particles are most
de nitely ‘open ended’ and should not be thought of as enclosed
spaces, as might appear from the simple example of Triangle Shape
Space. Platonia is not a claustrophobic vault but an ‘echoing canyon’
open to the sky. What is more, there is a sense in which its echo,
heard at any point within it, is what we call the past.

In completing this box, I note it has a nice unplanned symbolism.

Box 3 considers Shape Space, which is represented by the perfect
(equilateral) triangle, which itself, through each of its points,
represents all triangles, all of which are unities in the sense
mentioned on p.18. Shape Space illustrates Giordano Bruno’s monas
monadum, the unity of the unities.

If Laplace’s divinity contemplates a three-particle universe, its

history will be a curve in Triangle Land, which, omitting the triangle
size, we can show as a curve in Shape Space. The example of real
Newtonian three-body gravitational interaction shown in Figure 9
brings out very clearly the main fact that we associate with time,
that its instants come in a unique succession. This translates
beautifully into the winding path. (You can see why I am so indebted
to Dierck Liebscher, who crafted this diagram.) However, the other
two important attributes of time, duration and direction, are not yet
re ected in Figure 9. There are no marks along the curve to indicate
how much time elapses between any two points on it. It is also
impossible to say in which direction along the curve time increases.

What information must we give to determine a history of the kind

encoded in the path in Figure 9? According to Laplace, simply the
positions and velocities of the bodies at one instant (together with
their masses). Poincaré remarked, however, that the positions and
motions of the bodies are de ned in absolute space, and the speeds
of the bodies are de ned using absolute time. This is a subtle and
important quali cation.

If only relative quantities count, then Newton assumed too much

structure. In a universe of just three particles, only the three
distances between them (the triangle they form) should count. The
triangle universe cannot have an overall position and orientation in
some invisible containing space. Similarly, since the idea of an
external ‘grandstand clock’ is absurd, we cannot say ‘how fast’ the
universe travels along the curve in Figure 9. It simply occupies all
the points along it.

If Mach is right, so that time is nothing but change and all that
really counts in the world is relative distances, there should be a
perfect analogue of Laplace’s scenario with a divine intelligence that
contemplates Platonia. Machian dynamics in Platonia must be about
the determination of paths in that timeless landscape. It should be
possible to specify an initial point in Platonia and a direction at that
point, and that should be su cient to determine the entire path.
Nothing less can satisfy a rational mind. The history in Figure 10
starts at the centre of Shape Space, so that there the particles form an
equilateral triangle, and set o in a certain direction. In Machian
dynamics, the initial position and initial direction (strictly in
Triangle Land, not Shape Space) should determine the complete
curve uniquely. Now we can test this idea in the real world. The
heavens provide plenty of triple-star systems, and astronomers have
been observing their behaviour for a long time. They certainly meet
the Laplace-type condition when described in Newtonian terms. But
are their motions comprehensible from a Machian point of view?
This is the question Poincaré posed.

Figure 9 The (computer-generated) path traced in Shape Space by

the triangles formed by three mutually gravitating particles. It is
important to realize that the winding path shown is not traced by a
single particle moving over the page, but that each point on the path
represents the shape of a complete triangle. The curve shows the
succession of triangle shapes. As explained in the text, it is
impossible to say that time increases as you go in a particular
direction along this curve. However, suppose we imagine it starts in
the top left corner. This corresponds to particles A and C nearly
colliding, while particle B is far away. Then they move to a
con guration that is quite close to an equilateral triangle, after which
A and B get very close together near the bottom of the diagram. Then
the triangle shape evolves along the curve up to the top right. Where
the curve nearly touches the top line, all three particles are almost
on a line, with C between A and B. Finally the curve returns to the
bottom of the gure, where the wiggles indicate that particles A and
B are orbiting around each other, while C is far away. You see how
history is all coded in one curve, but you just cannot tell in which
direction it unfolds!

APPARENT FAILURE

The answer is very curious. The motions are nearly but not quite
comprehensible. This can be highlighted by showing how di erent
possible Newtonian motions look when represented as curves in
Triangle Land, our model Platonia – or rather Shape Space, since this
is much easier to represent. To create a vivid picture, let us imagine
that we are holding two cardboard triangles that are slightly
di erent. These can represent the relative con gurations of three
mutually gravitating bodies at two slightly di erent instants of
Newton’s absolute time.
Figure 10 Another possible path traced by the same three particles
as in Figure 9 (I refer to them now as bodies). This history starts (or
ends if time is assumed to run the other way) at the con guration in
which the three bodies form an equilateral triangle. The shape of the
triangle changes in a de nite way at all points along the curve. If it
left the initial equilateral triangle along one of the dotted lines, that
would mean two sides of the triangle remaining equal in length
while the third changes – the equilateral triangle would become an
isosceles triangle. In fact, for the example shown, the ratios of all
three sides change. For readers used to thinking of motions in
ordinary space, this example corresponds to particles that constantly
 orbit each other in a xed plane. The positions at which the curve
touches the dotted line that bound Shape Space correspond to
eclipses, when one particle is between the other two and on the line
joining them. Such a con guration is called a syzygy (that’s a nice
word to show o with). In ordinary space, the one particle passes
through the line joining the other two and comes out the other side.
But the points on the curve in Figures 9 and 10 stand for the
complete triangle, not one of the three particles. This is why the
curve approaches the syzygy frontier and then returns into the
interior of Shape Space. There are no triangles outside the syzygies!

Playing the role of Laplace’s divinity, we place the rst triangle,

at the instant when Newton’s grandstand clock says it is noon, at
some position in absolute space. A second later, we place the second
triangle somewhere near it in a slightly di erent position. The rst
triangle de nes the initial positions of the three bodies. Given the
position of the second triangle one second later, we can calculate the
initial motions, since we know where the particles have gone and
how long it took them. (Strictly, to calculate the instantaneous
velocities we must take an in nitesimal time interval, not one
second, but that is a minor detail). Imagine now that a strobe light
illuminates the bodies with a ash once every second, corresponding
to the seconds ticking on Newton’s clock, so that we can watch how
the triangle formed by the bodies moves through absolute space. We
have seen this already, in Figure 1. We can also plot the points
corresponding to the triangles in either Triangle Land or Shape
Space, obtaining a curve like those in Figures 9 and 10. This
abstracts away the extra Newtonian information – the positions in
absolute space and the time separations – that we possessed
originally.

Now, wherever we place the two triangles, the resulting curves in

either Triangle Land or Shape Space will all start at the same point,
since we always begin with the same triangle, and that corresponds
to just one xed point in Platonia. The curve must also have the
same initial direction, since that is determined by the position of the
second triangle in Platonia, which is also xed. This is explained in
the caption to Figure 10. The question is, how does the curve run
after that? What e ect do the positioning of the rst two triangles in
absolute space and the time separation have on the subsequent
evolution?

To answer this question, we need the notion of centre of mass

(Box 4). For a given triangle, there are two di erent things to bear in
mind when it comes to placing it in absolute space. First, we can
place its centre of mass anywhere. Since space has three dimensions,
this means that we can shift the centre of mass along three di erent
directions. Physicists say that in such cases there are three degrees of
freedom. Second, holding the centre of mass xed, we can change the
orientation of the triangle in space. This introduces three more
degrees of freedom. To see this, picture an arrow passing through the
centre of mass perpendicular to the triangle. It will point to
somewhere on the two-dimensional sky, giving two degrees of
freedom. The third arises because one can, keeping the arrow xed,
rotate the triangle around it as an axis.

BOX 4 Centre of Mass

The centre of mass of a system of bodies is the position of a ctitious

mass equal to the sum of the masses of the system. For two unequal
masses m and M, it lies on the line joining them at the position that
divides the line in the ratio M/m – that is, closer to the heavier mass
in that proportion. For any isolated system of bodies, the centre of
mass either remains at rest or moves uniformly in a straight line
through absolute space. The centre of mass for three bodies is shown
in Figure 11.
 Figure 11 Masses of 1, 2 and 3 units (indicated by their sizes) are
shown at the vertices of the two slightly di erent triangles (which
correspond to the two triangles discussed in the main text). The two
centres of mass are shown by the big blob (mass 6 units). The
position of the centre of mass is found by nding the centre of mass
of any pair and then the centre of mass of it and the remaining third
mass.

Now, wherever we place the centre of mass of the rst triangle,

and however we orient it, the sequence of triangles that then arises is
always the same. The path traced out in Platonia is the same. The
starting position in absolute space does not matter an iota. This is
rather remarkable. It is as if you could grow identical carrots in your
garden, at the bottom of the sea, and in outer space. Di erent
locations in absolute space have a decidedly shadowy reality. Unlike
real locations on the Earth, they do not have any observable e ects.

The ‘location in time’ is equally di cult to pin down. We started

our experiment at noon according to Newton’s clock. In fact, the
starting time has no in uence whatsoever: all the ‘carrots’ come out
just the same. So, as far as the position of the rst triangle in both
absolute space and time is concerned, it has no in uence whatsoever.
We begin to wonder whether they play any role at all. This doubt is
strengthened when we consider where to place the second triangle. It
turns out that we can position its centre of mass anywhere in
absolute space relative to the rst triangle. This too has no e ect at
all on the sequence of triangles that then follow. This absence of
e ect is due to so-called Galilean relativity, which is one of the most
fundamental principles of physics (Box 5).

BOX 5 The Galilean Relativity Principle

Galileo noted that all physical e ects in the closed cabin of a ship
sailing at uniform speed on a calm sea unfold in exactly the same
way as in a ship at rest. Unless you look out of the porthole, you
cannot tell whether the ship is moving. Quite generally, in
Newtonian mechanics the uniform motion of an isolated system has
no e ect on the processes that take place within it. The left-hand
diagram in Figure 12 shows (in perspective) the triangles formed by
the three gravitating bodies in the history of Figure 10 at equal
intervals of absolute time. The individual bodies move along the
‘spaghetti’ tubes. The centre of mass moves uniformly up the z axis.
(Despite appearances, the triangles are always horizontal, i.e. parallel
to the xy plane.) The right-hand diagram has two physically
equivalent interpretations. First, it is how observers moving
uniformly to the left past the system on the left would see that
system receding behind them. Second, it is also how observers at rest
relative to the system on the left would see a system identical to that
system except for a uniform motion of the centre of mass to the right.
This is how the happenings in the cabin of Galileo’s galley would be
‘sheared’ to the right for observers standing on the shore. Depending
on the speed of the system, the centre of mass will be shifted in unit
time by di erent amounts, but the actual sequence of triangles
remains the same. This corresponds to the freedom mentioned in the
text.

Because of the relativity principle, the laws of motion satis ed by

bodies take exactly the same form in any frame of reference moving
uniformly through absolute space as they do in absolute space itself.
Although Newton did not like to admit it, this fact makes it
impossible to say whether any such frame, which is called an inertial
frame of reference, is at rest in absolute space or moves through it
with some uniform velocity. Bodies with no forces acting on them
move in a straight line with uniform speed in any inertial frame of
reference (hence the name). It is impossible to say that you are at
rest in absolute space, only that you are at rest in some inertial frame
of reference. For historical reasons I use ‘absolute space’ in the text,
but strictly I should be using ‘any inertial frame of reference’.

Figure 12 Unlike Figures 9 and 10 (and the later Figure 14), the
lines followed by the spaghetti strands in this gure (and also Figure
13) show the tracks of the three individual particles in space. This is
why there are three strands and not a single curve. It will help you a
lot if you can get used to thinking about these two di erent ways of
representing one and the same state of a airs. Here we see
individual particles moving in absolute space. In Figures 9, 10 and
 14 we ‘see’ (in our mind’s eye) the ‘world’ or ‘universe’ formed by
the three particles moving in Platonia.

There are only four freedoms that remain. Having placed the
centre of mass of the second triangle at some position, we can
change its orientation (three freedoms). We can also change the
amount of Newton’s absolute time that elapses between the instants
at which the three bodies occupy the two positions (one freedom, the
fourth). If the time di erence is shortened, this means that the bodies
travel farther in less of Newton’s time – that is, they are moving
faster initially. In fact, since the motion of the centre of mass does
not matter, we can keep it xed and change only the orientation.
Now, at last, we come to something that does matter. Both these
changes – in the time di erence and in the relative orientation –
have dramatic consequences, which are illustrated in Figures 13 and
14.

Figures 13 and 14 express the entire mystery of absolute space

and time. Both of Newton’s absolutes are invisible, yet their e ects
show up in the evolutions of the triangles, which are more or less
directly visible. The astronomers do see stars and the spaces between
them (admittedly in projection) when they look through telescopes.
If time were merely change and only distances had dynamical e ect,
a decent Machian mechanics – one that would satisfy Laplace’s
divine intelligence – should lead to exactly the same evolutions in all
nine cases. This is manifestly not true for the real triple-star systems
that astronomers observe. All the di erent kinds of evolution shown
in Figures 13 and 14, and many more, are found. All the facts that
enabled Newton to win his argument against Leibniz are contained in
these diagrams, but it took about two centuries before Poincaré
found the best way to demonstrate them. He concluded regretfully
that a mechanics that uses only relative quantities, as Mach
advocated, cannot get o the ground. It lacks perfect Laplacian
determinism. Nevertheless, the failure is curious. Absolute space and
time could have had an e ect through all the freedoms allowed in
the placing of the two triangles. There are fourteen degrees of
freedom in total, of which ten have no e ect whatsoever. This is just
what the invisibility of space and time would lead us to expect. Yet
four degrees of freedom do have a profound in uence. Three are
associated with twists in space, the fourth with the overall speed put
into the system. These strange mismatches between expectation and
reality have kept the philosophers arguing and the physicists
puzzling for centuries.

The fact is that Newton’s absolute space and time play a

decidedly odd role. The rst problem is their invisibility. The more
serious problem is what little part they play in the whole story, and
how irrationally they enter the stage when they do participate in the
action. Once we have chosen the relative orientation and time
separation of the two triangles, we can take them anywhere in
absolute space and time. They will always give rise to the same
evolution. Absolute space and time seem to matter very little; only
the relative orientation and time separation count.
 Figure 13 These are ‘spaghetti diagrams’ of evolutions in absolute
space like the left-hand one in Figure 12 (the one at the top left is the
same evolution but with the triangles removed). The corresponding
curves in Shape Space are shown in Figure 14. In each diagram the
evolution commences with the three bodies forming an equilateral
triangle, and all the corresponding curves start in the same direction
in Triangle Land and Shape Space. This is because the second
triangle is the same in both cases. The di erent evolutions are
created by giving the bodies di erent initial speeds (they are
di erent in the three rows) and by giving the triangles di erent
orientational twists (di erent in the three columns).
Figure 14 These are the curves in Shape Space corresponding to the
nine evolutions in absolute space shown in Figure 13. They all start
from the same point with the same direction, but then diverge
strongly. Remember, as I explained in the caption to Figure 9, that
these curves represent not the motion of a single particle across the
page, but the shapes of continuous sequences of triangles. If you
 ‘stuck a pin’ into any point on one of these curves, the triangle
corresponding to it would ‘light up’. It is very important to
appreciate that Figures 13 and 14 show identical happenings in two
di erent ways. Since Newton’s time, nearly all physicists have
believed the Newtonian representation, Figure 13, to be the
physically correct way to think about these things.
   Following Leibniz and Mach, I believe Figure 14 is the right way.
However, this approach faces a severe di culty explained in the
text. It is only in Chapter 7 that I shall explain how it is overcome.

But these are our arbitrary choices. Once we have chosen two
triangles, nothing about the triangles in themselves gives any hint as
to how we should make the choices. Leibniz formulated two great
principles of philosophy that most scientists would adhere to. The
rst is the identity of indiscernibles: if two things are identical in all
their attributes, then they are actually one. They are the same thing.
The second, which we have already met, is the principle of su cient
reason: every e ect must have a cause. There must be some real
observable di erence that explains di erent outcomes.

Now we can see the problem. Considered in itself, the pair of

triangles is just one thing. Each di erent relative orientation and
time separation we give them depends on our whim. They should not
have any e ect. Yet each has momentous consequences: they create
quite di erent universes. An exactly analogous problem arises if the
universe consists of any number of particles. Two snapshots (the
analogues of the two triangles) of the relative con guration of the
universe are never quite enough to determine an entire history
uniquely.

Before we look at the one possibility that can resolve this puzzle,
it is worth considering how the four freedoms that do count show up
in practice. We shall then be able to see what a great discovery
Newton’s invisible framework was. We start with the twists.

SPACE AND SPIN

When I was a boy, there was only one sport at which I was any good:
the high jump. One year I went on a training course at the athletics
ground in Oxford. We were introduced to angular momentum and
how it could be exploited to improve the jump. As tallest of the
young hopefuls, I was chosen to give a demonstration. The instructor
made me lie on my side, arms and legs outstretched, on a small
bench turntable. He started to rotate it slowly and asked, ‘If you pull
yourself into a crouched position, what will happen?’ I knew, I was
studying physics: ‘Angular momentum will be conserved, and the
turntable will spin faster.’ ‘Right,’ he said, ‘do it.’ Proudly, I pulled in
my arms and legs with vigour. The e ect was frightening. The
turntable whizzed around so fast that I panicked, tried to get o , and
was thrown onto the oor. I escaped with bruises. I am still kicking
myself, not about the accident, but because I did not stay on another
day. I would have seen Roger Bannister run the rst four-minute
mile.
 Angular momentum is a kind of net spin about a xed axis. To
calculate it for the Earth, you multiply the mass of each piece of
matter in the Earth by its perpendicular distance from the rotation
axis and the speed of its circular motion about the axis. The Earth’s
total angular momentum is the sum of the contributions of all the
pieces. Clockwise and anticlockwise motions count oppositely. A jet
plane ying round the world in the opposite sense to the daily
rotation contributes with the opposite sign.

By Newton’s laws, this net spin cannot change for an isolated

system. This universal law applies equally to humans and planets.
When I pulled in my arms and legs in Oxford, I abruptly reduced the
distance of much of my mass from the rotation axis. This inescapably
enforced an equally abrupt increase of my rotational speed – with its
unfortunate consequences. The same law explains why the Earth’s
rotation axis stays xed, pointing towards the pole star, and why the
length of the day, the rotation period, does not change. The rotation
speed could change only if the Earth could expand or contract, but,
being rigid, it cannot. (Actually, both the axis and the day do change
very slowly due to the external in uence of the Sun and Moon.) For
rigid bodies like the Earth and a top, the e ects of angular
momentum are rather obvious. However, its e ects are far-reaching.

A globular cluster may contain a million stars. It has no rigidity –

all its stars move individually in di erent directions, though gravity
holds the cluster together. Its angular momentum is found by
choosing three mutually perpendicular axes, and calculating the net
spin around each of them. These correspond exactly to the three
degrees of freedom to make twists, mentioned in the previous
section. However, the three axes can always be chosen in such a way
that the spin about two of them is zero, and all the net spin is thus
about a single axis. This axis is a kind of arrow that points in a
certain direction in space. It and the net spin remain completely
unchanged as time passes. In astronomy, time passes in aeons. Since
the stars all move in di erent directions, the bookkeeping exercise
that nature performs is remarkable. A deep principle is at work.

The laws of nature are seldom seen to be operating in a pure

form, and are hard to recognize. Air resistance and friction distort
the basic laws of mechanics. But the greatest di culty arises because
the laws involve time, and we experience only one instant at a time.
If only we could see all the instants of time stretched out before us,
we could see the e ects of the laws of motion directly, as in some of
the diagrams earlier in the book.

However, a few phenomena reveal mechanics at work in a

striking fashion. They are often associated with angular momentum.
The humble top is one of the best examples. Riding a bicycle is
another: the reassuring way in which balance is maintained as you
speed down a hill with the air rushing past you is down to the
angular momentum in the spinning wheels. Once the wheels are
turning fast, they have a strong tendency to keep their axis of
rotation horizontal. Indeed, a child’s hoop illustrates beautifully how
the rotation axis maintains a xed direction. So does the frisbee,
spinning true as it oats through the air. Much grander examples
occur naturally. I have already mentioned the earth’s rotation, which
we see as the rising and setting of the Sun, Moon and stars and their
ceaseless march across the sky. Many of our images of time come
from this phenomenon, the child’s top writ large.

However, in all these examples there is a rigid body. The example

of globular clusters tells of a mighty invisible framework behind the
all too elusive phenomena. Newton knew it was there long before the
astronomers found the grandest examples of its handiwork: spiral
galaxies. In them, the initially invisible e ects of the framework have
become visible. Indeed, any isolated collection of matter, whatever
its nature – a million stars in a globular cluster or a huge cloud of
dust in space – has its associated xed axis of net spin. Laplace called
the plane perpendicular to it through the centre of mass the
invariable plane, because its orientation can never change. Sometimes
it can actually be seen. This is because some motions can be changed
or even lost through mutual interactions, whereas others cannot. For
example, objects moving parallel to the spin axis in opposite
directions may collide and be de ected into the invariable plane.
Over time, the matter in the system can ‘collect’ in or near it
provided there is still the correct amount of circular motion about
the axis. This has happened in spiral galaxies, in which the bright
stars in the spiral arms are formed from such accumulated matter.
They ‘light up’ the invariable plane, making it visible (Figure 15).
 A similar e ect has been at work in the solar system. About four
and a half billion years ago the Sun and planets formed from a huge
cloud of dust left over from a supernova explosion. The dust had
some net spin, and an associated invariable plane. The Sun formed
near the cloud’s centre of mass, and gathered up most of the mass in
the cloud. More or less all of the solar system’s rotation now takes
place in the plane of the ecliptic, in which the Earth moves around
the Sun. Although the Sun got the bulk of the mass, Jupiter has most
of the spin.

Figure 15 A spectacular spiral galaxy seen ‘from above’.

The fact that all the planets move in the same direction around
the Sun in nearly coincident planes is thus a remote consequence of
the relatively modest initial net spin of the primordial dust cloud. We
see the result in the sky, since all the celestial wanderers – the Sun,
Moon and planets – follow much the same track against the
background of the stars. Ironically, Newton underestimated the
power of his own laws. He could not bring himself to believe that the
solar system had arisen naturally. ‘Mere mechanical causes’, he said,
‘could not give birth to so many regular motions.’ He asserted that
‘this most beautiful system’ could only have proceeded ‘from the
counsel and dominion of an intelligent and powerful Being’. One
wonders what Newton would have made of the modern pictures of
Saturn and its rings (Figure 16). Of all the images created in the
heavens by gravity and the invariable plane, this is surely the most
perfect.

For three centuries, the best explanation for phenomena like the
rings of Saturn has remained Newton’s: inertia, the inherent
tendency of all objects to follow straight lines in the room-like arena
of absolute space. If these are accepted, then the rings of Saturn,
tops, frisbees and all the other manifestations of angular momentum
can be explained. However, Newton’s account is not so much an
explanation as a statement of facts in need of explanation. Since it is
always matter that we actually see, should we not try to account for
these things without the mysterious intermediaries of absolute space
and time? Before we attack this problem, we need to consider energy
and, in the next chapter, clocks and the measurement of time.
 Figure 16 Saturn and its rings.

ENERGY

Energy is the most basic quantity in physics. It comes in two forms:

kinetic energy measures the amount of motion in a system, while
potential energy is determined by its instantaneous con guration.
Like angular momentum, in an isolated system the sum of the two
remains constant. If one decreases, the other must increase. For
example, the potential energy of a falling body is proportional to its
height and decreases as it falls. The speed of descent, and with it the
kinetic energy, increases by an exactly compensating amount.
 Energy, like the whole of mechanics, has a curious hybrid nature.
Absolute space and time are needed to calculate kinetic but not
potential energy. Each body of mass m and speed v in a system
contributes a kinetic energy ½mv2. The speed is measured in
absolute space, which is why it is needed to calculate kinetic energy.
By contrast, the potential energy of a system depends only on its
relative con guration. For example, each pair of gravitating bodies
in a system contributes to the system’s total potential energy an
amount that is inversely proportional to their separation. If this is
doubled, the potential energy of the pair is halved. Since each point
in any Platonia corresponds to a di erent con guration of bodies,
the potential energy changes from place to place in Platonia. This is
illustrated for three bodies in Figure 17.
 Figure 17 The gravitational potential energy of three bodies of
di erent masses is shown as the height of a surface above Shape
Space (Figure 8), each point on which corresponds to a di erent
shape of the triangle formed by the three bodies. The overall scales
of the con gurations on the right are nine times greater, so the
magnitude of the potential energy is much lower. Since potential
energy is inversely proportional to separation, it increases sharply
towards the corners of Shape Space, corresponding to two-particle
coincidences, and becomes in nite at them. As this cannot be shown
 in the gure, the surfaces have been cut o at a certain height. The
most distant corner of Shape Space corresponds in this gure to
coincidence of the two most massive particles, so this is why the
potential increases most strongly there.

Like angular momentum, the energy a ects the appearance of

systems and the behaviour of individual objects. For gravity the
potential energy is negative, while the kinetic energy is positive.
Thus the total energy E can be either positive, zero or negative. If a
spacecraft is launched with su cient speed, it can escape from the
Earth’s gravity because its E is positive. If E is zero, the spacecraft
has exactly the escape velocity, and escape is just possible. If E is
negative, the spacecraft cannot escape from the Earth and will either
orbit the Earth or fall back to ground. The planets can never escape
from the Sun because they have negative E. Star clusters can remain
concentrated in a relatively small region of space only if their energy
is negative, otherwise they would rapidly disperse. This is why we do
see such ne objects as the galaxies and star clusters in the sky. It is
also largely the reason why the Sun and planets have their beautiful
round shapes.

Thus, the shapes of almost all the objects astronomers observe in

the sky re ect their energy and angular momentum. They, in turn,
seem impossible to explain unless absolute space and time do exist
and have a real in uence, just as Newton claimed. The evidence for
Newton’s invisible framework is written all over the sky. The
evidence can be summarized as the two-snapshots problem. Suppose
that snapshots of an isolated system taken at two closely spaced
instants show only the separations of its bodies, not the overall
orientations in absolute space. The separation in time between the
snapshots is also unknown. If the system is a globular cluster, the
snapshots contain millions of data. However, to determine the
evolution of the system, four pieces of data are still lacking. They
determine the kinetic energy (one piece of data) and the angular
momentum (three pieces of data). Although they cannot be deduced
from the two snapshots, they have a huge in uence on the evolution,
which can often be seen at a glance. A third snapshot will yield the
data, but also much redundant information. The four missing pieces
of data comprise the entire evidence for absolute space and time.
Every system in the universe proclaims their existence. This seems to
make nonsense of my claim that time does not exist. There appears
to be more to the universe than its relative con gurations. There is
invisible structure, of which no trace can be found in Platonia.
CHAPTER6
The Two Great Clocks in the Sky

WHERE IS TIME?

Newton’s mysterious ‘timepiece’ and speeds measured relative to it

gured prominently in the last chapter. But is it really there, and
how can we ever read its time if it is invisible? This chapter is about
these two questions.

A simple but famous experiment of Galileo provides strong

evidence for something very like Newton’s absolute time. He rolled a
ball across a table and o its edge. His analysis of its fall was a major
step in mechanics. First he noted the ball’s innate tendency to carry
on forward in the direction it had followed on the table. It also
started to fall under gravity, picking up speed. Galileo conjectured
that two processes were at work independently, and that each could
be analysed separately. The total e ect would be found by simply
adding the two processes together.

Galileo’s recognition of the tendency to keep moving forward

anticipated Newton’s law of inertia. He did not recognize it as a
universal law, but he did make it precise in some special cases. For
the example of the ball, he conjectured that but for gravity (and air
resistance) the ball would move for ever forward with uniform
speed. (He actually thought that the motion would be around the
Earth – Galileo’s inertia was circular. Luckily, the di erence was far
too small to a ect his analysis.)

As for the second process, Galileo had already found that if an

object is dropped from rest and in the rst unit of time falls one unit
of distance, then in the next it will fall a further three, in the next
ve, and so on. He was entranced by this, and called it the odd
numbers rule. Now consider the sequence:

at t = 1, distance fallen = 1,
at t = 2, distance fallen = 1 + 3 = 4,
at t = 3, distance fallen = 1 + 3 + 5 = 9,
at t = 4, distance fallen = 1 + 3 + 5 + 7 = 16,…

The distance fallen increases as the square of the time: 12 = 1, 22 =

4, 32 = 9, 42 = 16,... . Galileo’s originality was to seek for a deeper
meaning in this pattern.

Many teenagers can now do in seconds a calculation that took

Galileo a year or more – it was so novel. He asked: if the distance
fallen increases as the square of the time, how does the speed
increase with time? He eventually found that it must increase
uniformly with time. If after the rst unit of time the object has
acquired a certain speed, then after the second it will have twice that
speed, after the third three times, and so on. Galileo’s work showed
that, in the absence of air resistance, a falling body always has a
constant acceleration. It never ceases to amaze me what
consequences owed from Galileo’s simple but precise question. It
taught his successors how to read the ‘great book of nature’ (Galileo’s
expression). From a striking empirical pattern, he had found his way
to a simpler and deeper law.

To analyse the falling ball, Galileo simply combined the two

processes – inertia and falling – under the assumption that each acts
independently. He obtained the famous parabolic motion (Figure
18). In each unit of time, the ball moves through the same horizontal
distance, but in the vertical dimension the distance fallen grows as
the square of the time. The resulting curve traced by the ball is part
of a parabola. Newton applied Galileo’s method for terrestrial
motions to the heavens, and showed that the laws of motion had
universal validity. This was the rst great uni cation in physics.
There may be a lesson for us here in our present quest – the search
for time. We may have to look for it in the sky.

A search is needed. It is striking that all the elements in Galileo’s

analysis are readily visualized. You can easily call up a table and the
parabola traced by the ball in your imagination. Yet one key player
seems reluctant to appear on the stage. Where is time? This is the
question I have so far dodged. It presents a severe challenge to the
idea that con gurations are all that exist. Suppose that we take
snapshots of the ball as it rolls across Galileo’s table in Padua, where
he experimented. These snapshots can show everything in his studio.
However, only the ball is moving. We take lots of snapshots, at
random time intervals, until the ball is just about to fall over the
edge. We put the snapshots, all mixed up, in a bag and, supposing
time travel is possible, present it to Galileo and ask him whether, by
examining the snapshots, he can tell where the ball will land.

He cannot. Had we rolled the ball twice as fast, it would have

passed through the identical sequence of positions to the table’s
edge, and they are all the snapshots capture. The speed is not
recorded. But the ball’s speed determines the shape of the parabola,
and hence where the ball lands. In fact Galileo will not even know in
which direction the ball is going. Perhaps it will fall o the right-
hand side of the table. More clearly than with the three-body
evolutions, which mix the e ects of time and spin, we see here the
entire evidence for absolute time. The speed determines the shape of
the parabola. There is manifestly more to the world than the
snapshots reveal. What and where is it? Galileo himself provides an
answer of sorts. He tells us that he measured time by a water-clock –
a large water tank with a small hole in the bottom. His assistant
would remove a nger from the hole and let the water ow into a
measuring ask until the timed interval ended. The amount of water
measured the time.
 Figure 18 Galileo’s own diagram of parabolic motion. The ball
comes from the right and then starts to fall. Incidentally, this
diagram illustrates how conventions get established and become
rigid – a modern version of it would certainly show the ball coming
from the left and falling o on the right.) The uniformity of the
horizontal inertial motion is shown by the equality of the intervals
be, cd, de, ..., while the odd-numbers rule is re ected in the
increasing vertical descents bo, og, gl, ....

We have only to include the water tank and assistant in the

snapshots, and everything is changed. Galileo can tell us where the
ball will land because he can now deduce its speed. There are some
important lessons we can learn from this. First, it is water, not time,
that ows. Speed is not distance divided by time but distance divided
by some real change elsewhere in the world. What we call time will
never be understood unless this fact is grasped. Second, we must ask
what change is allowed as a measure of time. Galileo measured the
water carefully and made sure that it escaped steadily from the tank
– otherwise his measure of time would surely have been useless. But
the innocent word ‘steadily’ itself presupposes a measure of time.
Where does that come from? It looks as if we can get into an
unending search all too easily. No sooner do we present some
measure that is supposed to be uniform than we are challenged to
prove that it is uniform.

It is an indication of how slowly basic issues are resolved – and

how easily they are put aside – that Newton highlighted the issue of
the ultimate source of time nearly two hundred years before serious
attempts were made to nd it. Even then, the attempts remained
rather rudimentary and few scientists became aware of them. It is
interesting that Galileo had already anticipated the rst useful
attempt. This was actually forced upon him by the brevity of free fall
in the ball experiment: it was all over much too quickly for the
water-clock to be of any use. (It came into its own when Galileo
rolled balls down very gentle inclines.) To analyse the parabola, he
found a handy substitute. He noted that if the horizontal motion of
the falling ball does persist unchanged, then the horizontal distance
traversed becomes a direct measure of time. He therefore used the
horizontal motion as a clock to time the vertical motion. His famous
law of free fall was then coded in the shape of the parabola. Its
de ning property is that the distance down from the apex (where the
ball falls o the table) increases as the square of the horizontal
distance from the axis. But this measures time.

Thus, time is hidden in the picture. The horizontal distance

measures time. It would be nice if one could say ‘the horizontal
distance is time’. This is the goal I am working towards: time will
become a distance through which things have moved. Then we shall
truly see time as it ows, because time will be seen for what it is –
the change of things. However, there are many di erent motions in
the universe. Are they all equally suitable for measuring time? A
second question is this: what causal connections are at work here?
Galileo measured time by the ow of water, but it is hard to believe
that a little water owing out of a tank in the corner of his studio
directly caused the balls to trace those beautiful parabolas through
the air. If time derives from motion and change – and it is quite
certain that all time measures do – what motion or change, in the
last resort, is telling the ball which parabola to trace? The rst
question is more readily answered.

THE FIRST GREAT CLOCK

Nearly two thousand years ago, astronomers knew that some motions
are better than others as measures of time. This they discovered
experimentally. For the early astronomers, there were two obvious
and, on the face of it, equally good candidates for telling time. Both
were up in the sky and both had impressive credentials. The stars
made the rst clock, the Sun the second.
 The stars remain xed relative to each other and de ne sidereal
time. Any star can be chosen as the ‘hand’ of the stellar clock: one
merely has to note when it is due south. The stellar clock then ticks
whenever that star is due south (i.e. when it crosses the meridian).
Fractions of the ‘tick unit’ are measured by its distance from the
meridian. A mere glance at the night sky could tell the ancient
astronomers the time to within a quarter of an hour. With some care,
times could be told to a minute. There is something wonderful about
this great clock in the sky. It was a unique gift to the astronomers.
The discoveries that culminated with Kepler’s laws of planetary
motion, and many more made until well into the twentieth century,
are unthinkable without it. No other phenomenon in nature could
match it for convenience and accuracy. In millennia it has lost a few
hours.

But there is a rival – the Sun. It de nes solar time. This is the
clock by which humanity and all other animals have always lived.
The principle is the same: it is noon when the sun crosses the
meridian. You don’t even have to be an astronomer to tell the time
by this clock; a sundial will do.

Merely describing the clocks shows that speed is not distance

divided by time, but distance divided by some other real change,
most conveniently another distance. Roger Bannister ran one mile in
four minutes; normal mortals can usually walk four miles in one
hour. What does that mean? It means that as you or I walk four
miles, the sun moves 15° across the sky. But this is not quite the
complete story of speed and time, because there is a subtle di erence
between the two clocks in the sky – they do not march in perfect
step. One and the same motion will have a di erent speed depending
on which clock is used. One di erence between the clocks is trivial:
the solar day is longer than the sidereal. The Sun, tracking eastwards
round the ecliptic, takes on average four minutes longer to return to
the meridian than the stars do. This di erence, being constant, is no
problem. However, there are also two variable di erences (Box 6).

BOX 6 The Equation of Time

The rst di erence between sidereal and solar time arises from one
of the three laws discovered by Kepler that describe the motion of
the planets. The Sun’s apparent motion round the ecliptic is, of
course, the re ection of the Earth’s motion. But, as Kepler
demonstrated with his second law, that motion is not uniform. For
this reason, the Sun’s daily eastward motion varies slightly during
the year from its average. The di erences build up to about ten
minutes at some times of the year.

The second di erence arises because the ecliptic is north of the

celestial equator in the (northern) summer and south in the winter.
The Sun’s motion is nearly uniform round the ecliptic. However, it is
purely eastward in high summer and deep winter, but between,
especially near the equinoxes, there is a north-south component and
the eastward motion is slower. This can lead to an accumulated
di erence of up to seven minutes.
 The e ects peak at di erent times, and the net e ect is
represented by an asymmetric curve called the equation of time (it
‘equalizes’ the times). In November the Sun is ahead of the stars by
16 minutes, but three months later it lags by 14 minutes. This is why
the evenings get dark rather early in November, but get light equally
early in January. The stars, not the sun, set civil time.

Since the Sun is much more important for most human a airs
than the stars, how did the astronomers persuade governments to
rule by the stars? What makes the one clock better than the other?
The rst answer came from the Moon and eclipses. Astronomers have
always used eclipse prediction to impress governments. By around
140 BC, Hipparchus, the rst great Greek astronomer, had already
devised a very respectable theory of the Moon’s motion, and could
predict eclipses quite well.

Now, in the timing and predicting of eclipses, half an hour makes

a di erence. They can occur only when the Moon crosses the ecliptic
– hence the name – and the Moon moves through its own diameter in
an hour. There is not much margin for error. By about AD 150, when
Claudius Ptolemy wrote the Almagest, it was clear that eclipses came
out right if sidereal, not solar time was used. No simple harmonious
theory of the Moon’s motion could be devised using the Sun as a
clock. But the stars did the trick.

What Hipparchus and Ptolemy took to be rotation of the stars we

now recognize as rotation of the Earth. It is strikingly correlated with
the Moon’s motion. Even more striking is the correlation established
by Kepler’s second law, according to which a line from the Sun to a
planet sweeps out equal areas in equal intervals of sidereal time.
Whenever astronomers and physicists look carefully, they nd
correlations between motions. Some are simple and direct, as
between the water running out of Galileo’s water-clock and the
horizontal distance in his parabolas; others, especially those found by
the astronomers, are not nearly so transparent. But all are
remarkable.

If two things are invariably correlated, it is natural to assume that

one is the cause of the other or both have a common cause. It is
inconceivable, as I said, that water running from a tank in Padua can
cause inertial motion of balls in northern Italy. It is just as
inconceivable that the spinning Earth causes the planets to satisfy
Kepler’s second law. Kepler, in fact, thought that it arose because all
the planets were driven in their orbits by a spinning Sun, but we
must look further now for a common cause. We shall nd it in a
second great clock in the sky. This will be the ultimate clock. The
rst step to it is the inertial clock.

THE INERTIAL CLOCK

The German mathematician Carl Neumann took this rst step to a

proper theory of time in 1870. He asked how one could make sense
of Newton’s claim, expressed in the law of inertia, that a body free of
all disturbances would continue at rest or in straight uniform motion
for ever. He concluded that for a single body by itself such a
statement could have no meaning. In particular, even if it could be
established that the body was moving in a straight line, uniformity
without some comparison was meaningless. It would then be
necessary to consider at least two bodies. He introduced the idea of
an inertial clock. He supposed that one body was known to be free of
forces, so that equal intervals of its motion could then be taken to
de ne equal intervals of time. With this de nition, it would be
possible to see if the other body, also known to be free of forces,
moved uniformly. If so, then in this sense Newton’s rst law would
be veri ed.

Neumann’s idea illustrates the truth that time is told by matter –

something has to move if we are to speak of time. Unfortunately, it
left unanswered at least three important questions. How can we say
that a body is moving in a straight line? How can we tell that it is
not subject to forces? How are we to tell time if we cannot nd any
bodies free of forces?

The answers to these questions will tell us the meaning of

duration. If we leave aside for the moment issues related to Einstein’s
relativity theories and quantum mechanics, time as we experience it
has two essential properties: its instants come in a linear sequence,
and there does seem to be a length of time, or duration. I have tried
to capture the rst property by means of model instants. A random
collection of such model instants would correspond to points
scattered over Platonia. They would not lie on a single curve, and the
fact that they do is, if veri ed, an experimental fact of the utmost
importance. It enables us to talk about history.

But what enables us to talk so con dently of seconds, minutes,

hours? What justi cation is there for saying that a minute today has
the same length as a minute tomorrow? What do astronomers mean
when they say the universe began fteen billion years ago?
Conditions soon after the Big Bang were utterly unlike the conditions
we experience now. How can hours then be compared with hours
now? To answer this question, I shall rst assume that there are no
forces in the world and that the only kind of motion is inertial. This
simpli cation already enables us to get very close to the essence of
time, duration and clocks. Then we shall see what forces do.

Suppose Newton claims that three particles, 1, 2 and 3, are

moving purely inertially and that someone takes snapshots of them.
These snapshots show the distances between the particles but
nothing else (except for marks that identify the particles). We know
neither the times at which the snapshots were taken nor any of the
particles’ positions in absolute space. How can we test Newton’s
assertion? We shall be handed a bag containing triangles and told to
check whether they correspond to the inertial motion of three
particles at the corners of the triangles. The Scottish mathematician
Peter Tait solved this problem in 1883 (Box 7).

BOX 7 Tait’s Inertial Clock

Tait used the relativity principle (Box 5) to simplify things. If the
particles are moving inertially, one can always suppose that particle
1 is at rest; it is shown in Figure 19 as the black diamond where the
three coordinate axes x, y, z in absolute space meet. Now, unless the
particles collide at some time – and we need not bother about this
exceptional case – there must come a time at which particle 2 passes
1 at some least distance a. The coordinate axes can be chosen so that
the line of its motion is as shown by the string of black diamonds.
We can choose the unit of time so that particle 2 has unit speed. It
will be Neumann’s clock, and each unit of distance it goes will mark
one unit of time. Several positions of particle 2 are shown. They are
the ‘ticks’ of the clock. At time t = 0, let particle 2 be at the point
closest to particle 1 (at the black diamond on the x axis). At this
time, particle 3 can be anywhere (three unknown coordinates) and
have any velocity (three more unknowns). Thus, seven numbers are
unknown: six for particle 3 and the distance a.
 Figure 19 The arrangement of coordinates in Tait’s problem.

Now, each snapshot contains three independent data – the three

sides of the triangle at the instant of the snapshot. It would seem that
three snapshots give nine data – more than enough for the
accomplished Tait to solve the problem. But since we know none of
the times at which the snapshots are taken, each gives only two
useful data. Thus, four snapshots will give eight useful data, seven of
which will establish the Newtonian frame into which the triangles
 t, while the remaining one will verify that they are indeed obeying
Newton’s law. Figure 20 shows a typical solution.

Figure 20 A solution of Tait’s problem. The four ‘snapshots’ of the

triangles (the given data) are shown in plan, with an indication of
their positions in the ‘sculpture’ of four triangles created by the
solution.
 Tait’s solution creates nothing less than Newton’s invisible
framework. It is both absolute space and the positions of the
triangles in it at di erent instants of ‘time’. We start with triangles.
They are all that we have. We are told that they are not random
triangles but have arisen through a law. We test this assertion and
succeed in creating a simple ‘sculpture’ of three straight lines. The
order created is dramatic. Apart from exceptional cases, the distances
of the particles from one another do not change in a mutually
uniform manner. In fact, they vary with respect to one another in a
quite complicated way. Moreover, the triangles in themselves give no
hint that there is any space in which their corners lie on straight
lines. Yet such lines can be found.

It is equally remarkable that the motions along these lines are

mutually uniform. Each particle is the ‘hand’ of a clock for the
motion of the other two. The sculpture is a clock with many hands
(two in this case, since particle 1 de nes the origin). And, now that
we have the rigid structure, we see that absolute space is redundant.
The sculpture holds together on its own. The ‘room’ was never there
until it was created from the triangles and rules. It is they that give
an almost bodily tangibility to space and time.

They also explain the meaning of duration and the statement that
a second today is the same as a second tomorrow. Duration is
reduced to distance. If today or tomorrow any one of the ‘hands’ of
the inertial clock moves through the same distance, then we can say
that the ‘same amount of time’ has passed. The extra time dimension
is redundant: everything we need to know about time can be read o
from distances. But note how special is the distance that leads to a
meaningful de nition of duration. Any change of distance ‘labels’ the
instants of time. In statements like ‘Particle A hits B when C is ve
metres from D’, ‘ ve metres’ identi es an instant of time – it labels
the instant. That is su cient for history. However, the obvious
changes of distance – those between particles – do not lead to a
sensible de nition of duration. The secrets of time are rather well
hidden.

A similar construction can be repeated for any number of

particles – a hundred, a billion, quite enough to ll the sky and make
a galaxy or even a universe. It is important that if there are more
than ve particles then three snapshots are already su cient to solve
Tait’s problem, but two are never enough. This is very odd. For a
thousand bodies, three snapshots contain far more information than
we need, but two never give quite enough. The problem is exactly
the one we encountered earlier. Two snapshots tell nothing about the
relative orientations or the separation in time. We lack four pieces of
information, and all the mystery of absolute space and time resides
in them. We cannot make a clock until we get our hands on them.
But when we have them, the properties that are revealed are very
striking.

For example, Tait’s construction is a good model for the motions

of many thousands of stars that are relatively close neighbours of our
Sun. They all ‘fall’ in the gravitational eld of the Galaxy in much
the same way. That motion hardly shows up in the relative
separations. But also, because the stars are so far apart there is very
little gravitational interaction between them. Their motions are thus
well described by the construction. What is more, any three stars can
be chosen to make a ‘Tait clock’ and tell the time. Any other three
will make another. Thousands, millions of such clocks can be made.
All these clocks, light years apart, keep time with one another.

I mentioned earlier the importance of not being misled by the

special circumstances of our existence. One of them is the Earth.
Only the tiniest fraction of the matter in the universe is in solid form.
Indeed, only a small fraction of the Earth – its crust, on which we
live, and the innermost core – is solid. This is our home, and we take
it for the normal run of things. The ground, trees, buildings, hills and
mountains make a framework, which is so like absolute space. It
does seem quite natural that a body should move in a straight line in
such a space. But we need to think what the universe in its totality is
like. Take a billion particles and let them swarm in confusion – that
is the reality of ‘home’ almost everywhere in the universe. The stars
do seem to swarm, so do the atoms in the stars. To understand the
real issues of timekeeping, we must imagine trying to do it in typical
circumstances. We must master celestial timekeeping and not be
content with the short cuts that can be taken on the Earth, for they
hide the essence of the problem.

THE SECOND GREAT CLOCK

We have seen how to check whether bodies are moving inertially
without prior access to absolute space and time. But all matter in the
universe interacts. Interactions make things more complicated, and
not only because the calculations are hard. If objects are moving
inertially, any three will su ce to construct an inertial clock. But in
a system of interacting bodies it is not possible to treat any of them
separately because each is a ected by the others. In addition, we can
nd no framework at all in which the bodies move uniformly in
straight lines.

There are three parts to a clock: a mechanism, a clock-face and

hands. The main problem of celestial timekeeping arises because the
clock-face is invisible. A further problem is that the hands run at
varying rates. Imagine an isolated system of three gravitationally
interacting stars. We are again given only their relative positions,
from which we are to construct a clock. Because of their interactions,
no framework exists in which the three stars move along straight
lines. The best we can achieve is some ‘spaghetti sculpture’ (Figure
13). This is found by telling a computer that there does exist a
framework of absolute space and time in which the stars obey
Newton’s laws. However, the computer is given only the successive
relative positions, not the positions in the framework at given times.
But this is real information, and if the computer is given a su cient
number of snapshots it can search for an arrangement of them in a
spaghetti sculpture in which the stars do obey Newton’s laws. The
positions in the framework and the separations in time are found by
trial and error.
 Suppose we are given ten snapshots. We can mark the positions of
the triangles formed by the three bodies in Triangle Land (Figures 3
and 4). We can then tell the computer four of the positions. If the
snapshots have indeed been generated by bodies that satisfy
Newton’s laws, the computer will nd a curve that passes through
them and the other six. We obtain a curve like those in Shape Space
in Figures 9 and 10. We have to use both representations, in Triangle
Land and in absolute space, because the raw data come to us in the
former, but it is in the latter that we can make sense of them. Once
we have solved the problem in absolute space, a timing of the
evolution has been established. It is that timing of the events for
which Newton’s laws do hold. If the computer tried to assign other
timings, they would not. The timing that does work can then be
transferred back to the raw data: the curve in Triangle Land. We can
make marks along it corresponding to the passage of the time found
by the computer.

The same thing can be done for any number of bodies. Their
relative con gurations will correspond to di erent points along a
curve in the corresponding Platonia. To lay out ‘marks of equal
intervals of time’ on it, we have to go through the same procedure
with the computer, telling it to nd a framework and a time in which
the bodies do satisfy Newton’s laws. Only two facts about this
process are signi cant. First, because all the bodies interact, all their
positions must be used if the ‘time marks’ are to be found. To tell the
time by such a clock, we need to know where all its bodies are. Time
cannot be deduced from a small number of them, unlike inertial
time; the clock has as many hands as the system has bodies. Second,
no matter how many bodies there are in a system, the data in just
two snapshots are never enough to nd the spaghetti sculpture in
absolute space and construct a clock. We always need at least some
data from a third snapshot. As we have seen, this ‘two-and-a-bit
puzzle’ is the main – indeed the only – evidence that absolute space
and not Platonia is the arena of the universe.

You might think that this is all far removed from practical
considerations. It is true that scientists have learned to make
extremely accurate clocks using atomic phenomena. But this is a
comparatively recent development. Before then, astronomers faced a
tricky situation, which is worth recounting.

For millennia, the Earth’s rotation provided a clock su ciently

reliable and accurate for all astronomical purposes. It was unique –
the astronomers had access to no other comparable clock. However,
about a hundred years ago, astronomical observations had become so
accurate that de ciencies in it began to show up. Tidal forces of the
Moon acting on the Earth sometimes give rise to unpredictable
changes of the mass distribution in its interior. As my accident in
Oxford demonstrated, such changes in a rotating body must change
its rotation rate. The clock was beginning to fail the astronomers’
growing needs for greater accuracy. Such crises highlight
fundamental facts. What could the astronomers do?

They managed to nd a natural clock more accurate than the

Earth: the solar system. To make this into a clock, they assumed that
Newton’s laws governed it. (After the discovery of general relativity,
small corrections had to be made to them, but this did not change
the basic idea.) However, the astronomers had no direct access to
any measure of time. Instead, they had to assume the existence of a
time measure for which the laws were true. Making this assumption
and using the laws, they could then deduce how all the dynamically
signi cant bodies in the solar system should behave. Although they
had no access to it, they then knew where the various bodies should
be at di erent instants of the assumed time. Monitoring one body –
the Moon, in fact – they could check when it reached positions
predicted in the assumed time and verify that the other bodies in the
solar system reached the positions predicted for them at the
corresponding times. The astronomers were thus forced into the
exercise just described, and they used the Moon as the hand of a
clock formed by the solar system.

They originally called the time de ned in this manner Newtonian

time. It is now called ephemeris time. (An ephemeris is a publication
which gives positions of celestial objects at given times.) For a
decade or more it was actually the o cial time standard for civil and
astronomical purposes. More recently, atomic time, which relies on
quantum e ects, has been adopted. There are several important
things about ephemeris time. First, it is unthinkable without the laws
that govern the solar system. Second, it is a property of the complete
solar system (because all its bodies interact, all co-determine one
another’s positions). Third, it exists only because the solar system is
well isolated as a dynamical system from the rest of the universe.
 Ephemeris time may be called the unique simpli er. This is an
important idea. If, as Mach argued, only con gurations exist and
there is no invisible substance of time, what is it that we call time?
When we hold the con gurations apart in time and put a duration
between them, this something we put there is a kind of imagined
space, a fourth dimension. The spacing is chosen so that the
happenings of the world unfold in accordance with simple laws
(Newton’s or Einstein’s). This is a consequence of the desire to
represent things in space and time, and our inability hitherto to nd
laws of a simple form in any other framework.

Ephemeris time is the only standard we can use if clocks are to

march in step. If we could not construct such clocks, we could never
keep appointments and clocks would be useless. To see that there is
only one sensible de nition of duration, imagine that two teams of
astronomers were sent to two similar but nevertheless di erent
isolated three-body stellar systems. All they can do is observe their
motions. From them they must generate time signals. Each team
works separately, but the signals they generate must march in step –
one clock may run faster than the other, but the relative rate must
stay constant. There is only one measure of duration they can
choose. In general, no motion in one system marches in step with
any motion in the other. Only ephemeris time, deduced from the
system as a whole, does the trick. A clock is any mechanical device
constructed so that it marches in step with ephemeris time, the
unique simpli er.
 We can now see that there is only one ultimate clock: the
universe. Although it would not be practicable, if we wanted to
obtain time of ultra-high accuracy from the solar system, we would,
sooner or later, have to take into account the disturbances exerted by
bodies farther away. Since there are no perfectly isolated systems
within the universe, this process can only stop, if ever, when the
entire universe has been made into a clock. The universe is its own
clock.

In the light of this, let us think again about Galileo’s ball rolling
across the table in Padua. Snapshots of the ball alone were not
su cient to tell what would happen when it rolled over the edge. It
seemed inconceivable that the ball’s path could be determined by the
little bit of water owing from a tank used to tell the time. For such
reasons as this, Newton rejected speed relative to any one motion as
a fundamental concept and invoked instead speed relative to an
abstract time. However, if we conceive the universe as a single
dynamical entity, the abstract time becomes redundant. The speed of
Galileo’s ball that determines which parabola it will trace is its speed
as measured by the totality of motions in the universe. This explains
why some motions are distinguished from others for timekeeping.
They are those that march in step with the cosmic clock, the unique
true measure of time. This time is the distillation of all change. High
noon is a con guration of the universe.

But the two-and-a-bit puzzle persists. We still have no simple

direct way to measure time in Platonia, we always have to go
through the intermediary of absolute space. This re ects the hybrid
nature of energy. Kinetic energy is de ned in absolute space,
whereas potential energy is determined by instantaneous
con gurations and is thus independent of Newton’s invisible
framework. We shall be able to claim that Platonia is the arena of the
world only if we can dispense with absolute space in the de nition of
kinetic energy. That is the next topic.
CHAPTER 7
Paths in Platonia

NATURE AND EXPLORATION

The two-and-a-bit puzzle is the statement that two snapshots of a

dynamical system are nearly but not quite su cient to predict its
entire history. We need to know not only two snapshots, but also
their separation in time and their relative orientation in absolute
space. These are exactly the things that determine the energy and
angular momentum of any system, both of which, as we have seen,
have a profound in uence on its behaviour.

There are two di erent ways to approach this problem. Either

we assume the known laws of nature are correct and simply ask how
they can be veri ed, or we take a more ambitious stance and ask if
they arise from some deeper level that we have not yet
comprehended. The latter is the approach of this chapter. We shall
forget absolute space and time and take Platonia for real. I have
likened it to a country; countries are there to be explored. In
exploring a country, one follows a path through it. Any continuous
curve through Platonia is such a path. A natural question is whether
some paths are distinguished compared with others. It leads directly
to the idea of optimization.

Optimization problems arise naturally, and were already well

known to mathematicians in antiquity. It seems they were also
known and understood by Queen Dido, who when she came to
North Africa was granted as much land as she could enclose within
the hide of a cow. She cut it into thin strips, out of which she made
a long string. Her task was then to enclose the maximum area of
land within it. The solution to this problem of maximizing the area
within a gure of given perimeter is a circle. However, Dido’s
territory was to adjoin the coast, which did not count as part of the
perimeter. For a straight coastline the solution to this problem is a
semicircle, and this was said to be the origin of the territory of
Carthage. A rich body of mathematical and physical theory has
developed out of similar problems. It cannot explain why the
universe is, but given that the universe does exist it goes a long way
to explain why it is as it is and not otherwise.

In early modern times, Pierre de Fermat (of the famous last

theorem) developed a particularly fruitful idea due to Hero of
Alexandria, who had sought the path of a light ray that passes from
one point to another and is re ected by a at surface on its way.
Hero solved this problem by assuming that light travels at a
constant speed and chooses the path that minimizes the travel time.
Fermat extended this least-time idea to refraction, when light passes
from one medium to another, in which it may not travel at the same
speed as in the rst medium. When a ray of light passes from air
into water, the ray is refracted (bent) downward, towards the
normal (perpendicular) to the surface. If this behaviour is to be
explained by the least-time idea, light must travel slower in water
than in air. For a long time it was not known if this were so, so
Fermat’s proposal was a prediction, which was eventually
con rmed.

In 1696 John Bernoulli posed the famous ‘brachistochrone’

(shortesttime) problem. A bead, starting from rest, slides without
friction under gravity on a curve joining two points at di erent
heights. The bead’s speed at any instant is determined by how far it
has descended. What is the form of the curve for which the time of
descent between the two points is shortest? Newton solved the
problem overnight, and submitted his solution anonymously, but
Bernoulli, recognizing the masterly solution, commented that
Newton was revealed ‘as is the lion by the claw print’. The solution
is the cycloid, the curve traced by a point on the rim of a rolling
wheel.

Soon there developed the idea that the laws of motion – and thus
the behaviour of the entire universe – could be explained by an
optimization principle. Leibniz, in particular, was impressed by
Fermat’s principle and was always looking for a reason why one
thing should happen rather than another. This was an application of
his principle of su cient reason: there must be a cause for every
e ect. Leibniz famously asked why, among all possible worlds, just
one should be realized. He suggested, rather loosely, that God – the
supremely rational being – could have no alternative but to create
the best among all possible worlds. For this he was satirized as Dr
Pangloss in Voltaire’s Candide. In fact, in his main philosophical
work, the Monadology, Leibniz makes the more defensible claim that
the actual world is distinguished from other possible worlds by
possessing ‘as much variety as possible, but with the greatest order
possible’. This, he says, would be the way to obtain ‘as much
perfection as possible’.

Inspired by such ideas, the French mathematician and

astronomer Pierre Maupertuis (another victim of Voltaire’s satire),
advanced the principle of least action (1744). From shaky initial
foundations (Maupertuis wanted to couple his idea with a proof for
the existence of God) this principle grew in the hands of the
mathematicians Leonhard Euler and Joseph-Louis de Lagrange into
one of the truly great principles of physics. As formulated by
Maupertuis, it expressed the idea that God achieved his aims with
the greatest economy possible – that is, with supreme skill. In
passing from one state at one time to another state at another time,
any mechanical system should minimize its action, a certain quantity
formed from the masses, speeds and separations of the bodies in the
system. The quantities obtained at each instant were to be summed
up for the course actually taken by the system between the two
speci ed states. Maupertuis claimed that the resulting total action
would be found to be the minimum possible compared with all
other conceivable ways in which the system could pass between the
two given states. The analogy with Fermat’s principle is obvious.

Unfortunately for Maupertuis’s theological aspirations, it was

soon shown that in some cases the action would not have the
smallest but greatest possible value. The claims for divine economy
were made to look foolish. However, the principle prospered and
was cast into its modern form by the Irish mathematician and
physicist William Rowan Hamilton a little under a hundred years
after Maupertuis’s original proposal. A wonderfully general
technique for handling all manner of mechanical problems on the
basis of such a principle had already been developed by Euler and,
above all, Lagrange, whose Mécanique analytique of 1788 became a
great landmark in dynamics.

The essence of the principle of least action is illustrated by

‘shortest’ paths on a smoothly curved surface. In any small region,
such a surface is e ectively at and the shortest connection between
any two neighbouring points is a straight line. However, over
extended areas there are no straight lines, only ‘straightest lines’, or
geodesics, as they are called. As the idea of shortest paths is easy to
grasp, let us consider how they can be found. Think of a smooth but
hilly landscape and choose two points on it. Then imagine joining
them by a smooth curve drawn on the surface. You can nd its
length by driving pegs into the ground with short intervals between
them, measuring the length of each interval and adding up all the
lengths. If the curve winds sharply, the intervals between the pegs
must be short in order to get an accurate length; and as the intervals
are made shorter and shorter, the measurement becomes more and
more accurate. The key to nding the shortest path is exploration.
Having found the length for one curve joining the chosen points,
you choose another and nd its length. In principle, you could
examine systematically all paths that could link the two chosen
points, and thus nd the shortest.

This is indeed exploration, and it contains the seed of rational

explanation. There is something appealing about Leibniz’s idea of
God surveying all possibilities and choosing the best. However, we
must be careful not to read too much into this. There does seem to
be a sense in which Nature at least surveys all possibilities, but what
is selected is subtler than shortest and more de nite than ‘best’,
which is di cult to de ne. Nothing much would be gained by going
into the mathematical details, and it will be su cient if you get the
idea that Nature explores all possibilities and selects something like
a shortest path. However, I do need to emphasize that Newton’s
invisible framework plays a vital role in the de nition of action.

Picture three particles in absolute space. At one instant they are

at points A, B, C (initial con guration), and at some other time they
are at points A*, B*, C* ( nal con guration). There are many
di erent ways in which the particles can pass between these
con gurations: they can go along di erent routes, and at di erent
speeds. The action is a quantity calculated at each instant from the
velocities and positions that the particles have in that instant.
Because the positions determine the potential energy, while the
velocities determine the kinetic energy, the action is related to both.
In fact, it is the di erence between them. It is this quantity that
plays a role like distance. We compare its values added up along all
di erent ways in which the system could get from its initial
con guration to its nal con guration, which are the analogues of
the initial and nal points in the landscape I asked you to imagine.
The history that is actually realized is one for which the action
calculated in this way is a minimum. As you see, absolute space and
time play an essential role in the principle of least action. It is the
origin of the two-and-a-bit puzzle. Now let us see how it might be
overcome.

DEVELOPING MACHIAN IDEAS

After it became clear to me that Platonia was the arena in which to

formulate Mach’s ideas, I soon realized that it was necessary to nd
some analogue of action that could be de ned using structure
already present in Platonia. With such an action it would be possible
to identify some paths in Platonia as being special and di erent
from other paths. In Leibniz’s language, such paths could be actual
histories of the universe, as opposed to merely possible ones. The
problem with Hamilton’s action was that it included additional
structure that was present if absolute space and time exist, but
absent if you insist on doing everything in Platonia. In 1971, with a
growing family and nancial commitments, I was doing so much
translating work I had little time for physics. As luck would have it,
the postal workers in Britain went on an extended strike. No more
work reached me (no one thought of using couriers in those days) –
it was bliss. I got down to the physics and soon had a rst idea. It
still took quite a time to develop it adequately, but eventually I
wrote it up in a paper published in Nature in 1974. Mach’s principle
may be controversial, but it always attracts interest, and Nature also
published quite a long editorial comment on the paper. Perhaps it
was worth waiting ten years before getting my rst paper published.

It was certainly a turning point in my life. Some months after it

appeared, I received a letter with some comments on it from Bruno
Bertotti, who was, and still is, a professor of physics at the
University of Pavia in Italy. Bruno, who is a very competent
mathematician, has worked in several elds in theoretical physics.
In fact, he was one of the last students of the famous Erwin
Schrödinger, the creator of wave mechanics (Box 1). But he has also
been active in experimental gravitational physics, and he organized
the rst two – and very successful – international conferences in the
eld. Although I can never stop thinking about basic issues in
physics, I am at best an indi erent mathematician, so I was very
lucky that my correspondence with Bruno soon developed into
active collaboration. Sometimes Bruno came to work at College
Farm, but mostly I went to Pavia. For seven years I went there for
about a month, every spring and autumn. It was a very fruitful and
rewarding collaboration: my work on Mach’s principle would never
have developed into a real theory without Bruno’s input. I cannot
say that we discovered any really new physics, because in the end
we had to recognize that Einstein had got there long before us.
What I think was important was that in two papers, published in
1977 and 1982, we laid the foundations of a genuine Machian
theory of the universe. To our surprise, we then found that this
theory is already present within general relativity, though so well
hidden that no one (not even Einstein) suspected it. We had found a
quite new route to his theory, and had the consolation to know that
Einstein had by no means fully grasped the signi cance of his own
theory.

In this connection, a remarkable coincidence that happened to

me on my rst visit to Pavia is worth recounting. I arrived on a
Friday night. I was going to spend the rst weekend sightseeing, and
after breakfast on Saturday morning I wandered o with no set aim
through the streets of Pavia in the warm April sunshine. After about
twenty minutes I chanced upon a grand medieval house. A plaque
outside said that in the 1820s the poet Ugo Foscolo had lived there.
One could walk into the courtyard, which I did. It was Italy as you
dream of it. This, I thought, was the place to live. Six months later,
quite by chance, I learned that for two years, in the 1890s, it had
been Einstein’s home. In his teens, the electrical rm run by his
father and uncle in Munich had failed, and they had moved to Pavia
and started another rm (which also failed). Somehow that chance
episode in Pavia seems symbolic of my e orts in physics. Einstein
was there rst, long ago, but it was still worth the journey to see the
place from the inside. It yielded a perspective, quite di erent from
Einstein’s, which persuades me that Platonia is the true arena of the
universe. If it is, we shall have to think about time di erently.

The rst idea Bruno and I developed had several interesting and
promising properties. Above all, it showed that a mechanics of the
complete universe containing only relative quantities and no extra
Newtonian framework could be constructed. Hitherto, most people
had thought this to be impossible. Just as Mach had suspected, the
phenomenon that Newton called inertial motion in absolute space
could be shown to arise from motion relative to all the masses in the
universe. We also showed that an external time is redundant.
However, besides the desirable features we obtained e ects which
showed that the theory could not be right. While the universe as a
whole could create the experimentally observed inertial e ects that
we wanted, the Galaxy would create additional e ects, not observed
by astronomers, that ruled out our approach.

The idea that Bruno and I rst developed seemed so natural it

surprised us that no one had thought of it earlier. However, I
learned quite recently that something similar was proposed in 1904
(in an obscure booklet by one Wenzel Hofmann), and then
rediscovered in 1914 by the physicist Hans Reissner and again in
1925 by none other than Schrödinger, just before he discovered
wave mechanics. This was especially ironic since Bruno had been
Schrödinger’s student. I think the main reason why these papers got
overlooked was that they were completely overshadowed by
Einstein’s general relativity and the excitement of the discovery of
quantum mechanics in 1925/6. There is also an undoubted tendency
for physicists to work within a so-called paradigm (the American
philosopher of science Thomas Kuhn’s famous expression), and pay
at best eeting attention to ideas that do not t within the existing
established patterns of thought.

I mention these things because the next idea that Bruno and I
tried seems to me just as natural as our rst idea, if one approaches
the problems of describing motion and change with an open mind. It
does, however, seem very di erent from the present paradigm,
which has become deep-rooted with the long hegemony of
Newtonian ideas, which were only partly changed by Einstein.
Although, as we shall see, our second idea is actually built into
Einstein’s theory at its very heart, within the context of classical
physics it merely provides a di erent perspective on that theory.
However, for the study of quantum e ects it does represent a
genuine alternative, and the attempt to create a quantum theory of
the universe may force its adoption, alien though it may appear to
many working scientists.

EXPLORING PLATONIA

Let me now explain this second idea. So far, I have explained only
what the points of Platonia are. Each is a possible relative
arrangement, a con guration, of all the matter in the universe. If
there are only three bodies in it, Platonia is Triangle Land, each
point of which is a triangle (Figures 3 and 4). Can we somehow say
‘how far apart’ any two similar but distinct triangles are? If so, this
will de ne a ‘distance’ between neighbouring points in Triangle
Land, and just as mathematicians seek geodesics using ordinary
distances on curved surfaces, we can start to look for geodesics in
Platonia. If we can nd them, they will be natural candidates for
actual histories of the universe, which we have identi ed as paths in
Platonia. They will be Machian histories if the ‘distance’ between
any two neighbouring points in Platonia is determined by their
structures and nothing else, and we shall not need to suppose that
they are embedded in some extra structure like absolute space.

There is such a simple and natural solution to the problem of

nding geodesics in Platonia that I would like to spell it out. The
fact that it does seem to be used by nature is one of the two prime
pieces of evidence I have for suggesting that the universe is timeless.
(The second, equally simple in its way, comes from quantum
mechanics.) How it works out for the simplest example of a universe
of three bodies is described in Box 8.

BOX 8 Intrinsic Di erence and Best Matching

In Figure 21, triangle ABC is one point in Triangle Land, and the
slightly di erent triangle A*B*C* is a neighbouring point. A
‘distance’ between them can be found in many ways, but one of the
simplest is the following. Imagine that ABC is held xed, and
A*B*C* is placed in any position relative to it. This creates
‘distances’ AA*, BB*, CC* between the corresponding vertices, at
which we suppose there are bodies of masses a, b, c. We form a ‘trial
distance’ d by taking each mass and multiplying it by the square of
the corresponding distance, adding the results and taking the square
root of the sum. Thus

This is an arbitrary quantity, since the relative positioning of the

two triangles is arbitrary. It is, however, possible to consider all
relative positionings and nd the one for which d is minimized. This
is a very natural quantity to nd, and it is not arbitrary. Two
di erent people setting out to nd it for the same two triangles
would always get the same result. It measures the intrinsic di erence
between the two matter distributions represented by the triangles. It
is completely determined by them, and does not rely on any
external structure like absolute space.

The intrinsic di erence between two arbitrary matter

distributions can be found similarly. One distribution is supposed
xed, and the other moved relative to it. In any trial position, the
analogue of the above expression is calculated, and the position in
which it is minimized is sought. Because this special position
reduces the apparent di erence between the two matter
con gurations to a minimum, it may be called the best-matching
position.
 Figure 21 A trial relative placing of the two triangles.

Using the intrinsic di erence de ned in Box 8, we can determine

‘shortest paths’ or ‘histories’ in Platonia as explained above.
However, the intrinsic di erence by itself does not lead to very
interesting histories, and it is more illuminating to consider a
related quantity. The potential energy of any matter distribution
(Figure 17) is determined by its relative con guration, and is
therefore already ‘Machian’. Each matter distribution has its own
Newtonian gravitational potential energy. Two nearly identical
matter distributions have almost the same potential. Now, the
intrinsic di erence is determined by two nearly identical
con gurations. To obtain more interesting histories we can simply
multiply the intrinsic di erences by the potential (strictly speaking,
by the square root of minus the potential). This will change the
de nition of ‘distance’, but it will still enable us to determine
‘shortest distances’. You do not need to worry about these details,
but I do want to give you a avour of what is involved.

I think you will agree that nding shortest paths in an imagined

timeless landscape bears little direct resemblance to our powerful
sense of the passage of time. Yet the outcome turned out to be
remarkably like what happens in Newtonian theory. Let me explain,
taking again the example of a three-body universe, for which
Platonia is Triangle Land.

Any continuous path in it corresponds to a sequence of triangles:

they are the ‘points’ through which the path passes. But this is very
similar to what comes out of Newtonian theory (Figure 1) – which,
however, yields not only the triangles but also their positions in
absolute space and separations in time. Remember that the triangles
in Figure 1 are ‘lit up’ by ashes at each unit of absolute time, and
that we see them, in perspective, in absolute space at those times.
However, these are invisible aids. The astronomers see neither when
they look through telescopes, all they see are stars. Thus, as far as
observable things are concerned, both theories yield the same kind
of thing – sequences of triangles. The question is, what kind of
sequences do the two theories yield? In what respects do the
theories di er when it comes down to what is actually observable?
 The major di erence is that the Machian theory makes more
de nite predictions. As a theory of geodesics, it determines the
shortest path between any two xed points in Platonia. It covers
Platonia with such paths. These geodesics have the following
important property: at any one point in Platonia, many of them pass
through it. In fact, for every direction that you can go from a point,
there is just one geodesic. This is the crucial di erence. As Figures
13 and 14 highlighted, when the Newtonian histories are
represented as paths in Platonia, it turns out that many can pass
through the same point, and have at that point the same direction.
However, these paths then ‘splay out’ and go to quite di erent
places in Platonia. In Newtonian terms, they di er in energy and
angular momentum. The di erence is not apparent in the initial
point and direction, but comes to light dramatically in the later
evolution of the paths. This defect is absent in the Machian theory.
For any given point and direction at that point, there is just one
geodesic. Bruno and I constructed the theory precisely to achieve
that aim.

What did quite surprise us was to discover that the unique

Machian history with a given direction through a point is identical
to one of the many Newtonian histories through the point with the
same direction. It is, in fact, the Newtonian history for which the
energy and angular momentum are both exactly zero. The small
fraction of Newtonian solutions with this property are all the
solutions of a simpler timeless and frameless theory.
 This brought to light an unexpected reconciliation between the
positions of Newton and Leibniz in their debate about absolute and
relative motion. Both were right! The point is that in a universe
which, like ours, contains many bodies, there can be innumerable
subsystems that are e ectively isolated from one another. This is
true of the solar system within the Galaxy, and also for many of the
galaxies scattered through the universe. Each subsystem, considered
by itself, can have nonzero energy and angular momentum.
However, if the universe is nite, the individual energies and
angular momenta of its subsystems can add up to zero. In a universe
governed by Newton’s laws this would be an implausible uke. But
if the universe is governed by the Machian law, it must be the case.
It is a direct consequence of the law. What is more, the Machian law
predicts that in a large universe all su ciently isolated systems will
behave exactly as Newton predicted. In particular, they can have
nonzero energy and angular momentum, and therefore seem to be
obeying Newton’s laws in absolute space and time. But what
Newton took to be an unalterable absolute framework is shown in
the Machian theory to be simply the e ect of the universe as a
whole and the one law that governs it. What physicists have long
regarded as laws of nature and the framework of space and time in
which they hold are, as I said in Chapter 1, both ‘local imprints’ of
that one law of the universe.

You can see directly how absolute space and time are created out
of timelessness. Take some point on one of the Machian geodesics in
Platonia; it is a con guration of masses. Take another point a little
way along the geodesic; it is a slightly di erent con guration.
Without any use of absolute space and time, using just the two
con gurations, you can bring the second into the position of best
matching relative to the rst. You can then take a third
con guration, a bit farther along the path, and bring it into its best
matching position relative to the second con guration. You can go
along the whole path in this way. The entire string of con gurations
is oriented in a de nite position relative to the rst con guration.
What looks like a framework is created, but it is not a pre-existing
framework into which the con gurations of the universe are slotted:
it is brought into being by matching the con gurations.
Nevertheless, we get something like the Newtonian picture in Figure
1, except that we do not as yet have the ‘spacings in time’.

But this too emerges from the Machian theory. In the equations
that describe how the objects move in the framework built up by
best matching, it is very convenient to measure how far each body
moves by making a comparison with a certain average of all the
bodies in the universe. The choice of the average is obvious, and
simpli es the equations dramatically. No other choice does the
trick. For this reason it needs a special name; I shall call it the
Machian distinguished simpli er. It is directly related to the quantities
used to determine the geodesic paths in Platonia. To nd how much
it changes as the universe passes from one con guration to another
slightly di erent one, it is necessary only to divide their intrinsic
di erence by the square root of minus the potential. (The action, by
contrast, is found by multiplying it by the same quantity.) When this
distinguished simpli er is used as ‘time’, it turns out that each
object in the universe moves in the Machian framework described
above exactly as Newton’s laws prescribe. Newton’s laws and his
framework both arise from a single law of the universe that does not
presuppose them.

In such a universe, the ultimate standard of time that determines

which curve is traced by Galileo’s ball when it falls o his table in
Padua is unambiguous. It is the average of all the changes in the
universe that de nes the Machian distinguished simpli er. Time is
change, nothing more, nothing less.

The di erence between the Newtonian and Machian theories can

be summarized as follows. If we do not know the energy and
angular momentum of a Newtonian system, we always need at least
three snapshots of its con gurations in order to reconstruct the
framework of space and time in which they obey Newton’s laws.
The task is complicated, to say the least. If, however, the system is
Machian, the framework can be found with just two snapshots and
the task is vastly simpler. It simply requires best matching of the
two con gurations.

When, later, I suggest that the quantum universe is timeless in a

deeper sense than the classical Machian universe just described, that
will be a conjecture. But it is made plausible by the results of this
chapter. They are not speculation but mathematical truths. Every
phenomenon explained by Newton’s laws, including the beautiful
rings of Saturn and the spectacular structure of spiral galaxies, can
be explained without absolute space and time. They follow from a
simpler, timeless theory in Platonia.
PART 3
The Deep Structure of General Relativity

Now we come to relativity. My aim is not to give an extended

account, only to show how its fundamental features relate to the
book’s theme. But I have a tough nut to crack. My subject is the
non-existence of time, whereas time is almost everything in
relativity as it is usually presented. Is relativity Hamlet without the
Prince of Denmark?

In fact, the evidence for the non-existence of time in relativity

has long been hidden by accidents of historical development, and is
far stronger than many people realize. Yet the case is not quite
conclusive. We have seen how the space and time of Newton’s
theory can be constructed from instants of time as de ned in this
book. Taking them to be the true atoms of existence, we have shown
that no external framework is needed. Einstein’s space-time can also
be put together from instants in a strikingly similar way. However,
in the nished product they are knit together far more tightly than
in Newtonian theory. Explaining the wonderful way in which this
happens is the goal now. If the world were classical, no one would
try to pull space-time apart into instants. But quantum theory will
probably shatter space-time. It is therefore sensible to consider the
constituents into which it might shatter. This is what I shall do in
Part 3.

I begin by looking at the special theory of relativity, in which

gravity plays no role. I then go on to the general theory, in which
Einstein found a most brilliantly original way to describe gravity. In
both relativity theories time seems to be very real and to behave in
ba ing ways. But, as became clear only after Einstein’s death, his
theory has a deep structure which is revealed only by an analysis of
how it works as a dynamical theory. It is this deep structure that is
timeless. Quite a large proportion of Part 3 will explain the purely
historical accidents that obscured the deep structure of general
relativity for so long.
CHAPTER 8
The Bolt from the Blue

HISTORICAL ACCIDENTS

In the whole of physics, nothing is more remarkable than the

transformation wrought by a simple question that Einstein posed in
1905: what is the basis for saying that two events are simultaneous?
Einstein was not the rst to ask it. James Thomson, brother of Lord
Kelvin, had in 1883. More signi cantly, so had Poincaré – a great
gure in science – in 1898, in a paper that Abraham Pais, Einstein’s
biographer, calls ‘utterly remarkable’. In connection with historical
accidents, Poincaré’s paper is very interesting. He identi ed two
problems in the de nition of time.

First he considered duration: what does it mean to say that a

second today is the same as a second tomorrow? He noted that this
question had recently been widely discussed, and he outlined the
astronomers’ solution, the ephemeris time described in Chapter 6.
However, he then noted a second question, just as fundamental and
in some ways more immediate, which had escaped close attention.
How does one de ne simultaneity at spatially separated points? This
was the question that Einstein posed and answered seven years later
with such devastating e ect. I read the subsequent history of
relativity as follows. Einstein answered his question – Poincaré’s
second – with such aplomb and originality that it eclipsed interest in
the question of duration. It is not that duration plays no role in
relativity – quite the opposite, it plays a central role. But duration is
not derived from rst principles. It appears indirectly.

One of the main aims of Part 3 is to redress the balance, to treat

duration at the same level as simultaneity. There is, in fact, a
beautiful theory of duration at the heart of general relativity, but it is
hidden away in sophisticated mathematics. Einstein had no inkling of
this. He said of his own theory that no one who had grasped its
content could ‘escape its magic’. But the magic was more potent than
even he realized. It can, it almost certainly will, destroy time.

BACKGROUND TO THE CRISIS

Much of nineteenth-century physics can be seen as meticulous

preparation for the denouement over simultaneity. It had to come,
but what a coup de théâtre Einstein made of it. Many readers will be
familiar with the story, but since it introduces important ideas I shall
brie y recall some key elements. It all started with an investigation
of interference carried out in 1802 by the English polymath Thomas
Young, famous among other things for his decipherment of the
Egyptian hieroglyphs on the Rosetta Stone. In a sense, this was the
start of both relativity and quantum theory. Young observed that if
light from a single source is split into two beams that are
subsequently recombined and projected onto a screen, then bright
and dark fringes appear. He interpreted them in terms of a wave
theory of light. If light is some kind of wave motion, there will be
wave crests and troughs in both beams. When they are recombined,
there will be places where the crests from one beam coincide with
troughs in the other. They will cancel, giving dark fringes. But where
crests coincide, they will enhance each other, giving bright fringes
(Figure 22). Innumerable natural phenomena are explained by
interference.

Young’s insight, which was developed more or less independently

and much more thoroughly some years later by the Frenchman
Augustin Jean Fresnel, soon gave rise to the notion that light waves
must be vibrations of some elastic medium, which was called the
aether. Meanwhile, the study of electricity and magnetism developed
rapidly. In 1831, the English scientist Michael Faraday discovered
electromagnetic induction, which not only showed that electricity
and magnetism were related phenomena but rapidly became the
basis of all electrical machinery. Deeply impressed by the patterns
formed by iron lings sprinkled on paper held near a magnet (Figure
23), Faraday introduced the notion of lines of force and elds. A eld
can be thought of as a tension or excitation that exists throughout
space and varies continuously (as demonstrated by induction) in
both space and time. The eld concept eventually changed physicists’
picture of what the world is ‘made of.
 Figure 22 Thomas Young’s original explanation of the interference
fringes in accordance with the wave theory of light, which he
deduced by analogy with the behaviour of water waves. According to
this interpretation, the beam reaches the barrier in the form of a
plane wave, the successive parallel crests of which arrive
simultaneously at the two slits A and B. The wave is di racted at
each slit, and spherical waves spread out from each point of the two
slits towards the screen. At some points on the screen, the wave
crests (or troughs) from the two slits arrive simultaneously, and the
wave intensity is enhanced (bright regions). At other points, a wave
crest from one slit arrives with a wave trough from the other. The
wave intensity is cancelled at such a point (dark regions). This is the
classical explanation of the fringes in terms of interference.

In the decade from 1855, the Scottish physicist James Clerk

Maxwell took up Faraday’s qualitative eld notion and cast it into
mathematical form. His equations showed that electromagnetic
e ects should propagate through empty space as waves with a speed
determined by the ratio of certain constants. It had already been
noted that the ratio was equal to the known speed of light, leading to
the strong suspicion that light was an electromagnetic e ect.
Maxwell’s equations proved this. Electromagnetic e ects can
propagate as waves of many di erent wavelengths: from radio waves
(with wavelengths of around a metre to a kilometre), microwaves
(wavelengths measured in centimetres), infrared waves (some ten to
a thousand waves per centimetre), visible light (roughly ten
thousand waves to the centimetre), ultraviolet light (up to around a
million waves per centimetre), X-rays (of the order of ten million
waves in every centimetre) and gamma rays (billions or even trillions
of waves per centimetre). Hertz’s celebrated detection of waves from
an electromagnetic source in 1888 was the rst con rmation of this
consequence of Maxwell’s theory.
Figure 23 Magnetic lines of force as revealed by placing iron lings
in the magnetic eld of a bar magnet.

Virtually all physicists were convinced that these electromagnetic

excitations must be carried by some mechanical aether. This put a
remarkable twist into the theory of motion and the status of
Newton’s absolute space. Even in the framework of Newtonian
theory, there had always been one serious problem with the notion.
Newton was not entirely frank about it. In his guts, he certainly
believed in a state of absolute rest. When he introduced absolute
space, his words suggested the existence of one unique framework of
motion. Either you moved with respect to it or you did not.
However, in the body of the Principia he stated and used correctly
the relativity principle, according to which the motions within a
system are completely una ected by any uniform overall motion it
has (Box 5). This seriously undermined the idea of a unique state of
rest – no criterion could establish whether one were in it or moving
uniformly.

Figure 24 The pond argument. According to the spectator standing

on the bank, the ripples move to the left and right with equal speed.
But her partner, walking along the bank, sees things di erently. He
can almost keep up with the waves going to the right, while the left-
moving waves recede from him almost twice as fast.

It was soon seen that the aether should introduce an

experimentally veri able standard of rest. The argument is simple
and seems irrefutable (Figure 24). If you throw a stone onto the still
surface of a pond, waves spread out in concentric rings. The water
molecules do not move with the wave, they merely go up and down.
The water plays the role of the conjectured aether: itself at rest, it is
the material carrier of the waves. As seen by the woman standing on
the bank, the waves spread out in all directions with the same speed.
But for her partner walking along the bank, the wave process unfolds
di erently. The waves moving in the same direction as him have a
di erent velocity relative to him compared with the waves moving in
the opposite direction. A fast walker will even overtake some of the
waves. The relativity principle cannot hold for such processes, and it
was therefore expected that it would hold only for the mechanical
processes described by Newton’s laws, but not for optical and
electromagnetic phenomena. Moreover, as the Earth revolves in its
orbit around the Sun it must be continually changing its speed
through the aether. This ought to result in observable e ects.

In fact, the argument is not quite so simple. Everyone agreed that

light must be carried by an aether, but were all parts of the aether at
rest relative to one another? As the Earth orbits the Sun, might it not
carry some aether with it? It was also necessary to consider what
would happen to light passing through water owing on the Earth.
Would the aether be carried along, partially or completely, by the
water relative to the Earth? In fact, many issues had to be
considered, including aberration, which makes the stars seem to shift
slightly towards the point in the sky towards which the Earth is
moving at any instant as it orbits the Sun. The arguments, some of
which predated Maxwell’s work, developed over a period of eighty
years, and many important experiments were made. By 1895, when
the Dutch physicist Hendrik Anton Lorentz published an in uential
study, a consensus had more or less developed. It was that all known
experimental results, with one crucial exception, could be explained
by assuming the existence of a perfectly rigid aether.

The aether as proposed by Lorentz was actually devoid of all

physical properties except rigidity. It was simply there to carry the
excitations of Maxwell’s electromagnetic eld and, in Lorentz’s
words, to be the framework ‘relative to which all observable motions
of the celestial bodies take place’. It therefore supplied a standard of
rest like the water in the pond.

The one exception with which Lorentz had to contend was the
famous Michelson-Morley experiment performed with great accuracy
in 1887. Based on interference between light beams moving in the
direction of the Earth’s motion and at right angles to it, it was
designed to measure the change in the speed of the Earth’s motion
through the aether over the course of a year. Its accuracy was
su cient to detect even only one-hundredth of the expected e ect.
But nothing was observed. It was a great surprise, and very puzzling.

Lorentz’s response was piecemeal. In particular, he suggested

that, for some physical reason, the length of a body moving relative
to the aether could be reduced in the direction of its motion by the
amount needed to explain the Michelson-Morley result. Some years
earlier, the Irish physicist George Fitzgerald had made the same
proposal. Poincaré responded that some general principle should rule
out all possibility of detecting motion relative to the aether. It should
not be necessary to invoke ad hoc hypotheses. He began to think that
the relativity principle might hold universally and not just for
mechanical phenomena. Both he and Lorentz were working in this
direction when Einstein appeared on the scene with a stunning
solution.

EINSTEIN AND SIMULTANEITY

Two aspects of Einstein’s work ensured its triumphant success. First,

he took the relativity principle utterly seriously. It was the bedrock,
repeatedly exploited. Second, he took for real a ‘local time’ that
Lorentz had introduced as a formal device to describe phenomena in
a reference frame moving relative to the aether. Events simultaneous
in the ‘local time’ were not so in the real time of the aether frame.
But Einstein, committed to relativity, regarded one as just as real as
the other. He made a virtue out of an apparent vice, and saw that the
key to the entire mystery lay in the concept of simultaneity.

He deliberately highlighted an apparently irreconcilable paradox

and then deftly presented its unique resolution: a radical proposal for
saying when events are simultaneous. Hitherto this had seemed
obvious, but Einstein showed that simultaneity was not a property of
the world but a re ection of the way we describe it. By showing that
the paradox could be resolved only by changing the notion of
simultaneity – and with it time – he brought this issue to the fore.

The paradox was carefully prepared. He rst de ned the

relativity principle. As in mechanics (Box 5), he postulated
distinguished frames of reference in which all the laws of nature take
their simplest form, and required this form to be the same in each
frame. Any such frame, which constituted a kind of grid in space and
time, should be in a state of uniform rectilinear (i.e. straight) motion
relative to any other. He then postulated, in addition to this general
principle, just one actual law of nature: that light propagates with
the same speed c in all directions irrespective of the speed of the
source. This was exactly what everybody had always assumed would
hold in the unique frame of reference at rest in the aether. Einstein
insisted it should happen in all frames.

The pond argument suggests that this is absurd. But Einstein

realized that he had a hitherto unrecognized freedom: the grid lines
de ning simultaneity in space and time could be ‘drawn’ in a novel
way. Simultaneity at spatially separated points must be de ned in
some way – but how? There must be a physical transmission of
signals so as to synchronize clocks. The ideal would be in nitely fast
signals. Then there would be no argument. This is e ectively what
happens in the pond experiment – the man and woman observe the
water waves by light, which travels nearly a billion times faster than
they do. We now see that the analogy between water waves in ponds
and light waves in the aether is not perfect. For a full analogy, there
would have to be signals that travel faster than light itself.

But such signals were unknown in Einstein’s time, and his theory
would show that they could not exist. He therefore used the best
substitute – light. This completely changed things. Light was to be
analysed in a framework that light itself created, so the problem
became self-referential. It might seem that Einstein cheated, making
up the rules as he went along to ensure that he won the game.
However, he was simply confronting a fact of life: laws of nature will
be meaningful only if they relate things that can actually be
observed. We live inside, not outside, the universe, and to
synchronize distant clocks we have no alternative to the physical
means available to us. Einstein’s hunch that we should use light
because it would turn out to be the fastest medium available in
nature has so far been totally vindicated.

The magical touch was that his choice was the most natural thing
to do – in the theory of an aether and in the context of the relativity
principle. Given their apparent irreconcilability, his subsequent
demonstration of their compatibility was a coup. It also showed that
there was something inevitable about the result.

For suppose there is a pond-like aether and that nothing is faster

than light. It is natural to assume that it travels equally fast in all
directions. Then how are we to de ne simultaneity throughout the
aether? Einstein proposed setting up a master clock at a central
reference point, sending a light signal to some distant identical clock
at rest relative to it, and letting the signal be re ected back to the
master clock. If it measures a time T for the round trip, we would
obviously say that the light took ½T to reach the distant clock, which
can be synchronized to read t + ½T on the arrival of a signal sent by
the master clock when it reads t. In this way, clocks throughout the
aether can be synchronized with the master clock. Standard
measuring rods can be used to measure the distances between them.
This is the obvious way to set up a space-time grid if the aether
theory is correct.

However, it does assume that the aether is ‘visible’ and that we

know when we are at rest in it. But this the relativity principle
denies. Imagine a family of observers, equipped with clocks to
measure time and rods to measure length, distributed in space and at
rest relative to one another. Believing themselves at rest in the
aether, they de ne simultaneity by Einstein’s prescription. There is
also a second family, with identical rods and clocks, also at rest
relative to one another but moving uniformly relative to the rst. By
the relativity principle, they can equally believe themselves at rest in
the aether. So they too will use Einstein’s prescription to de ne
simultaneity. Just as belief in the aether theory makes the
prescription natural, belief in the relativity principle makes it natural
for both families to adopt it. Nothing in nature privileges one family
over the other. Whatever one family does, the other can do with
equal right. In particular, each can use Einstein’s prescription.

The inescapable consequence is that the two families will

disagree about which events are simultaneous. However, by
accepting this, Einstein achieved his rst goal – the demonstration
that light propagation takes an identical form for both families (Box
9). This remarkable fact – that the relativity principle holds for light
propagation and that simultaneity depends on the observer and on
convention – is thus the great denouement towards which so much
wonderful physics in the nineteenth century had been tending. It also
showed that the aether is a redundant concept, since no experiment
can establish whether we are moving relative to it.

Lack of simultaneity was only the beginning. Einstein went on to

draw further amazing consequences from his iron insistence that all
phenomena must unfold in exactly the same way for any two
families of observers in uniform motion relative to each other. In
particular, he was able to make some startling predictions about rods
and clocks. The point is that the facts of light propagation are
established by means of physical rods and clocks, but these tools are
not immune to the relativity principle. Using simple equations and
precise arguments, Einstein showed that two such families must each
come to the conclusion that the clocks of the other family, moving
relative to them, run slower than their own clocks. Each family also
concludes that the rods of the other family are shorter than their
own.

What is so remarkable about these results – and it seems so

impossible that many quite intelligent people still refuse to accept it
– is their mutual nature. How can it be that each family nds that
the clocks of the other family run slower than their own? Box 10
explains.

BOX 9 Relativity in One Diagram and 211 Words

 Figure 25 Alice (A) and Bob (B) believe they are at rest in the
aether, and therefore draw a grid (dashes) with time vertical and
space (only one dimension shown) horizontal. To synchronize their
clocks, Alice sends Bob a light signal (solid line), which reaches him
at X, where it is re ected back to Alice at T. Alice concludes that the
signal reached Bob when she was at H. Their identical twins Alice*
(A*) and Bob* (B*) are moving uniformly relative to them, but also
believe they are at rest. Alice* sends a light signal, just as her twin
does, at the moment they meet. It reaches Bob* at X*, and the
re ection of it returns to Alice* at T*. She therefore concludes that
her signal reached Bob* at H*, so she and Bob* have a grid (dots)
inclined relative to their twins’ grid. The pairs do not agree on which
events are simultaneous. Alice and Bob think H and X are, their twins
think H* and X* are. However, they con rm the relativity principle,
since both nd that light travels along rays parallel to the diagonals
(AX, XT and A*X*, X*T*) of their respective coordinate grids. Despite
appearances, the situation is symmetric – in Alice* and Bob*’s grid
their twins’ grid appears skew.

THE FORGOTTEN ASPECTS OF TIME

Fascinating as the results of Einstein’s two relativity theories are,

many of them are not directly relevant to my main theme. Popular
accounts that cover topics I omit are recommended in the section on
Further Reading. My aim in Part 3 is to show how Einstein’s
approach to relativity led him to an explicit theory of simultaneity
but an implicit theory of duration. It is the latter that is important for
this book, but it never got properly treated in relativity. The point is
that remarkable facts about duration, as revealed through the clocks
of di erent observers, follow inescapably from the de nition of
simultaneity and the relativity principle. Einstein did not need to
create a theory of clocks and duration from rst principles in order
to learn some facts about them: they already followed from his two
primary postulates.

BOX 10 The Impossible Becomes Possible

Figure 26 The horizontal and inclined strips, in which time increases
vertically and the horizontal represents one space dimension, show
the histories of two physically identical rods moving uniformly
relative to each other. For Bob and Alice, points on the continuous
line PPQQ are simultaneous and show the positions and lengths of
the rods at the corresponding times. Their rod, PP, appears longer
than Bob* and Alice*’s rod, QQ. But the starred twins think that
points on the line P’P’Q’Q’ are simultaneous, and conclude that their
rod Q’Q’ is longer than P’P’. A similar illustration could be given for
clocks. Such diagrams do not explain this behaviour of rods and
clocks, but do show that there is no outright logical contradiction.
Einstein’s conclusions are as secure as his premises. His con dence in
them has so far been totally vindicated.

The method Einstein used to create his relativity theories is an

important factor. During the nineteenth century, mainly through the
development of thermodynamics, physicists began to distinguish
between, on the one hand, theories of the world in terms of truly
basic laws and constituents (e.g. atoms and elds) and, on the other
hand, so-called principle theories. In the latter, no attempt would be
made to give an ultimate theory of things. Instead, the idea was to
seek principles that seemed to hold with great generality and include
them in the foundations of the description of phenomena. The
repeated failures of all attempts to build perpetual-motion machines,
of which two distinct types could be envisaged, became the basis of
the rst and second laws of thermodynamics. In the form in which it
was developed on this basis, thermodynamics was a theory of the
second kind – a principle theory.

In contrast, Lorentz’s combined theory of the electromagnetic

eld, electric charges and the aether was basically a theory of the
rst kind – it aspired to a fundamental description of the world in
terms of its ultimate constituents. Einstein deliberately decided not
to follow such a path in his own work on electrodynamics, from
which the special theory of relativity emerged. He based it as far as
possible on general principles. The fact is that Max Planck’s quantum
discoveries (Box 1) and Einstein’s own development of them a few
months before the relativity paper had persuaded Einstein that
something very strange was afoot. Despite his admiration for
Maxwell’s equations, he felt sure that they could not be the true laws
of electromagnetism because they completely failed to explain the
quantum e ects. He had no con dence in his ability to nd correct
alternatives. Then, and to the end of his days, Einstein found the
quantum ba ing. He felt deeply that it was a huge mystery. By
comparison, relativity (the special theory at least) was almost child’s
play.

It was this attitude that largely shaped Einstein’s strategy in

approaching the problem of the electromagnetic aether. He resolved
to make no attempt at a detailed description of microscopic
phenomena. Instead, he would rely on the relativity principle, for
which there seemed to be strong experimental support, and make as
few additional assumptions as possible. In the event, he was able to
limit these to his assumption about the nature of light propagation.
This was the one part of Maxwell’s scheme that he felt reasonably
con dent would survive the quantum revolution.

This had important consequences for the theory of time.

Poincaré’s 1898 paper showed that it must answer two main
questions: how simultaneity is to be de ned, and what duration is.
Associated with the second question is another, almost as important:
what is a clock? Because of his approach, Einstein answered only the
rst question at a fundamental level. He gave at best only partial
answers to the other two, and gave no explicit theories of either rods
or clocks. Instead, he tacitly assumed the minimal properties they
should possess. Otherwise, he relied to a very great extent on the
relativity principle. It took him far. Few things in physics are more
beautiful than the way he postulated the universal relativity
principle and the one particular law of light propagation, and then
deduced, from their combination, extraordinary properties of rods,
clocks and time. If the premises were true, rods and clocks had to
behave that way.

During his protracted creation of general relativity, Einstein used

this trick several times. The strategy was always to avoid speci c
assumptions, and instead to seek principles. In this way he avoided
ever having to address the physical working of rods and clocks: they
were always treated separately as independent entities in both
relativity theories. Their properties were not deduced from the inner
structure of the theory, but were simply required to accord with the
relativity principle. Einstein was well aware that this was ultimately
unsatisfactory, and said so in a lecture delivered in 1923. He made
similar comments again in 1948 in his Autobiographical Notes.

However, the tone of his comments does not suggest that he

expected any great insight to spring from the recti cation of this ‘sin’
(Einstein’s own expression). Only a ‘tidying up’ operation was
needed. This gap in the theory of duration and clocks has still not
been lled. I know of no study that addresses the question of what a
clock is (and how crucially it depends on the determination of an
inertial frame of reference) at the level of insight achieved in non-
relativistic physics by James Thomson, Tait and Poincaré.
Throughout relativity, both in its original, classical form and in the
attempts to create a quantum form of it (which we come to in Part
5), clocks play a vital role, yet nobody really asks what they are. A
distinguished relativist told me once that a clock is ‘a device that the
National Bureau of Standards con rms keeps time to a good
accuracy’. I felt that, as the theorist, he should be telling them, not
the other way round.

The truth is that a chapter of physics somehow never got written.

Despite his great admiration for Mach, Einstein was curiously
insensitive to the issues highlighted by Mach and Poincaré. He did
not directly address the nature and origin of the framework of
dynamics. Despite an extensive search through his published papers
and published and unpublished correspondence, I have found no
indication that he ever thought really seriously about issues like
those raised by Tait’s problem. This is rather surprising, since these
were ‘hot topics’ during the very period in which Einstein created
special relativity. He did not ask how the spatiotemporal framework
(i.e. the framework of space and time used by physicists) arose;
instead, he described the nished product and the processes that take
place within the arena it creates.

In fact, Einstein and Hermann Minkowski, whose work will

shortly be considered, brought about a marked change of emphasis
in physics. To use an expression of John Wheeler, the ‘royal high
road of physics’ from Galileo until Einstein was dynamics. Maxwell
saw his own work as an extension of the principles developed by
Galileo and Newton to new phenomena and to the eld notion
introduced by Faraday. At the same time, other scientists like Carl
Neumann and Mach became aware of the need for new foundations
of dynamics. In Poincaré’s writings of around 1900, one can see clear
hints of how dynamics might have been developed further as the
main stream of research. In particular, an explicit theory of the
origin of the spatiotemporal framework might have emerged. That is
more than evident from Poincaré’s 1898 paper on time and his 1902
comments, discussed in Chapter 5.

All this was changed by Einstein’s 1905 paper. Because of his

quantum doubts, he distrusted explicit dynamical models. Within a
few years a dualistic scheme appeared. Newton’s absolute space and
time were replaced by space-time, but this was not the complete
story. Actual physics emerged only with the statements about how
rods and clocks behaved in space-time. This is where the scheme was
dualistic. The behaviour of rods and clocks – and with it a theory of
duration – never emerged organically from the structure of space-
time, it was simply postulated. This is not to say the dualistic scheme
is wrong in the statements it makes. Einstein’s theory is as secure as
its foundations; there is no hint of failing there. However, insight
into the nature of time and duration was lost.

For all that, general relativity does contain, hidden away in its
mathematics (as I have already indicated), a theory of duration and
the spatiotemporal framework. However, this did not come to light
for many decades and even now is not properly appreciated. How
this came about, and an account of the ‘hidden dynamical core’ of
general relativity, are the subject of the next chapters.

It may help to end this chapter with a general remark on time. It

is impossible to understand relativity if one thinks that time passes
independently of the world. We come to that view only because
change is so all-pervasive and so many di erent changes all seem to
march in perfect step. Relativity is not about an abstract concept of
time at all: it is about physical devices called clocks. Once we grasp
that, many di culties fall away. If light did not travel so much faster
than normal objects, we would observe relativistic e ects directly
and they would not strike us as strange. There is nothing inherently
implausible in the idea that clocks travelling past us at high speed
should be observed to go slower than the watch on our wrist. Motion
of the clock might well alter the rate at which it ticks. After all, when
we swim through water, we feel the way our body responds. If there
were an aether, clocks could well be a ected by their motion
through it. What is di cult to grasp is how observers travelling with
the moving clocks think our wristwatch is running slow, while we
think just the same about their clocks (this apparent logical
impossibility has been dealt with in Box 10). However, the important
thing is to get away from the idea that time is something. Time does
not exist. All that exists are things that change. What we call time is
– in classical physics at least – simply a complex of rules that govern
the change.
 9
CHAPTER
Minkowski the Magician

THE NEW ARENA

Hermann Minkowski’s ideas have penetrated deep into the psyche of

modern physicists. They nd it hard to contemplate any alternative
to his grand vision, presented in a famous lecture at Cologne on 21
September 1908. Its opening words, a magical incantation if ever
there was one, are etched on their souls:

The views of space and time which I wish to lay before you
have sprung from the soil of experimental physics, and
therein lies their strength. They are radical. Henceforth
space by itself, and time by itself, are doomed to fade away
into mere shadows, and only a kind of union of the two will
preserve an independent reality.

The branch of knowledge that considers what exists is ontology.

These three sentences changed the ontology of the world – for
physicists at least.

For most physicists in the nineteenth century, space was the

most fundamental thing. It persisted in time and constituted the
deepest level in ontology. Space, in turn, was made up of points.
They were the ground of being, conceived as identical, in nitesimal
grains of sand close-packed in a block. Space was like glass. It was,
of course, three-dimensional. However, alerted by Einstein’s work to
how the relativity principle mixed up space and time, Minkowski
commented that ‘Nobody has ever noticed a place except at a time,
or a time except at a place.’ He had the idea that space and time
belonged together in a far deeper sense than anyone had hitherto
suspected. He fused them into space-time and called the points of
this four-dimensional entity events. They became the new ground of
being.

Such atoms of existence – the constituent events of space-time –

are very di erent from the entities that I suggested in Part 2 as the
true atoms of existence. The main aim of Part 3 is to show that
space-time can be conceived of in two ways – as a collection of
events, but also as an assemblage of extended con gurations put
together by the principle of best matching and the introduction of a
‘time spacing’ through a distinguished simpli er, as explained for
the Newtonian case at the end of Chapter 7. However, re ecting the
relativity of simultaneity, the assemblage has an additional
remarkable property that gives rise to the main dilemma we face in
trying to establish the true nature of time.

FROM THREE TO FOUR DIMENSIONS

In itself, the fusion of space and time was not such a radical step. It
can be done for Newtonian space and time. To picture this, we must
suppose that ordinary space has only two dimensions and not three.
We can then imagine space as a blank card, and the bodies in space
as marks on it. Any relative arrangement of these marks de nes an
instant of time.

The solution of Tait’s problem showed how relative

con gurations can, if their bodies obey Newton’s laws, be placed in
absolute space at their positions at corresponding absolute times. If
space is pictured as two-dimensional, absolute space is modelled not
by a room but by a at surface. The solution of Tait’s problem
places each card on the surface in positions determined by the
marks on the cards. In these positions, in which the centre of mass
can be xed at one point, any body moving inertially moves along a
straight line on the surface.

Keeping all the cards horizontal (parallel to the surface), we can

put a vertical spacing between them which is proportional to the
amount of absolute time between them. This is like imagining the
amount of time between 11 o’clock and 12 o’clock as a distance, and
is a very convenient way of visualizing things. The resulting
structure can be called Newtonian space-time. The one dimension of
time has been put together with the two of space. Newton’s laws can
be expressed very beautifully in this three-dimensional structure,
which is a kind of block. Whatever motion a body has, it must
follow some path in this block. Minkowski called this path its world
line. If the body does not move in space, which is a special case of
inertial motion, its world line is vertically upwards. If it is moving
inertially with some velocity, then it has a straight world line which
is inclined to the vertical. The faster the motion, the larger the angle
with the vertical.

In reality, ordinary space has three dimensions and Newtonian

space-time four. Instead of cards placed at vertical positions
representing di erent times, or simultaneity levels, we must imagine
three-dimensional spaces fused into a four-dimensional block. This
is impossible to visualize, but the model with only two space
dimensions is a good substitute.

Newtonian space-time di ers in an important respect from space,

in which all directions are on an equal footing and none is
distinguished from any other. In Newtonian space-time, one
direction is singled out. This is re ected in its representation as a
pack of cards. Directions that lie in a card, in a simultaneity level,
are quite di erent from the time line that runs vertically through
the cards. Newtonian space-time is ‘laminated’. If you were to ‘cut
through it’ at an angle, the ‘lamination’ would be revealed. You
would be ‘cutting through’ the simultaneity levels. The
inequivalence of directions can be expressed in the language of
coordinates.

Just as you can put a coordinate grid on a two-dimensional map,

you can ‘paint’ a rectangular grid on Newtonian space-time with one
of the axes perpendicular (parallel to the time line). The laws of
motion can be formulated in terms of the grid. For example, bodies
moving inertially travel along lines that are straight relative to the
grid. You can then ‘move’ the grid around as a complete unit into
di erent positions in space-time and see if the motions relative to
the new grid satisfy the same laws as they did in the old. For
Newton’s laws there is considerable but not complete freedom to
move the grid. Provided it is maintained vertical, it can be shifted
and rotated in ordinary space, just like a child’s climbing frame, and
it can also be raised and lowered in the vertical time direction.
However, tilting the vertical axis is not allowed. Newtonian forces
(in gravity and electrostatics, for example) are transmitted
instantaneously – horizontally in the model. If you tilt the grid from
the original time axis, you leave the old simultaneity levels. The
forces are not transmitted through the new levels.

Minkowski’s real discovery was that, in an analogous

construction using Maxwell’s electromagnetic equations instead of
Newton’s laws, the resulting space-time structure, now called
Minkowski space-time, has no special ‘lamination’. It is more like a
loaf of bread, through which you can slice in any way. The cut
surface always looks the same. The way this shows up in changes of
the coordinate grid is especially striking. Time becomes very like
space but not quite identical.

The di erence can be illustrated by the climbing frame. Here too

a vertically held frame can be shifted, rotated and raised or lowered
as a rigid unit. Maxwell’s laws still take the same form with respect
to the displaced grid. But you can also tilt it from the vertical
provided you do something else as well. For this, you need an
‘articulated’ grid, which we have in fact already encountered, in
Figure 25 in the discussion of simultaneity. It is a typical example of
the space-time diagrams that Minkowski introduced. Figure 26, with
its remarkable demonstration that two families of observers moving
relative to each other each see the rods of the others as contracted
relative to their own, is one of Minkowski’s actual diagrams, slightly
modi ed (merely to conform with the context of this book – the
physical content remains unchanged).

In Figure 25, the original grid is ‘painted’ onto space-time with

the dashes, while the dots show an alternative. As we saw, the law
of nature that describes the behaviour of light pulses allows them to
travel along the diagonals of either grid. A transformation of this
law from one coordinate grid to another is called a Lorentz
transformation, and the grids themselves are called Lorentz frames. I
have already mentioned that you should not think in terms of there
being one rectangular coordinate grid, and all the others oblique.
Alice thinks Alice* has an oblique system compared to her, but
Alice* thinks the same about Alice’s system. This is a consequence
of the relativity principle, and a special property of space-time that
we shall come to shortly. Minkowski pointed out that the
transformation shown in Figure 25 is a kind of rotation in four
dimensions. The possibility of making rotations in ordinary space is
a deep re ection of its unitary, block-like nature. Minkowski saw
the possibility of making a kind of rotation in space-time, which is
impossible in Newtonian space-time, as the clearest evidence for the
intimate fusion of space and time, even though the need for
‘articulation’ showed that time was still somewhat di erent in
nature from space.

Einstein, Minkowski and others were able to show that all the
laws of nature known in their time (except initially for gravitation)
either already had a form that was exactly the same in all Lorentz
frames or could be relatively easily modi ed so that they did. Even
though the modi cations were relatively easy once the idea was
clear, their implications, including Einstein’s famous equation E =
mc2 (a prediction at that time), were mostly very startling.
Minkowski, like Einstein and Poincaré, made a strong prediction
that all laws of nature found in future would accord with the
relativity principle, and emphasized that the guiding principle for
nding such laws was to treat time exactly as if it were space.

Except for the intermingling of space and time and the

distinguished role played by light, Minkowski’s space-time strongly
resembles Newtonian space-time. Matter neither creates nor changes
its rigid and absolute structure. It is like a football eld, complete
with markings, on which the players must abide by rules they
cannot change.

ARE THERE NOWS IN RELATIVITY?

It is often said that relativity destroyed the concept of Now. In
Newtonian physics the axes can never be tilted as they are in Figure
25. The simultaneity levels stay level, and there is a unique
sequence of instants of time, each of which applies to the complete
universe. This is overthrown in relativity, where each event belongs
to a multitude of Nows. This has important implications for the way
we think about past, present and future.

Even in Newtonian theory we can picture world history laid out

before us. In this ‘God’s-eye’ view, the instants of time are all ‘there’
simultaneously. The alternative idea of a ‘moving present’ passing
through the instants from the past to the future is theoretically
possible but impossible to verify. It adds nothing to the scienti c
notion of time. Special relativity makes a ‘moving present’ pretty
well untenable, even as a logical possibility.

Imagine that two philosophers meet on a walk. Each believes in

a present that sweeps through instants of time. But that implies a
unique succession of instants, or Nows. Which Nows are they? If the
two philosophers are to make such claims, they should be able to
‘produce’ the Nows through which time ows. Unfortunately, they
face the problem of the relativity of simultaneity. Each can de ne
simultaneity relative to themselves, but, since they are walking
towards each other, their Nows are di erent, and that puts paid to
any idea that there is a unique ow of time. There is no natural way
in which time can ow in Minkowski’s space-time. At least within
classical physics, space-time is a block – it simply is. This is known
as the block universe view of time. Everything – past, present and
future – is there at once. Some authors claim that nothing in
relativity corresponds to the experienced Now: there are just point-
like events in space-time and no extended Nows. At the
psychological level, Einstein himself felt quite disturbed about this.
Reporting a discussion, the philosopher Rudolf Carnap wrote:

Einstein said the the problem of the Now worried him

seriously. He explained that the experience of the Now
means something special for man, something essentially
di erent from the past and the future, but that this
important di erence does not and cannot occur within
physics. That this experience cannot be grasped by science
seemed to him a matter of painful but inevitable resignation.
So he concluded ‘that there is something essential about the
Now which is just outside the realm of science’.

The block universe picture is in fact close to my own, but the idea
that Nows have no role at all to play in physics, and must be
replaced by point-like events, would destroy my programme.
However, it is only absolute simultaneity that Einstein denied.
Relative simultaneity was not overthrown.

We are all familiar with at surfaces (two-dimensional planes) in

three-dimensional space. Such planes have one dimension fewer
than the space in which they are embedded, and are at.
Hyperplanes are to any four-dimensional space what planes are to
space. In Newtonian physics, space at one instant of time is a three-
dimensional hyperplane in four-dimensional Newtonian space-time.
It is a simultaneity hyperplane: all points in it are at the same time.
Such hyperplanes also exist in Minkowski space-time, but they no
longer form a unique family. Each splitting of space-time into space
and time gives a di erent sequence of them.

Now, what is Minkowski space-time made of? The standard

answer is events, the points of four-dimensional space-time. But
there is an alternative possibility in which three-dimensional
con gurations of extended matter are identi ed as the building
blocks of space-time. The point is that the three-dimensional
hyperplanes of relative simultaneity are vitally important structural
features of Minkowski space-time. It is an important truth that
special relativity is about the existence of distinguished frames of
reference. And an essential fact about them is that they are ‘painted’
onto simultaneity hyperplanes. As a consequence, simultaneity
hyperplanes, which are Nows as I de ne them, are the very basis of
the theory. They are distinguished features. You cannot begin to talk
about special relativity without rst introducing them. At this point,
the way both Einstein and Minkowski created special relativity
becomes signi cant.

The question is this: how is a four-dimensional structure built up

from three-dimensional elements? To make this easier to visualize,
consider the analogous problem of building up a three-dimensional
structure from cards with marks on them representing the
distribution of matter. From one set of cards with given marks,
many di erent structures can be built simply by sliding the cards
horizontally relative to one another and changing their vertical
spacings. Tait’s problem shows that in general the markings in a
structure built without special care will not satisfy the laws of
motion. What is more, to nd the correct positionings we have to
use the complete extended matter distributions. These are what I
have identi ed as instants of time. You simply cannot make the
space-time structure without using them.

The interesting thing is that neither Einstein nor Minkowski gave

serious thought to this problem – they simply supposed it had been
solved. They started their considerations at the point at which
space-time had already been put together. A comment by
Minkowski, more explicit than Einstein, makes this clear: ‘From the
totality of natural phenomena it is possible, by successively
enhanced approximations, to derive more and more exactly a system
of reference x, y, z, t, space and time, by means of which these
phenomena then present themselves in agreement with de nite
laws.’ He then points out that one such reference system is by no
means uniquely determined, and that there are transformations that
lead from it to a whole family of others, in all of which the laws of
nature take the same form. However, he never says what he means
by ‘the totality of natural phenomena’ nor what steps must be taken
to perform the envisaged successive approximations. But how is it
done? This is a perfectly reasonable question to ask. We are told
how to get from one reference system to another but not how to nd
the rst one. Had either Einstein or Minkowski asked this question
explicitly, and gone through the steps that must be taken, then the
importance of extended matter con gurations, and with them
instants of time as I de ne them, would have become apparent. This
is a key part of my argument. The accidents of the historical
development have obscured the vital role of extended Nows and
given the erroneous impression that events are primary.

I am not claiming that the description of space-time given by

Einstein or Minkowski is wrong. Far from it – they got it right, but
they described the nished product, and the complete story must
also include the construction of the product. This is best done
directly for the space-time of general relativity, the topic of the next
chapters. As preparation for them, I conclude this chapter with a
summary of the most important points.

Minkowski space-time is not some amorphous bulk in which

there is no simultaneity structure at all. We can ‘paint coordinate
lines’ – and an associated simultaneity structure – on space-time in
many di erent ways. But the whole content of the theory would be
lost if we could not do it one way or the other. There is no doubt
about it – simultaneity hyperplanes exist out there in space-time as
distinguished features.

Moreover, to give any content to relativity, we must, almost

paradoxically, assume a universality of three-dimensional things.
The clocks we can nd in one Lorentz frame must be identical to the
clocks we can nd in any other. This is a prerequisite of the
relativity principle, for it says that the laws of physics are identical
in any such frame. That would be impossible if a particular kind of
clock could exist in one frame but not in another. We can go further.
On any hyperplane in any Lorentz frame, the actual things in the
world (electric and magnetic elds, charged particles, etc.) can have
any one of a huge number of di erent arrangements. Each of them
is just like the possible distributions of particles from which we
constructed Platonia for Newtonian physics.

Exactly the same thing can be done in relativity. There is a

Minkowskian Platonia, whose points are all possible distributions of
elds and matter that one can nd on any simultaneity hyperplane
in Minkowski space. Whatever Lorentz frame we choose, the
Minkowskian Platonia always comes out the same. If it were not, the
relativity principle, with its insistence that the laws of nature are
identical in all Lorentz frames, would be meaningless. To be
identical, the laws must operate on identical things, which are
precisely the distributions that de ne the points of Platonia. For all
its four-dimensional integrity, space-time is built of three-
dimensional bricks. The beautiful four-dimensional symmetries hide
the vital role of the bricks.

It is just that space-time is not constructed from a unique set.

The analogy with a pack of cards is again quite apt. Newtonian
space-time is an ordinary pack; Minkowski space-time is a magical
pack. Look at it one way, and cards run through the block with one
inclination. Look at it another way, and di erent cards run with a
di erent inclination. But whichever way you look, cards are there.
CHAPTER 10
The Discovery of General Relativity

FUNNY GEOMETRY

This chapter is about how Einstein progressed from special relativity,

which does not incorporate gravity, to general relativity, which does.
Einstein believed that he was simultaneously incorporating Mach’s
principle as its deepest foundation, but later, as I said, he changed
his mind and left this topic in a great muddle. My view is that,
nevertheless, without being aware of it, Einstein did incorporate the
principle. This has important implications for time. We start with a
bit more about Minkowski’s discoveries, which is necessary if we are
to understand the way Einstein set about things.

One of the most important concepts in physics and geometry is

distance, which is measured with rods. Distances can be measured in
a space of any number of dimensions. You can measure them along a
line or curve, on a at or curved surface, or in space. In Part 2 we
saw how an abstract ‘distance’, the action, can be introduced in
multidimensional con guration spaces like Platonia. Minkowski
showed that a remarkable kind of four-dimensional distance exists in
space-time. Its existence is a consequence of the experimental facts
that underlie special relativity. These things are most easily
explained if we assume that space has just one dimension, not three;
space-time then has two dimensions. Such a space-time is shown in
Figure 27. We must rst of all learn about past, present, and future
in space-time.

One of the distinguished coordinate systems that exists in space-

time is shown in Figure 27, in which the x axis is for space and the t
axis for time, which increases upward. This is the Lorentz frame of
Alice in Figure 25. Her world line is the vertical t axis. The units of
time and distance are chosen to make the speed of light unity. Light
pulses that pass through event O at t = 0 in opposite directions in
space travel in space-time along the two lines marked future light
cone. Their continuations backward (the light’s motion before it
reaches O) de ne the past light cone.
Figure 27 Past and future light cones and the division of space-time
in time-like and space-like regions, as described in the text.

Each event has a light cone, but only O’s is shown. Relativity
di ers from Newtonian theory mainly through the light cone and its
associated distinguished speed c, which is a limiting speed for all
processes. Light plays a distinguished role in relativity simply
because it has that speed. No material object can travel at or faster
than it. If a material object passes through O, its world line must lie
somewhere inside the light cone, for example OA in Figure 27.

The light cone divides space-time into qualitatively di erent

regions. An event like A can be reached from O by a material object
travelling slower than light. Two such events are time-like with
respect to each other. For two such events there exists a Lorentz
frame in which they have the same space coordinates but di erent
time coordinates. For the points O and A this frame is shown in the
upper right of Figure 28.

Next we consider events like B and C in Figure 27, outside the

light cone of O. They are space-like with respect to O. No material
body can reach them from O, since to do so it would have to travel
faster than light. For two events that are mutually space-like there
exists a Lorentz frame in which they have the same time coordinate
but di erent space coordinates. For two space-like events, it is
impossible to say which is the earlier in any absolute sense. In some
Lorentz frames one will be earlier than the other (thus O is earlier
than both B and C in Alice’s frame in Figure 27), but in others the
temporal order will be reversed.
 Figure 28 Past, present and future in a space-time with two
dimensions of space. The object that moves along OA (bottom left) is
at rest in the starred frame (top right). Its world line is O*A (O and
O* are the same event).

Finally, two events that can be connected by a light ray have a

light-like relationship. All points on the light cone of event 0, for
example the point F, are light-like with respect to O.

These three basic relationships between events – being time-like,

spacelike or light-like – are the same in all Lorentz frames. This is
because the three types are determined by the light cones, which are
real features in space-time, just as rivers are real features of a
continent. In contrast, the coordinate axes are like lines ‘painted’ on
space-time – they are no more real than the grid lines on a map.
Moreover, in a change from one frame to another, the coordinate
axes never cross the light cones. The time axis moves but stays
within the light cone, while the space axes stay within the ‘present’
as de ned above. This is illustrated in Figure 28 for space with two
dimensions, which shows how the light cone gets its name. It also
highlights the great di erence between the Newtonian and
Einsteinian worlds. In the former, past, present and future are
de ned throughout the universe, and the present is a single
simultaneity hyperplane. In the latter, they are de ned separately for
each event in space-time, and the present is much larger.

Now we can talk about distance. In ordinary space it is always

positive. The distance relationships are re ected in Pythagoras’
theorem: the square of the hypotenuse in any right-angled triangle is
equal to the sum of the squares of the other two sides: H2 = A2+B2.

Minkowski was led to introduce a ‘distance’ in space-time by

noting a curious fact. For observers who use the xy frame in Figures
27 and 28, event A is separated from O by the space-like interval EA
and by the time-like interval DA. For observers who use the starred
frame, however, O and A are at the same space point and are merely
separated by the time-like interval OA. The xy observers measure EA
with a rod and DA with a clock, obtaining results X and T,
respectively. With their clock, observers in the starred frame can
measure only the time-like interval OA. Now, their clock runs at a
di erent rate to the xy clock, so they will nd that OA is not T but
Tstarred. Using Einstein’s results, Minkowski found that (Tstarred)2 =
T2−X2. This is just like Pythagoras’ theorem, except for the minus
sign.

There are several important things about this result. Einstein had
shown that observers moving relative to each other would not agree
about distances and times between pairs of events. However,
Minkowski found something on which they will always agree.
Measurements of the space-like separation (by a rod) and the time-
like separation (by a clock) of the same two events O and A can be
made by observers moving at any speed. They will all disagree about
the results of the separate measurements, but they will all nd the
same value for the square of the time-like separation minus the
square of the space-like separation. It will always be equal to the
square of the time-like separation, called the proper time, of the
unique observer for whom O and A are at the same space position.
This result created a sensation. Space and time, like rods and clocks,
seem to have completely di erent natures, but Einstein and
Minkowski showed that they are inseparably linked.
 What is more, Minkowski showed that it is very natural to regard
space and time together as a kind of four-dimensional country in
which any two points (events in space-time) are separated by a
‘distance’. This ‘distance’, found by measurements with both rods and
clocks, is to be regarded as perfectly real because everyone will agree
on its value. In fact, Minkowski argued that it is more real than
ordinary distances or times, since di erent observers disagree on
them. Only the ‘distance’ in space-time is always found to be the
same. But it is a novel distance – positive for the time-like OA in
Figure 27, zero for the light-like OF and negative for the space-like
OC. (It is a convention, often reversed, to make time-like separations
positive and space-like ones negative. What counts is that they have
opposite signs. Also, if the units of space and time are not chosen to
make the speed of light c equal to 1, the square of the space-time
‘distance’ becomes (cT)2 – X2.)

Almost everything mysterious and exciting about special

relativity arises from the enigmatic minus sign in the space-time
‘distance’. It causes the ‘skewing’ of both axes of the starred frame of
the starred twins in Figure 25, and leads to the single most startling
prediction – that it is possible, in a real sense, to travel into the
future, or at least into the future of someone else, since the future as
such is not uniquely de ned in special relativity. What we call space
and time simply result from the way observers choose to ‘paint
coordinate systems’ on space-time, which is the true reality.
Minkowski’s diagrams made all these mysteries transparent – and
intoxicatingly exciting for physicists. However, this is not the place
to discuss time travel and the other surprises of relativity, which are
dealt with extensively in innumerable other books.

EINSTEIN’S WAY TO GENERAL RELATIVITY

For physicists, ‘relativity’ has two di erent meanings. The more

common is the one employed by Einstein when he created relativity.
He related it to the empirical fact, rst clearly noted by Galileo in
1632, that all observations made within an enclosed cabin on a ship
sailing with uniform speed are identical to observations made when
the ship is at rest. Einstein illustrated this fact with experiments on
trains. The lesson he drew from it was that uniform motion as such
could not be detected by any experiment. The laws of nature could
therefore not be expressed in a unique frame of reference known to
be at rest. They could be expressed only in any one of a family of
distinguished frames in uniform motion relative to one another. The
relativity principle states that the laws of nature have the identical
form in all such frames. For reasons shortly to be explained, this later
became known as the restricted or special relativity principle.

This meaning of relativity is tied to a special feature of the world

– the existence of the distinguished frames and their equivalence for
expressing the laws of nature. The other meaning of relativity is
more primitive and less speci c. It simply recognizes that space and
time are invisible: all we ever see are objects and their relative
motions. We can speak meaningfully of the position and motion of
an object only if we say how far it is from other objects. Position and
motion are relative to other objects. This is often called kinematic
relativity, to distinguish it from Galilean relativity.

Both relativity principles have played important – often decisive –

roles in physics. Copernicus and Kepler used kinematic relativity to
great e ect in the revolution they brought about. Galileo used the
other relativity principle to explain how we can live on the Earth
without feeling its motion. That was almost as wonderful a piece of
work as Einstein’s, nearly three hundred years later. A natural
question is this: what is the connection between the two relativity
principles? Any satisfactory answer must grapple with and resolve
the issue of the distinguished frames of reference. How are they
determined? What is their origin? As we have seen, neither Einstein
nor Minkowski addressed these questions when they created special
relativity, and they have been curiously neglected ever since. This is
a pity, since they touch upon the nature of time. We cannot say what
time is – and whether it even exists – until we know what motion is.

Poincaré sought to unite the two relativity principles in a single

condition on the structure of dynamics, as formulated in the two-
snapshots idea. Had he succeeded, he would have derived the
empirical fact of Galilean relativity solely on the basis of a natural
criterion derived from kinematic relativity. He died without taking
this idea any further, but in any case it is doubtful whether the two
relativity principles can be fully fused into one. Poincaré formulated
his idea in 1902, before the relativistic intermingling of space and
time became apparent, and it is hard to see how that can ever be
derived from the bare fact of kinematic relativity. It is, however, of
great interest to see how far Poincaré’s idea can be taken. We shall
come to this when we have seen how Einstein thought about and
developed his own relativity principle and thereby created general
relativity.

It is important not to be overawed by the genius of Einstein. He

did have blind spots. One was his lack of concern about the
determination in practice of the distinguished frames that play such
a vital role in special relativity – he simply took them for granted. It
is true that they are realized approximately on the reassuringly solid
Earth in skilfully engineered railway carriages. But how does one
nd them in the vast reaches of space? This is not a trivial question.
Matching this lack of practical interest, we nd an absence of
theoretical concern. Einstein asked only what the laws of nature look
like in given frames of reference. He never asked himself whether
there are laws that determine the frames themselves. At best, he
sought an indirect answer and got into a muddle – but a most
creative muddle.

To see why, it is helpful to trace the development of his thinking

– a fascinating story in its own right. As an extremely ambitious
student, he read Mach’s critique of Newton’s absolute space. This
made him very sceptical about its existence. Simultaneously, he was
exposed to all the issues related to the aether in electrodynamics.
Lorentz, in particular, had e ectively identi ed absolute space with
the aether, in the form of an unambiguous state of rest. But, writing
to his future wife Mileva in August 1899, Einstein was already
questioning whether motion relative to the aether had any physical
meaning. This would develop into one of the key ideas of special
relativity. If it is impossible to detect motion relative to it, the aether
cannot exist. It was natural for Einstein to apply the same thought to
absolute space.

His 1905 paper killed the idea that uniform motion relative to
any kind of absolute space or aether could be detected. But Newton
had based his case for absolute space on the detection not of uniform
motion, but of acceleration. In 1933, Einstein admitted that in 1905
he had wanted to extend the relativity principle to accelerated as
well as uniform motion, but could not see how to. The great
inspiration – ‘the happiest thought of my life’ – came in 1907 when
he started to consider how Newtonian gravity might be adapted to
the framework of special relativity. He suddenly realized the
potential signi cance of the fact, noted by Galileo and con rmed
with impressive accuracy by Newton, that all bodies fall with exactly
the same acceleration in a gravitational eld.

Most physicists saw this as a quirk of nature, but Einstein

immediately decided to elevate it to another great principle and
exploit it as he had the relativity principle. Unable to divine new
laws of gravitation straight o , he formulated the equivalence
principle, according to which processes must unfold in a uniform
gravitational eld in exactly the same way as in a frame of reference
accelerated uniformly in a space free of gravity. He argued that pure
acceleration could not be distinguished from uniform gravitation.
Suppose that you awoke from a deep narcotic sleep in a dark
bedroom to nd that gravity was mysteriously stronger. There could
be two di erent causes. You might have been transported, bedroom
and all, to another planet with stronger gravity. But you might still
be on the Earth but in an elevator accelerating uniformly upward. No
experiments you could perform in your bedroom would enable you
to distinguish between these alternatives.

Einstein saw here a striking parallel with the relativity principle.

The relativity principle prevented an observer from detecting
uniform motion. In its turn, the equivalence principle prevented an
observer from detecting uniform acceleration – observed acceleration
could be attributed either to acceleration in gravity-free space or to a
gravitational eld. Einstein recognized the immediate short-term
potential of his new principle. He knew how processes unfolded in
gravity-free space. Mere mathematics showed how they would
appear in an accelerated frame, but by the equivalence principle it
was possible to deduce that these same processes must occur in a
uniform gravitational eld. Once again, Einstein’s inspired selection
of a simple universal principle – all bodies fall in the same way –
enabled him to perform a startling conjuring trick. He showed that
the rate of clocks must depend on their position in a gravitational
eld. Clocks closer to gravitating bodies must run slow relative to
clocks farther away.
 This fact is often said to show that ‘time passes more slowly’ near
a gravitating body. However, objective facts within relativity can
seem utterly mysterious and logically impossible if we imagine time
as a river. Such a time does not exist. Relativity makes statements
about actual clocks, not time in the abstract. It is easy to imagine –
and physicists now nd it comparatively easy to verify – that
otherwise identical clocks run at di erent rates at the top and
bottom of a higher tower. Incidentally, the ‘time dilation’ e ect in
gravity is much easier to accept than the similar e ect associated
with motion. There is no reciprocal slowing down. Thus, observers at
the top and bottom of the tower both agree that the clock at the top
runs faster.

By 1907 Einstein was also able to show that gravity must de ect
light. Both his early predictions, made precise in his fully developed
theory, have been con rmed with most impressive accuracy in recent
decades. However, Einstein saw his early predictions merely as
stepping stones to something far grander. The equivalence principle
persuaded him that inertia (i.e. the tendency of bodies to persist in a
state of rest or uniform motion) and gravity, which Newton and all
other physicists had regarded as distinct, must actually be identical
in nature. He started to look for a conceptual framework in which to
locate this conviction. At the same time, he saw a great opportunity
to abolish not only the aether but also all vestiges of absolute space.
So far he had managed to achieve two steps in this process by
showing that uniform motion and uniform acceleration could not
correspond to anything physically real in the world. However, much
more general motions could be imagined. Einstein aimed to show
that the laws of nature could be expressed in identical form whatever
the motion of the frame of reference.

The relativity he had so far established was very special. What he

wanted was complete general relativity. This idea, nurtured and
developed over eight years and involving intense and often agonizing
work during the last four, explains the name he gave to his uni ed
theory of gravitation and inertia that nally emerged in 1915.
Viewed in the light of the ancient debate about absolute and relative
motion, Einstein’s approach was very distinctive and somewhat
surprising since he made no attempt to build kinematic relativity
directly into the foundations of his theory. Unlike Mach and many
other contemporaries, he did not insist that only relative quantities
should appear in dynamics. He went at things in a roundabout way,
mostly because of his preference for general principles. However, I
think it was also a result of the way he thought about space and
time.

As far as I can make out, Einstein did conceive of space-time as

real and as the container of material things – elds and particles.
However, he recognized that all its points were invisible and that
they could be distinguished and identi ed only by observable matter
present at them. Since space-time was made ‘visible’ by such matter,
he supposed he could lay out coordinate grid lines on space-time and
express the laws of nature with respect to them.
 Now came the decisive issue. Einstein saw space-time without
any matter in it as a blank canvas. Nothing about it could suggest
why the coordinate grid lines should be drawn in one way rather
than another. Any choice would be arbitrary and violate the
principle of su cient reason. Einstein found this intolerable. That is
no exaggeration: his faith in rationality of nature – as opposed to
human beings – was intense. The only satisfactory resolution was
general relativity. In truth, there can be no distinguished coordinate
systems. It must be possible to express the laws of nature in all
systems in exactly the same form.

The only justi cation for the distinguished systems that appeared
in Newtonian dynamics and special relativity was the law of inertia.
But the equivalence principle had opened up the possibility of
unifying inertia and gravity. This insight sustained Einstein in his
long search for general relativity. His contemporaries would all have
been content simply to nd a new law of gravity. He was after
something sublime.

It is suggestive that both Poincaré and Einstein – the old and

young giants – began their attack on absolute space from the
principle of su cient reason. The di erence between their
approaches is interesting. Working within the traditional dynamical
framework, Poincaré said that only directly observable quantities –
the relative separations of bodies and their rates of change – should
be allowed as initial data for dynamics. In such a theory, we may say
that perfect Laplacian determinism holds (it doesn’t hold in Newtonian
theory, which uses invisible absolute space and time). Einstein had a
more general approach. He merely insisted that there should be no
arbitrary choice of the coordinate systems used to express the laws of
nature.

THE MAIN ADVANCES

The desire to express the laws of nature in progressively more

general coordinates led to all Einstein’s major breakthroughs.
Newton had argued that centrifugal forces proved the existence of
absolute space. The laws of nature looked di erent in rotating
systems. Einstein wanted to attack this problem head on. Could he
perhaps show that, if expressed properly, the laws of nature did after
all have the same form in rotating and non-rotating coordinates? The
principle of equivalence suggested that what Newton had taken to be
absolute inertial e ects in a rotating system might be the
gravitational e ects of distant matter. The point is that in a rotating
system the distant stars would themselves appear to be rotating.
Since rotating electric charges generate electric and magnetic elds,
rotating masses might generate new kinds of gravitational eld.
Nearly thirty years earlier, Mach had suggested that rotating matter
‘many leagues thick’ might generate measurable centrifugal forces
within it. Einstein now conjectured that the gravitational eld was
the mechanism through which such forces could arise.

He therefore started to consider what form the laws of nature

would take in a rotating system. This immediately led him to a
startling conclusion: the ordinary laws of Euclidean geometry could
not hold in such a system! His argument was based on the
contraction of measuring rods in motion which he had proved in
special relativity. First, imagine observers at rest on a surface who
measure the circumference and diameter of a circle painted on it.
They will nd that their ratio is π. That agrees with Euclidean
geometry – a recognized law of nature. Now imagine other observers
on a disk above the painted circle and rotating about its centre. Their
rods will undergo Fitzgerald-Lorentz contraction when laid out in the
direction of motion, around the circumference. However, when laid
out along the diameter, the rods will not contract. (The contraction
occurs only in the direction of motion.) Therefore, the rotating
observers will not nd π when they measure the ratio of the
circumference to the diameter. For them, Euclidean geometry will
not hold.

Because Einstein wanted so passionately to generalize the

relativity principle, he took this result seriously. According to the
hint from the equivalence principle, novel e ects in accelerated
coordinate systems (as a rotating one is) could be attributed to
gravitational e ects. He concluded that geometry would not be
Euclidean in a gravitational eld. This happened during 1911/12,
when he was working in Prague. Through either the suggestion of a
colleague or the recollection of lectures on non-Euclidean geometry
he had heard as a student, Einstein’s attention was drawn to a classic
study in the 1820s by the German mathematician Carl Friedrich
Gauss.
 Gauss had studied the curvature of surfaces in Euclidean space.
As a rule, material surfaces in space are not at but curved. Think of
the surface of the Earth or any human body. Gauss’s most important
insight was that a surface in three-dimensional space is characterized
by two distinct yet not entirely independent kinds of curvature. He
called them intrinsic and extrinsic curvature. The intrinsic curvature
depends solely on the distance relationships that hold within the
surface, whereas the extrinsic curvature measures the bending of the
surface in space. A surface can be at in itself – with no intrinsic
curvature – but still be bent in space and therefore have extrinsic
curvature. The best illustration of this is provided by a at piece of
paper, which has no intrinsic curvature. As it lies on a desk it has no
extrinsic curvature either: it is not bent in space. However, it can be
rolled into a tube. It is then bent – but not stretched – and acquires
extrinsic curvature.

In contrast to a sheet of paper, the surface of a sphere, like the

earth, has genuine intrinsic curvature. Gauss realized that important
information about it could be deduced from distance measurements
made entirely within the surface. Imagine that you can pace
distances very accurately, and that you walk due south from the
north pole until you reach latitude 85° north. Then you turn left and
walk due east all the way round the Earth at that latitude. All the
time you will have remained the same distance R from the north
pole. If you believed the Earth to be at, you would expect to have
to walk the distance 2πR before returning to the point of your left
turn. However, you will nd that you get there having walked a
somewhat shorter distance. This shows you that the surface of the
Earth is curved.

To describe these things mathematically for all smooth surfaces,

Gauss found it convenient to imagine ‘painting’ curved coordinate
lines on the surface. On a at surface it is possible to introduce
rectangular coordinate grids, but not if the surface is curved in an
arbitrary way. So Gauss did the next best thing, which is to allow the
coordinate lines to be curved, like the lines of latitude and longitude
on the surface of the Earth. He showed how the distance between
any two neighbouring points on a curved surface could be expressed
by means of the distances along coordinate lines, and also how
exactly the same distance relations could be expressed by means of a
di erent system of coordinates on the same surface. About thirty
years later, another great German mathematician, Bernhard
Riemann, showed that not only two-dimensional surfaces but also
three-dimensional and even higher-dimensional spaces can have
intrinsic curvature. This is hard to visualize, but mathematically it is
perfectly possible. Just as on the Earth, in a curved space of higher
dimensions, you can, travelling always in the same direction, come
back to the point you started from. These more general spaces with
curvature are now called Riemannian spaces.

Einstein realized that he had to learn about all this work

thoroughly, and it was very fortunate that he moved at that time to
Zurich, where Marcel Grossmann, an old friend from student days,
was working. Grossmann gave him a crash course in all the
mathematics he needed. When he had fully familiarized himself with
it, Einstein became extremely excited for two reasons.

First, Minkowski had shown that space-time could be regarded as

a four-dimensional space with a ‘distance’ de ned in it between any
two points. Except that the ‘distance’ was sometimes positive and
sometimes negative, whereas Riemann had assumed the distance to
be always positive and had never envisaged time as a dimension,
considered mathematically Minkowski’s space-time was just like one
of Riemann’s spaces. But it was special in lacking curvature – it was
like a sheet of paper rather than the Earth’s surface. Einstein had
meanwhile become convinced that gravity curves space-time. This
led to one of his most beautiful ideas: in special relativity, the world
line (path) of a body moving inertially is a straight line in space-
time. This is a special example of a ‘shortest curve’, or geodesic. The
corresponding path in a space with curvature would be a geodesic,
like a great circle on a sphere.

Einstein assumed that the world line of a body subject to inertia

and gravity would be a geodesic. In this way he could achieve his
dream of showing that inertia and gravity were simply di erent
manifestations of the same thing – an innate tendency to follow a
shortest path. This will be a straight line if no gravity is present, so
that space-time has no curvature, but in general it will be a curved
(but ‘straightest’) line in a genuinely curved space-time. Since matter
causes gravity, Einstein assumed that matter must curve space-time
in accordance with some law, for which he immediately started to
look. Bodies moving in such a space-time would follow the geodesics
corresponding to the curvature produced by the matter, so the
gravitational e ect of the matter would be expressed through the
curvature it produces. Another important insight was that in small
regions the e ect of curvature would be barely noticeable, just as the
Earth seems at in a small region, so that in those small regions
physical phenomena would appear to unfold just as in special
relativity without gravity. This gave full expression to the
equivalence principle.

The second reason why Einstein became so excited was that

Gauss’s method matched his own idea of general relativity. He
disliked the distinguished frames of special relativity because they
corresponded to special ways of ‘painting’ coordinate systems onto
space-time. He felt that this was the same as having absolute space
and time. They would be eliminated only if the coordinate systems
could be painted on space-time in an arbitrary way. But this was
what Gauss’s method amounted to. In fact, in a curved space it is
mathematically impossible to introduce rectangular coordinates.
Mathematicians call the possibility of using completely arbitrary
coordinate systems general covariance. Speci cally, laws are said to
be generally covariant if they take exactly the same form in all
coordinate systems. Einstein identi ed this with his requirement of
general relativity.

To summarize this part of the story, in 1912 Einstein became

aware of the possibilities opened up by non-Euclidean geometry and
the work initiated by Gauss. He had begun to suspect that
gravitational elds would make geometry non-Euclidean. He was
also almost desperate to nd a formalism that did not presuppose
distinguished frames of reference. He found that Gauss’s method of
arbitrary coordinates was tailor-made for his ambitions. He also saw
that, space and time having been so thoroughly fused by Minkowski,
the only natural thing to do was to make space-time into a kind of
Riemannian space. The ideas of Gauss and Riemann must be applied,
not to space alone, but to space and time. This is the incredibly
beautiful idea that Minkowski made possible: gravity was to be
explained by curvature in space and time. Einstein thus conjectured
that space-time is curved by gravity, and that bodies subject only to
gravity and inertia follow geodesics determined by the distance
properties of space-time, which encapsulate all its geometrical
properties. Einstein’s conjecture has been brilliantly con rmed to
great accuracy in recent decades.

THE FINAL HURDLE

Finding the law of motion of bodies in a gravitational eld was only

part of Einstein’s problem. He also had to nd how matter created a
gravitational eld. He needed to nd equations for the gravitational
eld somewhat like those that Maxwell had found for the
electromagnetic eld. They would establish how matter interacted
with the gravitational eld, and also how the eld itself varied in
regions of space-time free of matter (matching the way
electromagnetic radiation propagated as light through space-time).
This part of the problem created immense di culties for Einstein,
mostly through very bad luck.

Much as I would like to tell the complete story, which is

fascinating and full of ironies, I shall have to content myself with
saying that, after three nerve-wracking years, Einstein nally found a
generally covariant law that described how matter determined the
curvature of space-time. It involves mathematical structures called
tensors, all the properties of which had already been studied by
mathematicians. In particular, for space-time free of matter, Einstein
was able to show that a tensor known as the Ricci tensor (because it
had been studied by the Italian mathematician Gregorio Ricci-
Curbastro) must be equal to zero. Ironically, Grossmann had already
suggested to Einstein in 1912 that in empty space the vanishing of
the Ricci tensor might be the generally covariant law he was seeking.
However, some understandable mistakes prevented them from
recognizing the truth at that time.

It is a striking fact that all the mathematics Einstein needed

already existed. In fact, I believe it is signi cant that he did not have
to invent any of it. In 1915, he was immediately able to show that, to
the best accuracy astronomers could achieve at that time, his theory
gave identical predictions to Newtonian gravity except for a very
small correction to the motion of Mercury. All planetary orbits are
ellipses. A planet’s elliptical orbit itself very slowly rotates, under the
gravitational in uence of the other planets. This is known as the
advance of the perihelion, the perihelion being the point at which
the planet is closest to the Sun, marking one end of the ellipse’s
longest diameter. According to Einstein’s theory, Mercury’s
perihelion should advance by 43 seconds of arc per century more
than was predicted by Newtonian theory. This very small e ect
shows up for Mercury because it is closer to the Sun than the other
planets, and also has a large orbital eccentricity. For many years, the
sole discrepancy in the observed motions of the planets had been
precisely such a perihelion advance for Mercury of exactly that
magnitude. All attempts to explain it had hitherto failed. Einstein’s
theory explained it straight o .

GENERAL RELATIVITY AND TIME

Many more things could be said about general relativity and its
discovery. However, what I want to do now is identify the aspects of
the theory and the manner of its discovery that have the most
bearing on time.

First, the classical (non-quantum) theory as it stands seems to

make nonsense of my claim that time does not exist. The space-time
of general relativity really is just like a curved surface except that it
has four and not two dimensions. A two-dimensional surface you can
literally see: it is a thing extended in two dimensions. In their mind’s
eye, mathematicians can see four-dimensional space-time, one
dimension of which is time, just as clearly. It is true that time-like
directions di er in some respects from space-like directions, but that
no more undermines the reality of the time dimension than the
di erence between the east-west and north-south directions on the
rotating Earth makes latitude less real than longitude. However, the
quali cation ‘as it stands’ at the start of this paragraph is important.
In the next chapter we shall see that there is an alternative, timeless
interpretation of general relativity.

Next, there is the matter of the distinguished coordinate systems.

In one sense, Einstein did abolish them. Picture yourself in some
beautiful countryside with many varied topographic features. They
are the things that guide your eye as you survey the scene. The real
features in space-time are made of curvature, and hills and valleys
are very good analogies of them. Imagined grid lines are quite alien
to such a landscape. In general relativity, the coordinated lines truly
are merely ‘painted’ onto an underlying reality, and the coordinates
themselves are nothing but names by which to identify the points of
space-time.

For all that, space-time does have a special, sinewy structure that
needs to be taken into account. Distinguished coordinate systems still
feature in the theory. This is because the theory of measurement and
the connection between theory and experiment is very largely taken
over from special relativity. In fact, much of the content of general
relativity is contained in the meaning of the ‘distance’ that exists in
space-time. This is where the analogy between space-time and a
landscape is misleading. We can imagine wandering around in a
landscape with a ruler in our pocket. Whenever we want to measure
some distance, we just sh out the ruler and apply it to the chosen
interval. But measurement in special relativity is a much more subtle
and sophisticated business than that. In general, we need both a rod
and a clock to measure an interval in space-time. Both must be
moving inertially in one of the frames of reference distinguished by
that theory, otherwise the measurements mean nothing. The theory
of measurement in general relativity simply repeats in small regions
of space-time what is done in the whole of Minkowski space-time in
special relativity. No measurements can be contemplated in general
relativity until the special structure of distinguished frames that is
the basis of special relativity has been identi ed in the small region
in which the measurements are to be made.

This is something that is often not appreciated, even by experts. It

comes about largely because of the historical circumstances of the
discovery of general relativity and the absence of an explicit theory
of rods and clocks. There is also the stability of our environment on
the Earth and the ready availability in our age of clocks. It is easy for
us to stand at rest on the Earth, watch in hand, and perform a
measurement of a purely timelike distance. But nature has given us
the inertial frame of reference for nothing, and skilful engineers
made the watch. Finally, because we can and very often do see a
three-dimensional landscape spread out before our eyes, it is very
easy to imagine four-dimensional space-time displayed in the same
way. All textbooks and popular accounts of the subject positively
encourage us to do so. They all contain ‘pictures’ of space-time. Now
the picture is indeed there, and very wonderful it is too. But it arises
in an immensely sophisticated manner hidden away within the
mathematical structure of the Ricci tensor. The story of time as it is
told by general relativity unfolds within the Ricci tensor. It performs
the miracle – the construction of the cathedral of space-time by
intricate laying and interweaving of the bricks of time. I shall try to
explain this in qualitative terms in the next chapter. Let me conclude
this one by highlighting again the importance of the historical
development. It made possible the discovery of a theory without full
appreciation of its content.

At the end of November 1915, Einstein wrote an ecstatic letter to

his lifelong friend Michele Besso, telling him that his wildest dreams
had come true: ‘General covariance. Mercury’s perihelion with
wonderful accuracy.’ These two verbless sentences say it all. Einstein
was convinced that general covariance had deep physical
consequences and had led him to one of the greatest triumphs of all
time. Yet, barely two and a half years later, he admitted, in response
to a quite penetrating criticism from a mathematician called Erich
Kretschmann, that general covariance had no physical signi cance at
all.

In a way, this is obvious. Space-time is a beautiful sculpture.

What makes it beautiful is the way in which its parts are put
together. The fact that one can paint coordinate lines on the nished
product and measure distance on the sculpture between points on it
labelled by the arbitrary coordinates clearly leaves the sculpture
exactly the same. All this changing of coordinates is purely formal. It
tells you nothing about the true rules that make the sculpture.

Belatedly, Einstein came to see that his whole drive to achieve

general covariance as a deep physical principle had no foundation in
fact. It was just a formal mathematical necessity. Ever determined to
nd new and even more beautiful laws of nature, he never felt the
need to go back and see exactly how his sculpture was actually
created. In a book I wrote some years ago on the discovery of
dynamics, I commented on the fact that Kepler (so very like Einstein
in his dogged holding on to an idea that eventually transformed
physics) never realized quite what a wonderful discovery he had
made. I likened him to

a boy who nds for the rst time a ripe horse-chestnut with
the outer shell intact. Cherishing the golden and curiously
shaped object, he might take it home, quite unaware of the
shiny brown and perfectly smooth conker ready to spring
from the shell on application of a little directed pressure.
That was Kepler’s fate: he died without an inkling of what his
nut really contained.

The same thing happened to Einstein. Realizing while still at Prague

the sort of thing he needed, he hurried to a shop called ‘Mathematics’
owned by his friend Grossmann in Zurich. Straight o the shelf, at a
bargain price, he bought a wonderful device called the Ricci tensor.
Three years later, after agonizing struggles, he learned how to turn
the handles properly, and out popped the advance of Mercury’s
perihelion and the exact light de ection at eclipses.

But it never entered his head to ask how the device actually
worked. He died only half aware of the miracle he had created.
CHAPTER 11
General Relativity: The Timeless Picture

THE GOLDEN AGE OF GENERAL RELATIVITY

Strange as it may seem, general relativity was little studied for about
forty years. This was not for want of admiration, for it was soon
recognized as a supreme achievement. Con rmation of the predicted
bending of starlight near the Sun by Arthur Eddington’s eclipse
expedition in 1919, communicated by telegram to The Times, made
Einstein into a world celebrity overnight. The problem was that there
seemed to be little one could do except wonder at the miracle of the
theory he had created.

The main di culty was the extreme weakness of all readily

accessible gravitational elds. Apart from three small di erences
from Newtonian theory, which were all reasonably well con rmed,
no further experimental tests seemed possible. A further problem was
the mathematical complexity of the theory. Its solutions contained
fascinating structures, above all black holes, but it was decades
before these were discovered and fully understood. Finally, interest
in general relativity was overshadowed by the discovery in 1925/6
of quantum mechanics. In fact, truly active research in general
relativity commenced only in 1955, ironically the year Einstein died,
with a conference held in Bern (where Einstein had worked as a
patent clerk in 1905) to mark the ftieth anniversary of special
relativity.

Since then, research has concentrated in three main elds. First,

there have been tremendous experimental advances, made possible
above all by technological developments, including space
exploration. The foundations and some detailed predictions of the
theory have been tested to a very high accuracy. Particularly
important was the discovery a quarter of a century ago of the rst
binary pulsar, observations of which have provided strong evidence
for the existence of the gravitational waves predicted by the theory.
General relativity has also played a crucial role in observational
astronomy and cosmology.

There have been two broad avenues of theoretical research. First,

general relativity has been studied as a classical four-dimensional
geometrical theory of space-time, a systematical and beautiful
development of Minkowski’s pioneering work. Roger Penrose has
probably done more than anyone else in this eld, though many
others, including Stephen Hawking, have made very important
contributions. Second, the desire to understand the connection
between general relativity and quantum mechanics (Box 2) has
stimulated much work. Here it is necessary to distinguish two
programmes. The less ambitious one accepts space-time as a classical
background and seeks to establish how quantum elds behave in it.
This work culminated in the amazing discovery by Hawking that
black holes have a temperature and emit radiation. In Black Holes
and Time Warps, Kip Thorne has given a gripping account of this
story. Although the full signi cance of Hawking’s discovery is still far
from understood, nobody doubts its importance for the more
ambitious programme, which is to transform general relativity itself
into a quantum theory (Box 2). This transformation, which has not
yet been achieved, is called the quantization of general relativity.

In fact, many researchers believe that it is a mistake to try to

quantize general relativity directly before gravity has been uni ed
with the other forces of nature. This they hope to achieve through
superstring theory. However, a substantial minority believe that
general relativity contains fundamental features likely to survive in
any future theory, and that a direct attempt at its quantization is
therefore warranted. This is my standpoint. In particular, I regard
general relativity as a classical theory of time. It must surely be
worth trying to establish its quantum form. Even if we have to await
a future theory for the nal details, the quantization of general
relativity should give us important hints about the quantum theory
of time.

It was the desire to quantize general relativity that led to the

work described in this chapter. One important approach, called
canonical quantization, is based on analysis of the dynamical structure
of the classical theory. This is how general relativity came to be
studied in detail as a dynamical theory nearly half a century after its
creation as a geometrical space-time theory. The ‘hidden dynamical
core’, or deep structure, of the theory was revealed. The decisive
analysis was made in the late 1950s by Paul Dirac and the American
physicists Richard Arnowitt, Stanley Deser and Charles Misner. They
created a particularly elegant theory, now known universally as the
ADM formalism. (Because it is regarded as controversial by some, the
initials are occasionally reshu ed as MAD or DAM.)

The dynamical form of general relativity is often called

geometrodynamics. The term, like ‘black hole’ and several others, was
coined by John Wheeler, who, together with his many students at
Princeton, did much to popularize this form of the theory. The
interpretation of it proposed in this chapter is very close to one put
forward by Wheeler in the early 1960s. However, I believe it brings
out the essentially timeless nature of general relativity rather more
strongly than Wheeler’s well-known writings of that period. What is
at stake here is the plan of general relativity. What are its ultimate
elements when it is considered as a dynamical theory, and how are
they put together?

This is what Dirac and ADM set out to establish. The answer was
manifestly a surprise for Dirac at least, since it led him to make the
remarkable statement quoted in the Preface. They found that if
general relativity is to be cast into a dynamical form, then the ‘thing
that changes’ is not, as people had instinctively assumed, the four-
dimensional distances within space-time, but the distances within
three-dimensional spaces nested in space-time. The dynamics of
general relativity is about three-dimensional things: Riemannian
spaces.

PLATONIA FOR RELATIVITY

To connect this with the topics of Part 2, let me tell you about the
work that Bruno Bertotti and I did after the work described there. We
began to wonder whether we could be more ambitious and construct
not merely a non-relativistic, Machian mechanics, but perhaps an
alternative to general relativity. At the time, we believed that
Einstein’s theory did not accord with genuine Machian principles.
Experimental support for it was beginning to seem rather convincing,
but tiny e ects have often led to the replacement of a seemingly
perfect theory by another with a very di erent structure. We were
aware of quite a lot of the work of Wheeler and ADM, and various
arguments persuaded us that the geometry of three-dimensional
space might well be Riemannian, possess curvature and evolve in
accordance with Machian principles. We wanted to nd a Machian
geometrodynamics, which we did not think would be general
relativity. The rst task was to select the basic elements of such a
theory. What structures should represent instants of time and be the
points of the theory’s Platonia?

This question was easily answered. Any class of objects that di er

intrinsically but are all constructed according to the same rule can
form a Platonia. So far, we have considered relative con gurations of
particles in Euclidean space. There is nothing to stop us considering
three-dimensional Riemannian spaces, especially if they are nite
because they close up on themselves. This is di cult for a non-
mathematician to grasp, but the corresponding things in two
dimensions are simply closed, curved surfaces like the surface of the
Earth or an egg. The points of Platonia for this case are worth
describing. The surface of any perfect sphere is one point; each
sphere with a di erent radius is a di erent point. Now imagine
deforming a sphere by creating puckers on its surface. This can be
done in in nitely many ways. There can be all sorts of ‘hills’ and
‘valleys’ on the surface of a sphere, just as there are on the Earth and
the Moon. And there is no reason why the surface should remain
more or less spherical: it can be distorted into innumerable di erent
shapes to resemble an egg, a sausage or a dumbbell. On all of these
there can be hills and valleys. Each di erent shape is just one point
in Platonia, and could be a model instant of time. In this case you
can form a very concrete image of what each point in Platonia looks
like. These are things you could pick up and handle. Note that only
the geometrical relationships within the surface count. Surfaces that
can be bent into each other without stretching, like the sheet of
paper rolled into a tube, count as the same. However, this is a mere
technicality. The important thing is that the points of any Platonia
are real structured things, all di erent from one another.

Imagining the points that constitute this Platonia is easy enough.

It is much harder to form a picture of Platonia itself because it is so
vast and has in nitely many dimensions. Triangle Land has three
dimensions, and we can give a picture of it (Figures 3 and 4). But
Tetrahedron Land already has six dimensions, and is impossible to
visualize. When there are in nitely many dimensions, all attempts at
visualization break down, but as mathematical concepts such
Platonias do exist and play important roles in both mathematics and
physics.

Riemannian spaces are actually empty worlds since they contain

nothing that we should recognize as matter. You might wonder in
what sense they exist. They certainly exist as mathematical
possibilities, and the proof of this was one of the great triumphs of
mathematics in the nineteenth century. But they can also contain
matter, just like at familiar Euclidean space. Its properties and
existence were originally suggested by the behaviour of matter
within it, and evidence for curved space can be deduced through
matter as well, as the experimental con rmations of general
relativity show. I hope that this disposes of any worries you might
have. In fact, the Platonia of three-dimensional Riemannian spaces is
well known in the ADM formalism as superspace (another Wheeler
coining, and not to be confused with a di erent superspace in
superstring theory).

The Platonia that models the actual universe certainly cannot

consist of only empty spaces, since we see matter in the world. To
get an idea of what is needed, imagine surfaces with marks or
‘painted patterns’ on them to represent con gurations of matter or
electric, magnetic or other elds in space. This will hugely increase
the number of points in Platonia, since now they can di er in both
geometry and the matter distributions. Any two con gurations that
di er intrinsically in any way count as di erent possible instants of
time and di erent points of Platonia.

Within classical general relativity, the concept of superspace is

not without di culties, which could undermine my entire
programme. Since the issues are decidedly technical, I have put the
discussion of them in the Notes. However, I can say here that
marrying general relativity and quantum mechanics is certain to
require modi cation of the patterns of thought that have been
established in the two separate theories. Superspace certainly arises
as a natural concept in the framework of general relativity. The
question is whether it is appropriate in all circumstances.

I feel that, when everything has been taken into account,

superspace is the appropriate concept, though its precise de nition
and the kinds of Nows it contains are bound to be very delicate
issues. Now, making the assumption they can be sorted out, what can
we do with the new model Platonia?

BEST MATCHING IN THE NEW PLATONIA

The key idea in Part 2 is the ‘distance’ between neighbouring points

in Platonia based solely on the intrinsic di erence between them. It
was obvious to Bruno and me that if we were to make any progress
with our more ambitious goal, we should have to nd an analogous
distance in the new Platonia. We had to look for some form of best
matching appropriate in the new arena.
 To explain the problem, let me rst recall what best matching
does and achieves in the Newtonian case of a large (but xed)
number of particles. Each instant of time, each Now, is de ned by a
relative con guration of them in Euclidean space. We modelled each
Now as a ‘megamolecule’, and compared two such Nows, without
reference to any external space or time, by moving one relative to
the other until they were brought as close as possible to coincidence
as measured by a suitable average. This is where the real physics
resides, since the residual di erence between the Nows in the best-
matching position de nes the ‘distance’ between them in Platonia.
Once we possess all such ‘distances’ between neighbouring Nows, we
can determine the geodesics in Platonia that correspond to classical
Machian histories. Besides de ning these ‘distances’, the best
matching automatically brings the two Nows into the position they
have in Newton’s absolute space, if we want to represent things in
that way.

However, to complete that Newtonian-type picture, we have still

to determine ‘how far apart in time’ the two Nows are. This is the
problem of nding the distinguished simpli er, the time separation
that unfolds the dynamical history in the simplest or most uniform
way. As we saw in the nal section of Chapter 6, in the discussion of
ephemeris time, the choice of distinguished simpli er is unique if we
want to construct clocks that will enable their users to keep
appointments. Our ability to keep appointments is a wonderful
property of the actual world in which we nd ourselves, and we
must have a proper theoretical understanding of its basis. This is
achieved if we insist that a clock is any mechanism that measures, or
‘marches in step with’, the distinguished simpli er. This is the theory
of duration and clocks that Einstein never addressed explicitly.
However, the most important thing is that history itself is
constructed in a timeless fashion. The distinguished simpli er is
introduced after the event to make the nal product look more
harmonious. Duration is in the eye of the beholder.

In Newtonian best matching, the compared Nows are moved

rigidly relative to one another. We could conceive of a more general
procedure, but since the Nows are de ned by particles in Euclidean
space its atness and uniformity make that an additional
complication. We should always try to keep things simple.

However, if we adopt curved three-dimensional spaces, or 3-

spaces as they are often called, as Nows, any best-matching
procedure for them will have to use a more general pairing of points
between Nows. For example, two 3-spaces (which may or may not
contain matter) may have di erent sizes. It will then obviously be
impossible to pair up all points as if they were sitting together in the
same space. More generally, the mere fact that both spaces are
curved – and curved in di erent ways – forces us to a much more
general and exible method for achieving best matching.

In a talk, I once illustrated what has to be done by means of two

magni cent fungi of the type that grow on trees and become quite
solid and rm. For reasons that will become apparent, I called them
Tristan and Isolde. Tristan was a bit larger than Isolde, and both
were a handsome rich brown, the darkness of which varied over
their curved and convoluted surfaces. I wanted to explain how one
could determine a ‘di erence’ between the two by analogy with the
best-matching for mass con gurations in at space. In some way, this
would involve pairing each point on Tristan to a matching point on
Isolde. A little re ection shows that the only way to do this is to
consider absolutely all possible ways of making the matching.

I took lots of pins, numbered 1, 2, 3, ..., and stuck them in

various positions into Tristan. I then took a second set, also
numbered 1, 2, 3, ..., and stuck them into Isolde. Since they had
similar shapes, I placed the pins in corresponding positions, as best
as I could judge. I could then say that, provisionally, pin 1 on Tristan
was ‘at the same position’ as pin 1 on Isolde. All the other points on
them were imagined to be paired similarly in a trial pairing.

This made it possible to determine a provisional di erence. For

example, I could compare the two fungi using the darkness of their
brown surfaces. Alternatively, and much closer to what happens in
general relativity, I could compare the curvatures at matching points.
The essential point is that some intrinsic property is compared at
each pair of matched points, and an average of all the resulting
di erences is then determined. This average, one number, is the
provisional di erence. I leave out the mathematical details, which
are intricate even though the underlying idea is simple.

This provisional di erence is clearly arbitrary since the pairing on

which it is based could have been made di erently. To nd an
intrinsic di erence that can have real physical meaning, we must
now embark – in imagination at least – on an immensely laborious
task. Keeping the pins on Tristan xed, we need to rearrange the pins
(reasonably continuously so that the mathematics works) on Isolde in
every conceivable way. For each trial pairing of all points on Isolde
to all points on Tristan, we must nd the provisional di erence. We
shall know that we have found the best-matching pairing and
corresponding intrinsic di erence when the provisional di erence
remains unchanged if we go from the given pairing to any other
pairing that di ers from it ever so slightly. (In mathematics, the
ful lment of this condition indicates that one has found a maximum,
a minimum or a so-called stationary point of the quantity being
considered. It turns out that a stationary point is what is found in
this case, but that is a mere technicality.) Since there is an immense
– indeed in nite – number of ways of changing the pairings, the best-
matching requirement imposes a very strong condition. It is
impossible to conceive of a more re ned and delicate comparison of
two things that are di erent but of the same kind. However, as
Bruno and I realized, it is made necessary by the nature of the
compared things.

It leads immediately to the ne plus ultra of best matching – and

rationality.

CATCHING UP WITH EINSTEIN

It was around 1979 that Bruno and I developed the new best-
matching idea. We did quite a lot of technical work, and were
beginning to get quite hopeful. We knew that we could construct
various forms of Machian geometrodynamics, and we began to think
that one of them might be a serious rival to general relativity. But it
is not easy to beat Einstein, as we were soon to nd. This came about
through the intervention of another friend, Karel Kuchař, whom I
had got to know in 1972, when we had several discussions. Karel is
Czech and studied physics at the Charles University in Prague,
specializing in relativity. In 1968 he won an award to study at
Princeton with John Wheeler, where he quickly established himself
as a leading expert in the canonical quantization of gravity (the most
straightforward quantization procedure (Box 2) that can be used in
the attempt to quantize gravity), in which Dirac and ADM had been
the pioneers. Some years later he became a professor of physics at
the University of Utah in Salt Lake City, where he still works. Over
the years I have pro ted greatly from discussions with Karel, and
certainly would not have been in the position to write this book
without assistance from him at some crucial points. However, I
hasten to add that Karel is sceptical about my idea that time does not
exist at all. As we shall see, general relativity presents a great
dilemma. Karel gives more weight to one Born of this dilemma, I to
the other.

The issue came into clear focus for me in 1980. In April of that
year, Karel gave a memorable review talk at an international
conference in Oxford, during which I had an opportunity to discuss
with him the ideas that Bruno and I were developing. He invited me
to come to Salt Lake City, which I did in the late fall, just in time to
see the pale gold of the aspens in the Wasatch mountains. Getting to
know Utah and the magni cent deserts of the western United States
has been a great bonus from the study of physics for me and my
family. But as this is a book about physics, not travel, I had better
not digress.

To come straight to the point, it soon became clear in the

discussions with Karel that the idea of best matching and the whole
way of thinking about duration as a measure of di erence were
already both contained within the mathematics of general relativity,
though not in a transparent form. These facts are still not widely
known, mainly, I think, because of a certain inertia. General
relativity was discovered as a theory of four-dimensional space-time,
and that is still essentially the way it is presented. The fact that it is
simultaneously a dynamical theory describing the changes of three-
dimensional things is given much less weight. This is why so few
people are aware that there is such a deep issue and crisis about the
nature of time at the heart of general relativity.

I think that the nature of the problem can be explained to a non-

scientist. Here, at least, is my attempt. Figure 29 is a very schematic
representation of the three di erent kinds of four-dimensional space-
time that have been considered in this book. As usual, only one of
the three dimensions of space is shown. It and its material contents
are represented by the horizontal direction, while time runs
vertically. Thus, the more or less horizontal lines and curves in the
three parts of the diagram represent space and its material contents
at di erent ‘times’. They are each Nows in my sense. As we have
seen, Newtonian space-time is like a pack of ordinary cards. Each
card is a Now, and they are all horizontal. I called Minkowski space-
time a magical pack of cards because its Nows, or hyperplanes of
simultaneity, can be drawn in di erent ways. Depending on the
Lorentz frame that is chosen, di erent families of parallel Nows are
obtained. Time has become relative to the frame. In general
relativity, this relativity of time is taken much further: provided the
Nows do not cut through the light cone, they can be drawn in an
immense number of di erent ways. It is the complete absence of
uniqueness in the way this is done that led Einstein to comment that
the concept of Now does not exist in modern physics. However, this
re ects the space-time viewpoint. The dynamical viewpoint puts
things in a di erent perspective.

Figure 29 The three di erent kinds of space-time: on the left,

Newtonian space-time, with ‘horizontal’ Nows; in the middle,
Minkowski space-time, with alternative ‘tilted’ Nows; on the right,
 the space-time of general relativity, with Nows running in arbitrary
directions.

To see this, suppose we consider two neighbouring Nows, as

shown in Figure 30, in a space-time that satis es the equations of
general relativity. Each Now is a 3-space with its own intrinsic three-
dimensional geometry and material contents embedded within space-
time. This four-dimensional space-time has its own geometry too,
and permits the construction of ‘struts’ between the two Nows. The
struts are the world lines of bodies that follow geodesics in space-
time, leaving the earlier Now along the space-time direction that is
perpendicular to it at the point of departure. Each ‘strut’ is, so to
speak, erected on the rst Now. It will pierce the second Now at
some point. Taken altogether, such struts uniquely determine a
pairing of each point of the rst Now with a point of the second
Now. They do something else, too. If a clock travels along each strut
between its two ends, it will measure the proper time between them
as it goes. Because the two Nows have been chosen arbitrarily, the
proper time will in general be di erent for each strut.
 Figure 30 The two continuous curves represent (in one dimension)
the two slightly di erent 3-spaces mentioned in the text; the more or
less vertical lines are the ‘struts’.

What has this to do with best matching? Everything. Imagine

mean-minded mathematicians who stick ‘pins’ like those that I stuck
into Tristan and Isolde into the two 3-spaces to identify the two ends
of all the struts in Figure 30. The pins carry little ags with the
‘lengths’ of the corresponding struts – the proper time – along them.
However, all this information, which tells us exactly how the two 3-
spaces are positioned relative to each other in space-time, is made
invisible to other mathematicians who are ‘given’ just the two 3-
spaces, the Nows with their intrinsic geometries and matter
distributions, and set the task of nding the struts’ positions and
lengths. Will they succeed?

Despite niggling quali cations, the answer is yes. When you

unpack the mathematics of Einstein’s theory and see how it works
from the point of view of geometrodynamics, it appears to have been
tailor-made to solve this problem. This was shown in 1962 in a
remarkable, but not very widely known paper of just two pages by
Ralph Baierlein, David Sharp and John Wheeler (the rst two were
students of Wheeler at Princeton). I shall refer to these authors,
whose paper has the somewhat enigmatic title ‘Three-dimensional
geometry as a carrier of information about time’, as BSW. Initials can
become a menace, but the BSW paper is so central to my story that I
think they are warranted.
 It is the implications of the BSW paper that I discussed with Karel
in 1980. They can be quickly summarized. The basic problem that
BSW considered was what kind of information, and how much, must
be speci ed if a complete space-time is to be determined uniquely.
This is exactly analogous to the question that Poincaré asked in
connection with Newtonian dynamics, and then showed that the
information in three Nows was needed. As we have seen, a theory
will be Machian if two Nows are su cient. What BSW showed is that
the basic structure of general relativity meets this requirement.

In fact, the all-important Einstein equation that does the work is

precisely a statement that a best-matching condition between the
two 3-spaces does hold. The pairing of points established by it is
exactly the pairing established by the orthogonal struts. In fact, the
key geometrical property of space-times that satisfy Einstein’s
equations re ects an underlying principle of best matching built into
the foundations of the theory. I think that Einstein, with his deep
conviction that nature is supremely rational, would have been most
impressed had he lived to learn about it.

Equally beautiful and interesting is the condition that determines

‘how far apart in time’ the 3-spaces are. It is closely analogous to the
rule by which duration can be introduced as a distinguished
simpli er in Machian dynamics and the method by which the
astronomers introduced ephemeris time. There is, however, an
important di erence. In the simple Machian case, the distinguished
simpli er creates the same ‘time separation’ across the whole of
space. In Einstein’s geometrodynamics, the separation between the 3-
spaces varies from point to point, but the principle that determines it
is a generalization, now applied locally, of the principle that works
in the Newtonian case and explains how people can keep
appointments. This is why I say that, quite unbeknown to him,
Einstein put a theory of Mach’s principle and duration at the heart of
his theory.

I go further. The equivalence principle too is very largely

explained by best matching. To model the real universe, the 3-spaces
must have matter distributions within them. The analogue in two
dimensions is markings on bodies or paintings on curved surfaces.
When we go through the best-matching procedure, sticking pins into
Isolde, it is not only points on her skin that are matched to points on
Tristan, but also any tattoos or other decorative markings. All these
decorations – matter in the real universe – contribute with the
geometry in determining the best-matching position and the
distinguished simpli er that holds the 3-spaces apart and creates
proper time between them. When this idea is combined with the
relativity requirement, the equivalence principle comes out more or
less automatically.

Since the equivalence principle is essentially the condition that

the law of inertia holds in small regions of space-time, and all clocks
rely in one way or another on inertia, this is the ultimate explanation
of why it is relatively easy (nowadays at least) to build clocks that all
march in step. They all tick to the ephemeris time created by the
universe through the best matching that ts it together.

A SUMMARY AND THE DILEMMA

We have reached a crucial stage, and a summary is called for. In all

three forms of classical physics – in Newtonian theory, and in the
special and general theories of relativity – the most basic concept is a
framework of space and time. The objects in the world stand lower in
the hierarchy of being than the framework in which they move. We
have been exploring Leibniz’s idea that only things exist and that the
supposed framework of space and time is a derived concept, a
construction from the things.

If it is to succeed, the only possible candidates for the

fundamental ‘things’ from which the framework is to be constructed
are con gurations of the universe: Nows or ‘instants of time’. They
can exist in their own right: we do not have to presuppose a
framework in which they are embedded. In this view, the true arena
of the world is timeless and frameless – it is the collection of all
possible Nows. Dynamics has been interpreted as a rule that creates
histories, four-dimensional structures built up from the three-
dimensional Nows. The acid test for the timeless alternative is the
number of Nows needed in the exercise. If two su ce, perfect
Laplacian determinism holds sway in the classical world. It will have
a fully rational basis. There will be a reason for everything, found by
examination and comparison of any two neighbouring Nows that are
realized. There is perfection in such dynamics: every last piece of
structure in either Now plays its part and contributes, but nothing
more is needed.

In non-relativistic dynamics, Newton’s seemingly incontrovertible

evidence for a primordial framework and the secondary status of
things can be explained if the universe is Machian. Then the roles
will be reversed, things will come rst, and the local framework
de ned by inertial motion will be explained. However, without
access to the complete universe such a theory cannot be properly
tested. In any case, the Newtonian picture is now obsolete even if it
did clarify the issues. In general relativity the situation is much more
favourable and impressive, since the best matching is in nitely
re ned and its e ects permeate the entire universe. We can test for
them locally. Finding that they are satis ed at some point in space-
time is like nding a visiting card: ‘Ernst Mach was here’. The strong
evidence that Einstein’s equations do hold suggests that physics is
indeed timeless and frameless.

For all that, the manner in which space-time holds together as a

four-dimensional construct is most striking. It is highlighted by the
fact that there is no sense in which the Nows follow one another in a
unique sequence. This is what, in the Newtonian case, gives rise to
the beautifully simple image of history as a curve in Platonia. But in
special relativity and, much more strikingly, in general relativity
such a unique curve of history is lost. One and the same space-time
can be represented by many di erent curves in Platonia. Even
though no extra structure beyond what already exists in Platonia is
needed to construct space-time, the way it holds together convinces
most physicists that space-time (with the matter it contains) is the
only thing that should be regarded as truly existing. They are very
loath to accord fundamental status to 3-spaces in the way the
dynamical approaches of Dirac, ADM and BSW require. Even though
most of them grant that quantum theory will almost certainly modify
drastically the notion of space-time, they are still very anxious to
maintain the spirit of Minkowski’s great 1908 lecture. They are
convinced that space and time hang together, and they want to
preserve that unity at all costs. Within the purely classical theory, it
seems to me that the argument is nely balanced. Perhaps an
unconventional image of space-time will show how delicate this issue
– space-time as against dynamics – is.

Wagner’s opera Tristan und Isolde is widely regarded as a

highpoint of the Romantic movement in music. General relativity is
the ne plus ultra of dynamics. More explicitly, the way in which two
3-spaces are tted together in its dynamical core is like two lovers
seeking the closest possible embrace. This is the level of re nement
at work in the principles that create the fabric of space-time. It is
vastly more than just a four-dimensional block. Everywhere we look,
it tells the same great story but in countless variations, all
interwoven in a higher-dimensional tapestry. This is what Einstein
made out of Minkowski’s magical pack of cards. Look at space-time
one way, and we see Tristan and Isolde hanging, Chagall-like, in the
sky. Look another way, and we see Romeo and Juliet, yet another
way and it is Heloise and Abelard. All these pairs, each perfect in
themselves, are all made out of each other. They and their stories
stream through each other. They create a criss-cross fabric of space-
time (Figure 31).

Figure 31 Space-time as a tapestry of interwoven lovers. Given just

the ‘intrinsic structure’ of Tristan and Isolde, the BSW formalism
determines in principle all the points on Tristan that will be paired
with points on Isolde. The lengths of the struts (proper time between
matched points) are obtained as a by-product of the basic problem –
nding the ‘best position’ for the closest possible embrace. They are
therefore shown as dashes. The lengths of the struts are local
analogues of ephemeris time and, as they separate Tristan and Isolde,
 are simply the most transparent way of depicting the intrinsic
di erence between the two of them. The struts between the other
pairs of lovers are determined similarly. We can see how the
di erence that keeps Tristan apart from Isolde is actually part of the
body of Romeo (and Juliet). The struts between Romeo and Juliet
are drawn with short dashes because they have a space-like
separation. Einstein’s equations and the best-matching principle hold,
however space-time is sliced.

It stretches to the limit the notion of substance. For the body of

space-time, its fattening in time, is just the way we choose to hold
things apart so that the story unfolds simply. At least, it is in
Newtonian space-time. All the dynamics – what actually happens – is
in the horizontal placing. We pull the cards apart in a vertical
direction that we call time as a device for achieving simplicity of
representation. Time is the distinguished simpli er. The substance is
in the cards. They are the things; the rest is in our mind.

General relativity adds an amazing twist to this seemingly

de nitive theory of time. Considered alone, Tristan and Isolde are
substance, and the separation between them is just the measure of
their di erence. They cannot come together completely simply
because they are di erent. This di erence we call time. But what is
representation of di erence between Wagner’s lovers is part of the
very substance of Shakespeare’s lovers. Romeo and Juliet would not
be what they are if Tristan and Isolde were not held apart by their
di erence. The time that holds Tristan apart from Isolde is the body
of Romeo. This interstreaming of essence and di erence all in one
space-time is even more remarkable than Minkowski’s diagram
containing two rods each shorter than the other.

Several profound ideas are uni ed and taken to the extreme in

Figure 31: Einstein’s relativity of simultaneity, Minkowski’s fusion of
time with space, Poincaré’s idea that the relativity principle should
be realized through perfect Laplacian determinism, Poincaré’s idea
that duration is de ned so as to make the laws of nature take the
simplest form possible, and the astronomers’ realization that it is
measured by an average of everything that changes. Since best
matching in general relativity holds throughout the universe in all
conceivable directions, both time and space appear as the distillation
of all di erences everywhere in the universe. Machian relationships
are manifestly part of the deep structure of general relativity. But are
they the essential part?

If the world were purely classical, I think we would have to say

no, and that the unity Minkowski proclaimed so con dently is the
deepest truth of space-time. The 3-spaces out of which it can be built
up in so many di erent ways are knitted together by extraordinarily
taut interwoven bonds. This is where the deep dilemma lies. Four
decades of research by some of the best minds in the world have
failed to resolve it. On the one hand, dynamics presupposes – at the
foundation of things – three-dimensional entities. Knowing nothing
about general relativity, someone like Poincaré could easily have
outlined a form of dynamics that was maximally predictive, exible,
re ned and made no use of eternal space or time. Such dynamics,
constrained only by the idea that there are distinct things, must have
a certain general form. A whole family of theories can be created in
the same Machian mould.

On the other hand, a truly inspired genius might just have hit on
one further condition. Let dynamics do all those things with
whatever three-dimensional entities it may care to start from. But let
there be one supreme overarching principle, an even deeper unity.
All the three-dimensional things are to be, simultaneously with all
their dynamical properties, mere aspects of a higher four-
dimensional unity and symmetry.

If certain simplicity conditions are imposed, only one theory out

of the general family meets this condition. It is general relativity. It is
this deeper unity that creates the criss-cross fabric of space-time and
the great dilemma in the creation of quantum gravity. As we shall
see, quantum mechanics needs to deal with three-dimensional things.
The dynamical structure of general relativity suggests – and
su ciently strongly for Dirac to have made his ‘counter-
revolutionary’ remark – that this may be possible. Yet general
relativity sends ambivalent signals. Its dynamical structure says ‘Pull
me apart’, but the four-dimensional symmetry revealed by
Minkowski says ‘Leave me intact.’ Only a mighty supervening force
can shatter space-time.
Note added for this printing. New work summarized on p. 358 could
signi cantly change the situation discussed in this nal section of the
chapter. It suggests that the timeless Machian approach is capable of
leading to a complete derivation of general relativity and that it is not
necessary to presuppose ‘a higher tour-dimensional unity and
symmetry.’ Since this new work has only just been published and has
not yet been exposed to critical examination, I decided to leave the
original text intact. However, as already indicated in the note at the
end of the Preface, this new work does have the potential to
strengthen considerably the arguments for the nonexistance of time.
PART 4
Quantum Mechanics and Quantum
Cosmology

If the di erence between Newtonian and Einsteinian physics is

great, quantum mechanics seems separated from both by a chasm.
Most accounts of it, however, do not question the framework,
essentially absolute space and time, in which it was formulated.
They describe how very small systems – mostly atoms and molecules
– behave in an external framework. This may make quantum
mechanics appear more ba ing than need be.

If quantum mechanics is universally true and applies not only to

atoms and molecules but also to apples, the Moon, the stars and
ultimately the universe, then we ought to consider quantum
cosmology. What does the quantum mechanics of the universe look
like? It cannot be formulated in an external framework. Like
classical physics, quantum cosmology needs a description without a
framework. We shall see that many apparent di erences between
classical and quantum mechanics then appear in a di erent light.
What remains is one huge di erence. We shall soon begin to get to
grips with it.
CHAPTER 12
The Discovery of Quantum Mechanics

About a hundred years ago, a dualistic picture of the world took

shape. The electron had just been discovered, and it was believed
that two quite di erent kinds of thing existed: charged particles and
the electromagnetic eld. Particles were pictured as little billiard
balls, possessing always de nite positions and velocities, whereas
electromagnetic elds permeated space and behaved like waves.
Waves interfere, and recognition of this had led Thomas Young to
the wave theory of light (Figure 22).

By the end of the nineteenth century, the evidence for the wave
theory of light was very strong. However, it was precisely the failure
of light, as electromagnetic radiation, to behave in all respects in a
continuous wavelike manner that led rst Max Planck in 1900 and
then Einstein in 1905 to the revolutionary proposals that eventually
spawned quantum mechanics. A problem had arisen in the theory of
ovens, in which radiation is in thermal equilibrium with the oven
walls at some temperature. Boltzmann’s statistical methods, which
had worked so well for gases, suggested that this could not happen,
and that to heat an oven an in nite amount of energy would be
needed. The point is that radiation can have any wavelength, so
radiation with in nitely many di erent wavelengths should be
present in the oven. At the same time, the statistical arguments
suggested that, on average, the same nite amount of energy should
be associated with the radiation when in equilibrium. Therefore
there would be an in nite amount of energy in the oven – clearly an
impossibility. Baking ovens broke the laws of physics! Planck was
driven to assume that energy is transferred between the oven walls
and the radiation not continuously but in ‘lumps’, or ‘quanta’.

Accordingly, he introduced a new constant of nature, the

quantum of action, now called Planck’s constant, because the same
kind of quantity appears in the principle of least action. Until
Planck’s work, it had been universally assumed that all physical
quantities vary continuously. But in the quantum world, action is
always ‘quantized’: any action ever measured has one of the values
0, ½h, h, ¾h, 2h, .... Here h is Planck’s constant. (The fact that half-
integer values of h, i.e. ½h, ¾h, ..., can occur in nature was
established long after Planck’s original discovery. By then it was too
late to take half the original quantity as the basic unit.) The value of
h is tiny.

Most people are familiar with the speed of light, which goes
seven times round the world in a second or to the Moon and back in
two and a half seconds. The smallness of Planck’s constant is less
well known. Comparison with the number of atoms in a pea brings
it home. Angular momentum is an action and can be increased only
in ‘jerks’ that are multiples of h. Suppose we thread a pea on a string
30cm long and swing it in a circle once a second. Then the pea’s
action is about 1032 times h. As we saw, the atoms in a pea,
represented as dots a millimetre apart, would comfortably cover the
British Isles to a depth of a kilometre. The number 1032, represented
in the same way, would ll the Earth – not once but a hundred
times. Double the speed of rotation, and you will have put the same
number of action quanta into the pea’s angular momentum. It is
hardly surprising that you do not notice the individual ‘jerks’ of the
hs as they are added.

When people explain how our normal experiences give no

inkling of relativity and quantum mechanics, the great speed of light
and the tiny action quantum are often invoked. Relativity was
discovered so late because all normal speeds are so small compared
with light’s. Similarly, quantum mechanics was not discovered
earlier because all normal actions are huge compared with h. This is
true, but in a sense it is also misleading. For physicists at least,
relativity is completely comprehensible. The mismatch between the
relativistic world and its non-relativistic appearance to us is entirely
explained by the speed of light. In contrast, the mere smallness of
Planck’s constant does not fully explain the classical appearance of
the quantum world. There is a mystery. It is, I believe, intimately
tied up with the nature of time. But we must rst learn more about
the quantum.

Einstein went further than Planck in embracing discreteness. His

1905 paper, written several months before the relativity paper, is
extraordinarily prescient and a wonderful demonstration of his
ability to draw far-reaching conclusions from general principles. He
showed that in some respects radiation behaved as if it consisted of
particles. In a bold move, he then suggested that ‘the energy of a
beam of light emanating from a certain point is not distributed
continuously in an ever increasing volume but is made up of a nite
number of indivisible quanta of energy that are absorbed or emitted
only as wholes’. Einstein called the putative particles light quanta
(much later they were called photons). In a particularly beautiful
argument, Einstein showed that their energy E must be the radiation
frequency ω times Planck’s constant: E = hω. This has become one
of the most fundamental equations in physics, just as signi cant as
the famous E = mc2.

The idea of light quanta was very daring, since a great many
phenomena, above all the di raction, refraction, re ection and
dispersion of light, had all been perfectly explained during the
nineteenth century in terms of the wave hypothesis and associated
interference e ects. However, Einstein pointed out that the intensity
distributions measured in optical experiments were invariably
averages accumulated over nite times and could therefore be the
outcome of innumerable ‘hits’ of individual light quanta. Then
Maxwell’s theory would correctly describe only the averaged
distributions, not the behaviour of the individual quanta. Einstein
showed that other phenomena not belonging to the classical
successes of the wave theory could be explained better by the
quantum idea. He explained and predicted e ects in ovens, the
generation of cathode rays by ultraviolet radiation (the
photoelectric e ect), and photoluminescence, all of which de ed
classical explanation. It was for his quantum paper, not relativity,
that Einstein was awarded the 1921 Nobel Prize for Physics.

The great mystery was how light could consist of particles yet
exhibit wave behaviour. It was clear to Einstein that there must be
some statistical connection between the positions of the conjectured
light quanta and the continuous intensities of Maxwell’s theory.
Perhaps it could arise through signi cantly more complicated
classical wave equations that described particles as stable,
concentrated ‘knots’ of eld intensity. Maxwell’s equations would
then be only approximate manifestations of this deeper theory.
Throughout his life, Einstein hankered after an explanation of
quantum e ects through classical elds de ned in a space-time
framework. In this respect he was surprisingly conservative, and he
famously rejected the much simpler statistical interpretation
provided for his discoveries by the creation of quantum mechanics
in the 1920s.

In the following years, Einstein published several important

quantum papers, laying the foundations of a quantum theory of the
speci c heats of solids. However, the next major advance came in
1913 with Danish physicist Niels Bohr’s atomic model. It had long
been known that atoms emit radiation only at certain frequencies,
called lines because of their appearance in spectra. These spectral
lines, which had been arranged purely empirically in regular series,
were a great mystery. Everyone assumed that each line must be
generated by an oscillatory process of the same frequency in the
atoms, but no satisfactory model could be constructed.

Bohr found a quite di erent explanation. In a famous

experiment, the New Zealander Ernest Rutherford had recently
shown that the positive charge in atoms (balanced by the negative
charge of the electrons) was concentrated in a tiny nucleus. This
discovery was itself very surprising and is illustrated by a well-
known analogy. If the space of an atom – the region in which the
electrons move – is imagined as being the size of a cathedral, the
nucleus is the size of a ea. Bohr supposed that an atom was
something like the solar system, with the nucleus the ‘Sun’ and the
electrons ‘planets’.

However, he made a seemingly outrageous ad hoc assumption.

Using the electrostatic force for the known charges of the electron
and positive nucleus, he calculated the electron orbits in Newtonian
mechanics for the hydrogen atom, which has only one electron.
Each such orbit has a de nite angular momentum. Bohr suggested
that only orbits for which this angular momentum is some exact
multiple of Planck’s constant, i.e. 0, h, 2h, ..., can occur in nature.
These orbits also have de nite energies, now called energy levels. He
made the further equally outrageous conjecture that radiation in
spectral lines arises when an electron ‘jumps’ (for some unexplained
reason) from an orbit with higher energy to one with lower energy.
He suggested that the di erence £ of these energies is converted
into radiation with frequency ω, determined by the relation E = hco
found by Einstein for the ‘lump of energy’ associated with radiation
of frequency co. Thus, according to Bohr’s theory, an atom emits a
light quantum (photon) of a well-de ned energy by jumping from
one orbit to another.

For hydrogen atoms, it was easy to calculate the energy levels

and hence the frequencies of their radiation. Subject to certain
further conditions, Bohr’s theory had an immediate success. His
hotchpotch of Newtonian theory and strange quantum elements had
hardly explained the enigmatic spectral lines, but it did predict their
frequencies extraordinarily well, and there could be no doubting
that he had found at least some part of a great truth.

During the next decade the Bohr model was applied to more and
more atoms, often but not always with success. It was clearly ad
hoc. The need for an entirely new theory of atomic and optical
phenomena based on consistent quantum principles became ever
more transparent, and was keenly felt. Finally, in 1925/6 a
complete quantum mechanics was formulated – by Werner
Heisenberg in 1925 and Erwin Schrödinger in 1926 (and called,
respectively, matrix mechanics and wave mechanics). At rst, it
seemed that they had discovered two entirely di erent schemes that
miraculously gave the same results, but quite soon Schrödinger
established their equivalence.
 Heisenberg’s scheme, or picture, is based on abstract algebra and
is often regarded as giving a truer picture. In the form in which
quantum theory currently exists, it is more exible and general.
Unfortunately, it is rooted in abstract algebra, making it very
di cult to describe in intuitive terms. I shall therefore use the
Schrödinger picture. Luckily, this will not detract from what I want
to say. In fact, one of the main ideas I want to develop is that the
Schrödinger picture is actually more fundamental than the
Heisenberg picture, and is the only one that can be used to describe
the universe quantum-mechanically. Many physicists will be
sceptical about this, but perhaps this is because they study
phenomena in an environment and do not consider how local
physics might arise from the behaviour of the universe as a whole.

Schrödinger’s work developed out of yet another revolutionary

idea, put forward by the Frenchman Louis de Broglie in 1924. It
nally overthrew the dualistic picture of particles and elds that
had crystallized at the end of the nineteenth century. Einstein had
already shown that the electromagnetic eld possessed not only
wave but also particle attributes. De Broglie wondered whether,
since light can behave both as wave and particle, might not electrons
do the same? Together with its position, the most fundamental
property of a particle of mass m and velocity v is its momentum, mv.
De Broglie assumed that particles are invariably associated with
waves with wavelength λ related by Planck’s constant to their
momentum: λ = h/mv.
 He applied this idea to Bohr’s model. At each energy level, the
electron has a de nite momentum and hence a wavelength. We can
imagine moving round an orbit, watching the wave oscillations. In
general, if we start from a wave crest, the wave will not have
returned to a crest after one circuit. De Broglie showed that crest-to-
crest matching, or resonance, would happen only for the orbits with
quantized angular momentum that gured so prominently in the
Bohr model.

Although he had not, strictly, made any new discovery, his

proposal was suggestive. It restored a semblance of unity to the
world – both electrons and the electromagnetic eld exhibited wave
and particle properties. De Broglie’s thesis was sent to Einstein, who
was impressed and drew attention to its promise. Schrödinger got
the hint, and, as they say, the rest is history. During the winter of
1925/6 and the following months he created wave mechanics. This
will be the subject of the following chapters.

In 1927 de Broglie’s conjecture was brilliantly con rmed for

electrons rst in an experiment by the Englishman George Thomson,
and then in a particularly famous experiment by the Americans
Clinton Davisson and Lester Germer. These experiments paralleled
those made about a decade and a half earlier by the German
physicist Max von Laue, in which he had directed X-rays onto
crystals and observed very characteristic di raction patterns, from
which the structure of the crystals could be deduced. The patterns
were explained in terms of the interaction of waves with the regular
lattice of the atoms forming the crystals. They demonstrated
graphically the wave-like behaviour of the electromagnetic eld (X-
rays are, of course, electromagnetic waves, like light, but with much
higher frequency and shorter wavelength). In the 1927 experiments,
electrons were directed onto crystals, and di raction patterns
identical in nature to those produced by X-rays were seen. Thus, the
particle nature of electrons was observed long before their wave
nature was suspected. With light it was the other way round – wave
interference was observed a century before Einstein suspected that
light could have a particle aspect too.

Although it was now clear that both light and electrons exhibited
wave-particle duality, there were important di erences between
them. A brief description of the picture as it now appears will help.
All particles are associated with elds, and can be described as
excitations of those elds. To get some idea of what this means, we
can liken the particles to water waves, which are excitations of
undisturbed water. However, the analogy is only partial. The classic
example of a particle associated with a wave is the photon, which is
an excitation of the Maxwell eld. Fields and associated particles of
di erent kinds exist. There are elds described by a single number
at each point, called scalar elds, and vector elds, which are
described by three numbers. Scalar elds represent a simple
intensity, while the vector elds – such as Maxwell’s eld – are a
kind of ‘directed’ intensity. In general relativity we also encountered
tensors. Mathematically, scalar, vector and tensor elds belong to
one family and obey the same kind of rule under rotations of the
coordinate system. In particular, after one rotation they return to
the values they had before. However, in 1927 yet another
sensational quantum discovery was made, this one by Dirac. He
found a quite di erent family of elds, called spinor elds, which are
associated with electrons and protons (as well as many other
particles). In their case, one rotation of the coordinate system brings
them back to minus the value they had before, and two rotations are
needed to restore their original value. Dirac found spinors by trying
to make the newly discovered quantum principles compatible with
relativity, and achieved a spectacular success even though it was
subsequently found that his arguments were not totally compelling.
However, the main point is that electrons are associated with a
spinor eld, photons with a vector eld.

Both electrons and photons can, depending on the circumstances,

exhibit wave or particle behaviour. Otherwise they behave very
di erently. Many photons can be present simultaneously in the
same state (a state being a characteristic set of properties of
particles, such as position and direction of motion), but for electrons
this is impossible – there can be at most one in any given state. The
two kinds of particle have di erent statistical behaviour, so-called
Fermi-Dirac statistics for electrons and Bose-Einstein statistics for
photons. In fact, there are now known to be many di erent
particles, each with an associated eld. They satisfy either Fermi-
Dirac statistics, and are thus called fermions, or Bose-Einstein
statistics, in which case they are called bosons. In addition, nearly all
particles have an antiparticle. An antiparticle is identical to the
original particle in some respects, but opposite to it in others; in
particular, a particle and its antiparticle always have opposite
charges.

In many ways, the story of fundamental physics during the last

seventy years has been the discovery of particles and the
understanding of the manner in which they interact. All particles
that have so far been discovered – there is a whole ‘zoo’ of them –
are either spinor or vector particles. Ironically, particles
corresponding to the simplest scalar elds have not yet been
discovered, though it is con dently believed that they will be soon,
mainly on the grounds of indirect but rather persuasive theoretical
arguments. Currently, an immense amount of work is being done in
the attempt to unify the two broad categories of particles – fermions
and bosons – by means of an idea called supersymmetry. In the last
two or three years, there has been another great surge of excitement
in the eld of superstring theory. This combines the idea of
supersymmetry with the idea that the complete ‘zoo’ of particles
known at present are simply di erent manifestations of the
vibrations of a string, much as a violin string can vibrate at its
di erent harmonics. This is the dream of the theory of everything
(TOE). Some readers may be familiar with these ideas, originally
embodied in the acronym GUT – grand uni ed theory. This was the
aim of physicists who wished to describe within a single, uni ed
theoretical framework all the forces of nature except gravity (long
recognized as especially di cult to include). More recent, and more
ambitious since it aims to include gravity, is the quest for the big
TOE.

I am not going to make any attempt to discuss this work, nor will
I try to explain the connection between a particle and its associated
eld. If a theory of everything is found, it may well change the
framework of physics. We may nd ourselves in a quite new arena
and have to change our ideas about space and time yet again.
However, as of now I believe we can glimpse the outlines of an
arena large enough to accommodate not only the present ‘zoo’ but
also whatever entities some putative theory of everything will come
up with. The arena I have in mind is vast and timeless. I see it not as
a rival to the theory of everything, but as a general framework in
which such a theory can be formulated.

Now it is time to talk about the ideas that Schrödinger

introduced in the winter of 1925/6. That was when the door was
opened onto the vast arena.
CHAPTER 13
The Lesser Mysteries

INTRODUCTION

Most accounts of quantum mechanics concentrate on the simplest

situations – the behaviour of a single particle. That is already very
surprising. But the really mysterious properties come to light only in
composite systems of several particles, whose behaviour can become
ba ingly correlated. The situation is currently very exciting because
experimentalists are now able to study two widely separated but
strongly correlated particles. Their observations con rm quantum
mechanics brilliantly but stretch human intuition to the limit. How
can such things happen in space and time? And what unbelievable
scenarios will a quantum universe present?

I suspect that the present astonishment exists because most

quantum theoreticians do not think enough about quantum
cosmology. The rst issue is its arena. Quantum mechanics is
currently presented in a hybrid framework of two arenas at once.
One is an abstract mathematical construct known as Hilbert space,
but its elements are essentially de ned by absolute space and time,
which comprise the second arena. Quantum mechanics takes both for
granted. But they provide only a dubious foundation for quantum
cosmology. Clarity cannot be achieved until this hybrid state is
ended: the space-time framework must go. The answer to the
question of how such things can happen in space and time is that
they do not. They neither happen nor are they to be found in space
and time. But these things are, and their being is in Platonia, which
must replace the Hilbert space erected on the shaky foundations of
absolute space and time. That, at least, is my view.

My account of wave mechanics will aim to show that the demise

of space and time is inevitable. We shall rst see how a single
particle is described in space and time, and then see what happens
when we try to describe the universe. Space and time ‘evaporate’,
and we are left with the one true arena – timeless Platonia. In this
arena, quantum mechanics seems to me to take on a totally
transparent form. Whether we can believe in it is another matter.

THE WAVE FUNCTION

Every account of quantum mechanics includes the famous two-slit

experiment, and mine is no exception (Box 11). Di erences come
later. The two-slit experiment is to quantum mechanics what the
Michelson-Morley experiment is to relativity. The facts are simple,
and show that a radical change is unavoidable. The great beauty is
that the bare experimental facts directly suggest the need for and the
basic form of wave mechanics.

BOX 11 The Two-Slit Experiment

If a beam of photons or electrons, all with the same energy,
encounters a slit in a barrier and then impinges on a screen behind
it, individual localized ‘hits’ invariably occur (Figure 32). This is so
even if the beam has a very low density, so that at most one particle
at a time is passing through the system. This strongly suggests that
individual particles leave the beam generator, pass through the slit,
and strike the screen. The impacts have a characteristic distribution
over a region.

Now introduce a second identical slit in the barrier (Figure 33).

The interpretation of the rst experiment in terms of individual
particles yields an unambiguous prediction for what will happen. The
argument is as follows. All particles travel towards the barrier at
right angles to it, and can be assumed to be uniformly distributed in
space. The pattern behind a single slit is presumably created by the
interaction between the particles and the slit as they pass through it.
Entering the slit at di erent positions, the particles will have
di erent de ections and will thus strike the screen at di erent
points. When two slits are open, each should have an e ect identical
to that of the single slit, so the combined pattern should be simply
the sum of the e ects of two single slits.
Figure 32 The distribution of hits behind one slit.
Figure 33 The expected distribution of hits behind two slits.
 Figure 34 The actual distribution behind two slits.

Nothing remotely resembling this is observed. The hits are

distributed in the bands or fringes (Figure 34) characteristic of the
interference that led Young to the wave theory of light (Figure 22).
When, in the nineteenth century, it was believed that these fringes
are built up continuously, and not in individual ‘hits’, it seemed that
only a wave eld could produce them.

In the absence of a detailed theory, the pattern observed behind a

single slit can be explained equally well by particles or waves. But
the pattern behind two slits seems totally inexplicable on the
assumption of particles. For surely a particle can pass through only
one slit, and what it does then will depend solely on the properties of
that slit. It cannot ‘know’ whether the other slit is open or closed and
change its behaviour accordingly. Moreover, we can do similar
experiments with many slits of di erent shapes and sizes. Invariably,
wave theory correctly predicts the pattern produced on the screen.
As far as the total intensity pattern is concerned, there is no way to
explain it except by a wave theory.

Yet the patterns are always built up by individual ‘hits’. This is

extraordinarily strong evidence for particles. But if particles are
creating the patterns, they must somehow explore all the slits at
once. They must do what the very concept of a particle denies – be
everywhere at once. Moreover, this ability to be present at several
places at once gives rise to self-interference. Dirac put it memorably:
‘Each photon ... interferes only with itself.’ It is an important
observational fact that the possibility for interference to occur
continues until something like the screen forces the particle ‘to reveal
itself.

As long as the particle is not forced to make a choice, its

behaviour in quantum mechanics is described by what Schrödinger
called a wave function, which he denoted by the Greek letter psi, ψ,
and this has become traditional. Sometimes the capital is used: ψ. I
shall use this suitably grander capital in quantum cosmology,
keeping ψ for the things that happen in laboratories. The wave
function is like an intensity. If x is a point in space, ψ(x) is the value
of ψ at x. In general ψ has a di erent value for each x. The wave
function represents something completely new in physics. A further
novelty is that the wave function is not an ordinary number, as it
would be for a simple intensity, but a complex number (Non-
mathematicians should not get alarmed: it will be quite su cient to
think of a complex number as a pair of ordinary numbers. ‘Complex’
in this context means ‘composite’, not ‘complicated’.)

The status of the wave function is contentious to say the least.

Some claim it merely represents knowledge, while others want to
make it as physical as Faraday’s magnetic eld. As I see things, the
wave function is incorporeal (not some physical thing like a eld or
particle) and establishes a ranking of things. The real things are the
points of Platonia, the instants of time. Quantum cosmology – at
least in one embryonic form – will associate a value of ψ (note the
capital) with each point of Platonia. To emphasize how di erent the
wave function is, I like to think of it as some ‘mist’ that hangs or
hovers over Platonia, its intensity varying from point to point.

Actually, there are two mists because the wave function, being
complex, contains two numbers, which are its two components. I shall
call them the red mist and green mist, respectively. I shall also
introduce a third number, calling it the blue mist. The intensity of this
third mist is determined by the two primary components as the sum
of the squares of the red and green intensities. This is the mist
mentioned in the early chapters. Those in the know will recognize
the three mists as the real and imaginary parts of the wave function
and the square of its amplitude.
 The prominence that I give to these mists could be regarded by
most theoretical physicists (above all Dirac and Heisenberg, were
they still alive) as a one-sided, if not to say distorted and naive
picture of quantum mechanics. The mists (as opposed to things called
operators) are not particularly appropriate for talking about most
quantum experiments currently performed in laboratories. However,
the experiment I have in mind is not done in a laboratory. It is what
the universe does to the instants of time. For this experiment, the one
that really counts, I think the language of mists is appropriate. Those
who disagree might have second thoughts if they really started to
think of how inertial frames and duration arise. I come back to these
issues later.

I shall now give, in familiar space-time terms, a quantum-

mechanical account of the two-slit experiment (Figure 35). At an
initial time, the wave function associated with a particle is in a
‘cloud’ well to the left of the barrier. Inside the cloud, ψ is not zero.
Outside, it is zero. As time passes, this cloud moves to the right and,
in general, changes its shape. It evolves (in accordance with some
de nite rules). Typically, it ‘spreads’. At the barrier, some of the
cloud is re ected back to the left but some passes through the two
slits. Initially there are two separate clouds, but they spread rapidly
if the slits are narrow, and soon overlap. Characteristic wave
interference occurs. Thus, when the merged wave reaches the screen,
ψ is not the same everywhere, and fringes can form. In fact, the best
fringes are formed by a steady ‘stream’ of wave function, not a cloud.
INTERPRETING THE WAVE FUNCTION

The question now arises: where will the particle in Figure 35 be

observed? The answer, given already by the German physicist Max
Born in 1926, is that ψ determines, through the intensity of the blue
mist, the probability of where the particle will be observed. The blue
mist enables you to guess where the particle will ‘hit’ – twice the
intensity means twice the probability.

There are many mysteries in quantum mechanics, and the rst is

the probabilities. We can send identical clouds through the slits
many times. The fringe patterns are always exactly reproduced, but
the hits are distributed randomly. Only after many ‘runs’ does a
pattern of hits build up. The blue mist gives that pattern. Where its
intensity is high, many hits occur; where it is low, few; where it is
zero, none. Quantum mechanics determines these probabilities
perfectly, but says nothing about where the individual hits will
occur.
Figure 35 A ‘cloud’ of wave function ψ approaching two slits (at t =
0 and t = 1), passing through them, dividing into two (at t = 2),
spreading and overlapping (t = 3) and impinging on a screen (t =
4).

Einstein found this decidedly disturbing. He could not believe

that God reaches for a die every time physicists set up such an
experiment and force the particle to show up somewhere. For that is
what standard quantum mechanics implies – brute chance
determines outcomes. But there are even more puzzling things. It is
worth saying that quantum mechanics has a remarkably beautiful
and self-contained structure. Examined mathematically, it is a very
harmonious whole. It is hard to see how its structure could be
modi ed naturally to make it determine where individual hits occur,
especially when relativity is taken into account.

The next mystery is the collapse of the wave function. Just before
the particle hits the screen, its ψ can be spread out over a large
region. What happens to ψ when the particle is suddenly found
somewhere? The standard answer is that the wave is instantaneously
annihilated everywhere except where the particle is now known to
be.

If we want to determine what now happens, we have to start

afresh from a small, reduced cloud. The large cloud has been
‘collapsed’ and has no more relevance. This too provokes much
puzzling, especially for those (like Schrödinger in 1926) who would
wish to think of ψ as something real, a density of charge, say. How
can something real disappear instantaneously? Nothing in the
equations describes the collapse – it is simply postulated. Lawful
evolution, in accordance with the rules (equations) of quantum
mechanics, continues until an observation is made, but then the rules
are simply set aside. Quite di erent rules apply in measurements, as
they are called. (In quantum mechanics, the term ‘measurement’ is
used a very precise way. It means that some de nite arrangement of
instruments is used to establish the value of some physical property –
say the speed or position of a particle.) The abrupt and schizophrenic
change of the rules when measurements are made is a major part of
the notorious measurement problem. There are rules for evolution and
rules for measurement – and they are even more di erent than chalk
and cheese. Nevertheless, both are excellently con rmed, though we
have to be careful when saying that the collapse is instantaneous,
and even when it occurs.

STATES WITHIN STATES

Just as mysterious as the rule change when measurements are made

is a certain mutual exclusivity about the kinds of measurement that
can be made. So far, I have talked only about particle positions.
However, we can also measure other quantities – for example, a
particle’s energy, momentum or angular momentum. It is particularly
fascinating that information about them all is coded at once in ψ.
This is another big di erence from classical mechanics.

Imagine a perfect sinusoidal wave that extends with constant

wavelength from in nity to in nity. For the moment, suppose that it
is ‘frozen’, like the wave patterns you see in damp sand at low tide.
Let me call this the red wave, because it represents the red mist. Now
imagine another identical though green wave, shifted forward by a
quarter of a wavelength relative to the red wave (Figure 36). Then
the red peaks lie exactly at the green wave’s nodes, where the green
wave has zero intensity. As time passes, the red and green waves
move to the right, maintaining always their special relative
positioning. A wave function in this special form represents a particle
that has a de nite momentum: if it hit something, it would transmit
a de nite impulse to it. A particle with the opposite momentum is
represented similarly, but travels in the opposite direction and has
the green peaks a quarter of the wavelength behind the red peaks.
According to the quantum rules, the particle has a de nite
momentum because its ψ has a de nite wavelength and is perfectly
sinusoidal. Such wave functions give the best interference e ects in
two-slit experiments. They are called momentum eigenstates. (The
German word eigen means ‘proper’ or ‘characteristic’.)

Figure 36 The wave function of a particle with a de nite

momentum.

The striking thing about this situation is that the probability for
the position of the particle, given by the sum of the squares of the
red and green intensities, is completely uniform in space. The reason
is that for two sinusoidal waves displaced by a quarter of a
wavelength, this sum is always 1 if the wave’s amplitude (its height
at the peaks) is 1. This is a consequence of the well-known
trigonometric relation sin2A + cos2A = 1, which itself is just
another expression of Pythagoras’ theorem. Thus, for a particle in
this state, we have absolutely no information about its position, but
we do know that it has a de nite momentum.

So far we have considered waves of only one wavelength.

However, we can add waves of di erent wavelengths. Whenever
waves are added, they interfere, enhancing each other here and
cancelling out there. By playing around with waves of di erent
wavelengths we can make a huge variety of patterns (Figure 37 is an
example). In fact the French mathematician Joseph Fourier (one of
Napoleon’s generals) showed that more or less any pattern can be
made by adding, or superposing, sinusoidal waves appropriately. Any
wave pattern created in this way and concentrated in a relatively
small ‘cloud’ is called a wave packet. The same pattern can be made
by superpositions of quite di erent kinds. The primary meaning of ψ
is that its value at x determines, through the squares of its two
intensities, the probability that the particle will be ‘found’ at x. Now,
a ‘cloud’ could be so narrow that it becomes a ‘spike’ at some value
of x. The particle can then be at only one place – at the spike. Such a
wave function is called a position eigenstate.
Figure 37 Superposition of the two waves at the top gives rise to the
very di erent wave pattern at the bottom.

Thus, the same wave pattern can be regarded either as a

superposition of plane waves or as a superposition of many such
spikes added together with di erent coe cients (Figure 38). Any
wave function is a superposition of either position or momentum
eigenstates. There is a duality at the heart of the mathematics. What
is remarkable – and constitutes the essential core of quantum
mechanics in the standard form it was given by Dirac – is that it
perfectly re ects a similar duality found in nature. This is where the
measurement problem becomes even more puzzling. We need to
consider the ‘o cial line’, known as the Copenhagen interpretation
because it was established by Heisenberg and Bohr at the latter’s
institute in Copenhagen shortly after the creation of quantum
mechanics.

Figure 38 Two ‘spiky’ wave patterns (thin curves) are superposed to

make a much smoother pattern (heavy curve).

THE COPENHAGEN INTERPRETATION

The wave function of a particle is assumed to be a maximal

representation of its physical state. It codes everything that can ever
be deduced about the particle at an instant. Using it, we can predict
the outcomes of experiments performed on the particle. There are
two cardinal facts about these predictions. First, they are
probabilistic. Only if, for example, the particle is in a momentum
eigenstate (represented by the two special plane waves described
above) will measurement of the momentum con rm that the particle
has the corresponding momentum. If it is in a superposition of
momentum eigenstates, then any one of the momenta in the
superposition may be found as a result of the measurement. The
probabilities for them are determined by the strengths with which
the corresponding momentum eigenstates are represented in the
superposition.

It is a basic Copenhagen tenet that the probabilistic statements

re ect a fundamental property of nature, not simply our ignorance. It
is not that before the measurement the particle does have a de nite
momentum and we simply do not know it. Instead, all momenta in
the superposition are present as potentialities, and measurement
forces one of them to be actualized. This is justi ed by a simple and
persuasive fact. If we do not perform measurement but instead allow
ψ to evolve, and only later make some measurement, then the things
observed later (like the two-slit fringes) are impossible to explain
unless all states were present initially and throughout the subsequent
evolution. Outcomes in quantum mechanics are determined by
chance at the most fundamental level. This is the scenario of the
dice-playing God that so disturbed Einstein.
 If anything, the second cardinal fact disturbed him even more.
There seems to be a thoroughgoing inde niteness of nature even
more radical than the probabilistic uncertainties. As we have seen,
one and the same state can be regarded as a superposition of either
momentum or position eigenstates. It is the way this mathematics
translates into physics that is startling. The experimentalist has
complete freedom to choose what is to be measured: position or
momentum. Both are present simultaneously as potentialities in the
wave function. The experimentalist merely has to choose between
set-ups designed to measure position or momentum. Once the choice
is made, outcomes can then be predicted – and one outcome is
actualized when the measurement is made. In fact, the inde niteness
is even greater since other quantities, or observables as they are
called, such as energy and angular momentum, are also present as
potentialities in ψ.

Only one experiment can be made – for position or momentum,

say, but not both. Every measurement ‘collapses’ the wave function.
After the collapse, the wave function, which could have been used to
predict outcomes of alternative measurements, has been changed
irrevocably: there is no going back to the experiment we opted not to
perform. It is a very singular business. Whatever observable we
decide to measure, we get a de nite result. But the observable that is
made de nite depends on our whim. The many people who, like
Einstein, believe in a real and de nite world nd this immensely
disconcerting. What is out there in the world seems to depend on
mere thoughts that come into our mind. Most commentators believe
that this radical inde niteness – the possibility to actualize either
position or momentum but not both – is the most characteristic
di erence between classical and quantum physics. In classical
physics, position and momentum are equally real, and they are also
perfectly de nite.

The fact that in quantum mechanics one can choose to measure

one but not both of two quantities was called complementarity by
Bohr. Pairs of quantities for which it holds are said to be
complementary.

HEINSENBERG’S UNCERTAINTY PRINCIPLE

Heisenberg’s famous uncertainty relation gives quantitative

expression to complementarity for position and momentum. De
Broglie’s relation λ=h/mv=h/p determines the wavelength of a
particle of momentum p=mv, where m is its mass and v its velocity.
Now, it follows from Fourier’s work on the superposing of waves that
a wave packet restricted to a small spatial region contains many
waves in a broad spread of wavelengths. To narrow down the spatial
positions q, it is necessary to broaden the range of momenta p.
Conversely, to get a nearly de nite p, we must accept a wide range
of positions q.

Mathematically, we can in fact construct wave packets in which

the positions are restricted to a small range, from q to q + Δq, and
the momenta to a correspondingly small range, from p to p + Δp.
Any attempt to make Δq smaller necessarily makes Δp larger, and
vice versa. Heisenberg’s great insight – his uncertainty relation – was
the physics counterpart of this mathematics. There is always a
minimum uncertainty: the product ΔqΔp is always greater than or, at
best, equal to Planck’s constant h divided by 4π. If you try to pin
down the position, the momentum becomes more uncertain, and vice
versa. This is the uncertainty relation. Moreover, a wave packet of
minimum dimensions will in general spread: the uncertainty in the
position will increase. This is what in quantum mechanics is known
as the ‘spreading of wave packets’.

Since Planck’s constant h is so small, an object like a pea or even

a grain of sand can e ectively have both a de nite position and
de nite momentum, and the spreading of its wave packet takes place
extremely slowly. This explains why all the macroscopic objects we
see around us can seem to have de nite positions. But though the
quantum laws allow objects to be localized in space and to have
e ectively de nite velocities, there is no apparent reason in the
equations why this should habitually be so. They also allow –
encourage, one might even say – a pea’s wave packet to be localized
in two or more places at once. Nothing forces ψ to ‘localize’ around a
single point. Einstein used to look at the Moon and ask why we do
not see two. It is a real problem. Quantum measurements on
microscopic systems are actually designed to create situations in
which a macroscopic instrument pointer is, according to the
equations, in many places at once. Yet we always see it at only one.

THE ENIGMATIC GEM

We shall come back to this mystery, which is one aspect of another:
Hilbert space and transformation theory. If you nd this section a bit
abstract, don’t worry; it is helpful at least to mention these things. In
quantum mechanics, position and momentum (and other
observables) play a role rather like coordinates – ‘grid lines’ – on a
map. Just as in relativity the coordinates on space-time can be
‘painted’ in di erent ways, so too in quantum mechanics there are
many mathematically equivalent ways of arranging the coordinates.
This was one of Dirac’s rst great insights, and it led to his
transformation theory.

According to this, the state of a quantum system is some de nite

but abstract thing in an equally abstract Hilbert space. The one state
can, so to speak, be looked at from di erent points of view. A Cubist
painting might give you a avour of the idea. In relativity, di erent
coordinate systems on space-time correspond to di erent
decompositions into space and time. In quantum mechanics, the
di erent coordinate systems, or bases, are equally startling in their
physical signi cance. They determine what will happen if di erent
kinds of measurement, say of position or of momentum, are made on
the system by instruments that are external to the system. The state
in Hilbert space is an enigmatic gem that presents a di erent aspect
on all the innumerable sides from which it can be examined. As
Leibniz would say, it is a city multiplied in perspective. Dirac was
entranced, and spoke of the ‘darling transformation theory’. He knew
he had seen into the structure of things. What he saw was some real
but abstract thing not at all amenable to easy visualization. But the
multiplication of viewpoints and the mathematical freedom it
furnished delighted him.

In The Principles of Quantum Mechanics, a veritable bible for

quantum mechanicians, Dirac says that in classical physics ‘one could
form a mental picture in space and time of the whole scheme’ but ‘It
has become increasingly evident that Nature works on a di erent
plan. Her fundamental laws do not govern the world as it appears in
our mental picture in any very direct way ...’. I have quoted these
words because, with all respect to the greatness of his discoveries
and the clarity of his thought, Dirac may have gone too far with his
dismissal of simple mental pictures. But what kind of mental pictures
are we talking about here? Dirac was reacting against Einstein and
Schrödinger, who longed to form mental pictures in space and time.
Schrödinger, for example, had commented in his second paper on
wave mechanics that some people

had questioned whether the things that happen in the atom

could be incorporated in the space-time form of thought at
all. Philosophically, I would regard a nal decision in this
sense as the same as complete capitulation. For we cannot
actually change the forms of thought, and what we cannot
understand within them cannot be understood at all. There
are such things – but I do not think atomic structure is one of
them.
 This appeal to ineluctable forms of thought, an echo of the
eighteenth-century German philosopher Immanuel Kant’s belief that
space and time are an a priori framework without which we cannot
even begin to form a picture of the world, is doubly ironic.
Schrödinger was strongly drawn to the holistic notions of eastern
mysticism but would not accept them in his own theory, where they
seem inescapable. Even more ironically, he himself changed the
forms of thought. He created new mental images just as transparent
as the space and time to which he and Einstein clung for dear life.
That is the topic of the next chapter.
CHAPTER14
The Greater Mysteries

SCHRÖDINGER’S VAST ARENA

The true heart of quantum mechanics and the way to quantum

cosmology is the way in which it describes composite systems – that
is, systems consisting of several particles. It is an exciting, indeed
extraordinary story, though it is seldom well told. When Schrödinger
discovered wave mechanics, he said it could be generalized and
‘touches very deeply the true essence [wahre Wesen] of the quantum
prescriptions’. But it was not just the Bohr quantization prescriptions
that came into focus: at stake here are the rules of creation. A bold
claim, but one I hope to justify as the book goes on. First, we have to
see how Schrödinger opened the door onto a vast new arena.

The central concept of this book is Platonia. It is a relative

con guration space. The new arena that Schrödinger introduced is
something similar, a con guration space (without the ‘relative’). The
notion is easily explained. Each possible relative arrangement of
three particles is a triangle and corresponds to a single point in the
three-dimensional Triangle Land. But now imagine the three particles
located in absolute space. Besides the triangle they form, which is
speci ed by three numbers (the lengths of its sides), we now have to
consider the location of its centre of mass in absolute space, which
requires three more numbers, and also its orientation in absolute
space, which also requires three more numbers. Location in Triangle
Land needs three numbers, in absolute space it needs nine. Just as
each triangle corresponds to one point in three-dimensional Triangle
Land, the triangle and its location in absolute space correspond to
one point in a nine-dimensional con guration space. The tetrahedron
formed by four particles corresponds to one point in six-dimensional
Tetrahedron Land and one point in the corresponding twelve-
dimensional con guration space. For any Platonia corresponding to
the relative arrangements of a certain number of particles, the
matching con guration space has six extra dimensions. Schrödinger
called such a space a Q, and I shall follow his example. Such a Q is a
‘hybrid Platonia’, since it contains both absolute and relative
elements. This hybrid nature is very signi cant, as will become
apparent.

The most important thing about Schrödinger’s wave mechanics is

that it is formulated not in space and time, but in a suitably chosen Q
and time. This is not apparent for a single particle, for which the
con guration space is ordinary space. Since most accounts of
quantum mechanics consider only the behaviour of a single particle,
many people are unaware that the wave function is de ned on
con guration space. That is where ψ lives. It makes a huge
di erence.
 An illustration using a plastic ball-and-strut model of molecules
may help to bring this home. Imagine that you are holding such a
model in some de nite position in a room, which can represent
absolute space. There are three digital displays – I shall call them ψ
meters – that show red, green and blue numbers on the wall. These
numbers give the intensities of the three ‘mists’ represented by ψ for
the system at the time considered. Suppose you take just one ball,
representing one particle of the system, and detach it from the
model. Keeping all the other balls xed, you can move the one ball
around and, courtesy of the ψ meters, see how ψ changes. As you
move in each direction in space, each ψ value will change. For each
point of space you can nd the value of ψ. The blue ψ meter will
always tell you the positions for which the probability is high or low.
Suppose you do this and then return the ball to its original place.

Now move a second ball to a slightly di erent position, and leave

it there. The ψ meters will change to new values. Once again, explore
space with the rst ball, watching the ψ meters. The values of ψ will
be (in general) quite di erent. The ψ values on the displays embody
information. The amount is staggering. For every single position in
space to which you move any one of the other balls, you get a
complete new set of values in space for the ball chosen as the
‘explorer’. And any ball can be the explorer. Each explorer will have
its own distinctive three-dimensional patterns of ψ for every
conceivable set of positions of the others.
 Now, what is a molecule? When Richard Dawkins described the
haemoglobin molecule and its six thousand million million million
perfect copies in our body, he said that in its intricate thornbush
structure there is ‘not a twig nor a twist out of place’. That is in a
molecule containing perhaps twenty thousand atoms. But molecules
are even more remarkable than that. The twig and the twist are
averaged structures corresponding to the most probable
con guration in which the molecule will be found. In the
Schrödinger picture, the molecule is not just one structure but a huge
collection of potentially present structures, each with its own
probability.

In fact, the complete structure of complicated protein molecules

like haemoglobin cannot be understood solely on the basis of wave
mechanics. This is because of the way they are put together from
amino acid units. But for simpler molecules, which may still contain
many particles, you could (in imagination at least) do what I have
just described for the ball-and-strut model. Start with one of the
model con gurations shown in chemistry textbooks, and look at the
ψ meters, especially the blue one. It will give a high reading. Around
that highly probable structure are other similar structures, all with a
high – but not quite so high – blue intensity. Individual units of the
structure – simpler forms of Dawkins’s ‘twigs’ – could be moved as a
whole, say by twisting them, from the most probable con guration,
and the blue intensity would drop. It would also drop if one atom of
the few dozen within the twig were moved from the most-favoured
position. The molecule is not just the most probable con guration. It
is all possible con gurations with their ψ values, held in balance by
the laws of wave mechanics. The existence and most-favoured shape
of molecules can be understood in no other way.

Contrary to the impression given in many books, quantum

mechanics is not about particles in space: it is about systems being in
con gurations – at ‘points’ in a Q, or ‘hybrid Platonia’. That is
something quite di erent from individual probabilities for individual
particles being at di erent points of ordinary space. Each ‘point’ is a
whole con guration – a ‘universe’. The arena formed by the ‘points’
is unimaginably large. And classical physics puts the system at just
one point in the arena. The wave function, in contrast, is in principle
everywhere.

This is what I mean by saying that Schrödinger opened the door

onto a vast new arena. Compared with Schrödinger’s vistas, grander
than any Wagnerian entrance into Valhalla, the Heisenberg
uncertainty relation for a single particle captures little of quantum
mechanics. All revolutions in physics pale into insigni cance beside
Schrödinger’s step into the con guration space Q. Not that he did it
happily.

CORRELATIONS AND ENTANGLEMENT

It is not possible to observe the extraordinary quantum arena

directly. Some people do not believe it exists at all. To a large degree
it has been deduced, or surmised, from phenomena observed in
systems of a few particles. Getting clear, direct evidence for the
quantum behaviour of single particles was di cult. It was long after
Dirac made his memorable remark about each photon interfering
with itself that the development of sources which release individual
particles with long time intervals between releases con rmed the
build-up of interference patterns in individual ‘hits’. In the last two
decades, it has become possible to create in the laboratory pure
quantum states of two particles, whose Q therefore has six
dimensions. The quantum predictions, all veri ed, are not easy to
explain in many words, let alone a few, and a serious attempt to do
so would take me too far from my main story. The simplest possible
illustration is given by two particles moving on a single line; each
has a one-dimensional Q and together they have a two-dimensional
con guration space (Figure 39).

As for a single particle, the maximally informative description of

a quantum system at any instant t is speci ed by a complex wave
function ψ which, in principle, has a di erent value at each point of
the con guration space. As t changes, ψ changes. All information that
can be known about the system at t is encoded in ψ at f, and consists
of predictions that can be made about it. Many di erent kinds of
prediction can be made, but they are often mutually exclusive. In a
very essential way, the predictions refer to the system, not its parts.

Let us start with position predictions. Just as we did for a single

particle, we can form from ψ the sum of the squares of its intensities,
nding the intensity of the ‘blue mist’ (Figure 40). This gives the
relative probability that the system will be found at the
corresponding point in Q if an appropriate measurement is made.
The important thing is that a single point in Q corresponds to
positions of both particles. Anyone who has not understood this has
not understood quantum mechanics. It is this fact, coupled with
complementarity, that leads to the most startling quantum
phenomena.

Figure 39 The two-dimensional con guration space Q of two

particles on one line. The line is shown in multiple copies on the left.
Nine di erent con gurations of the two particles on it are shown.
The positions of particles 1 and 2 are indicated by the black and
white triangles, respectively. The axes of Q on the right show the
distances of particle 1 (horizontal axis) and particle 2 (vertical axis)
from the left-hand end of the line. The points on the 45° diagonal in
Q correspond to con gurations for which the two particles coincide
(points 1, 4 and 8). You might like to check how the nine
con gurations on the left are represented by the nine corresponding
points on the right.

In Chapter 3 we imagined tipping triangles out of a bag. That

exercise was presented because it mimics one of the ways in which
we can interpret a quantum state. Imagine now that the blue mist
has the distribution shown in Figure 40. To avoid problems with
in nite numbers of con gurations, we divide up Q by a grid of cells
su ciently ne that ψ hardly changes within any one of them
(Figure 40, on the right). The intensity of the blue mist at the central
point of each cell then gives the relative probability of the nearly
identical con gurations in that cell. On a piece of cardboard, let us
depict one of these con gurations (as shown on the left in Figures 39
and 40). This will serve as the representative of all the con gurations
of that cell. For the grid in the gure with 100 cells, there are 100
relative probabilities whose sum should be conveniently large, say a
million. Then we shall not distort things seriously by replacing exact
relative probabilities like 127.8543... by the rounded-up integer 128.

We now imagine putting into a bag the number of copies of each

representative con guration equal to its rounded probability, 128 for
example. In quantum mechanics, performing a measurement to
determine the positions of both particles is like drawing at random
one piece of cardboard from the bag. We get some de nite
con guration. In the process, we destroy the wave function and
replace it by one entirely concentrated around the con guration we
have found. If we recreate the original wave function, by repeating
the operations that we used to set it up, and repeat the experiment
millions of times, then the relative frequencies with which the
various con gurations are ‘drawn from the bag’ will match,
statistically, the calculated relative probabilities.

Figure 40 Like Figure 39, this shows nine di erent con gurations of
two particles (black and white triangles) on a line and the points
corresponding to them in the con guration space Q (on which a grid
has been drawn). A possible distribution of the intensity of the blue
probability mist is shown as the height of a surface over Q in the top
part of the gure (you are seeing the surface in perspective from
above, and rotated). In the state of the system shown here, the
probabilities for con gurations 4, 6 and 9 are high, while 5 has a
very low probability.
Figure 41 (a) The e ect of measuring, for the probability density of
Figure 40, the position of the particle represented by the horizontal
axis, and nding that it lies in the interval on which the vertical strip
stands. All the wave function outside the strip is instantaneously
collapsed.
 This is only the start. We can select from a menu of di erent
kinds of measurement. For example, we can opt to nd the position
of only one particle, which has remarkable implications for what we
can say about the other one. Suppose rst that we measure the
position of just one of the particles. According to the quantum rules,
this instantaneously collapses the wave function from its original
two-dimensional ‘cloud’ to a one-dimensional pro le (Figure 41).
The point is that we now know the position of one particle to within
some small error, so none of the wave function outside the narrow
strip is relevant any longer. It is annihilated. If the particle whose
position is measured is represented along the horizontal axis, only a
vertical strip of ψ survives (Figure 41(a)); if the position of the other
particle is measured, only a horizontal strip survives (Figure 41(b)).
 Figure 41 (b) The same for a position measurement of the other
particle.

Either pro le then gives conditional information. If we know

where one particle is, the possible positions of the other are
restricted to a narrow strip. The relative probabilities for the position
of the second particle are determined by the values of ψ within the
strip. Provided we know the original wave function, acquired
knowledge about one particle sharpens our knowledge –
instantaneously – about the other. This is the place to explain
entangled states, or quantum inseparability (Box 12).

BOX 12 Entangled States

Figure 42 again shows our two-particle Q and two di erent quantum

states. In the unentangled state at the bottom, all the horizontal ψ
pro les are identical, and so are all the vertical pro les (only their
shapes count). Such a wave function is said to be unentangled
because if we gain information about particle 1 (black triangles) –
that it is at some de nite position – we do not gain any new
information about particle 2 (white triangles). This is because all the
horizontal pro les and all the vertical pro les are identical: they give
identical relative probabilities. This is shown by two pro les that
result from an exact position measurement of particle 1.
Measurement on particle 1 leads to no new information about
particle 2.
 Figure 42 Entangled (top) and unentangled (bottom) states for two
particles (black and white triangles).

Much more interesting is the entangled state at the top, for which
the horizontal and vertical pro les are not identical. Figure 42 shows
two pro les that result from exact position measurement of particle
1. They give very di erent probability distributions for particle 2: the
gain in information about particle 2 is considerable. This is typical of
quantum mechanics, since virtually all wave functions are entangled
to a greater or lesser extent.

A particular feature of entangled states should be noted. Particles

normally interact (a ect each other) when they are close to each
other. For the two-particle Q in Figure 39, the particles coincide on
the diagonal line, and the region in which the particles are close to
each other and can interact strongly is therefore a narrow strip
around that line. However, the wave function of an entangled state
may be located completely outside this region – the particles may be
very far apart when one of them is observed. Yet the other particle is
apparently immediately a ected. It can jump to one or other of two
hugely di erent possibilities. Moreover, if the ‘ridge’ in the top part
of Figure 42 is made thinner and thinner, shrinking to a line, then
position determination of one particle immediately determines the
position of the other to perfect accuracy. Such situations are not easy
to engineer for position measurements (and in general will not
persist because of wave-packet spreading), but there are analogous
situations for momentum and angular-momentum measurements that
are easy to set up.

The facts discussed in Box 12 are immensely puzzling if we wish

to nd a physical and causal mechanism to explain how
measurement on one particle can have an immediate e ect on a
distant particle. As I have already explained, innumerable
interference phenomena indicate that, in some sense, the particles
are, before any measurement is made, simultaneously present
wherever ψ extends. Since there is no restriction on the distance
between the particles, any causal e ect on the second particle after
the rst has been observed would have to be transmitted
instantaneously. However, relativity theory is supposed to rule out
all causal e ects that travel faster than the speed of light. Moreover,
in the mid-1980s Alain Aspect in Paris performed some very famous
experiments in which such wave-function collapses were tested, and
the predictions of quantum mechanics con rmed with great
accuracy. The experiments were so arranged that any physical e ect
would have had to be transmitted faster than the speed of light to
bring about the collapse.

The situation is actually delicate and intriguing. Relativity

absolutely prohibits the transmission of information faster than light.
But, curiously, wave-function collapse does not transmit information.
When information about particle 1 has been obtained by an
experimentalist, he or she will know immediately what a distant
experimentalist can learn about particle 2. But there is no way such
information can be transmitted faster than light. There is no con ict
with the rules of relativity, though many physicists are concerned
that its ‘spirit’ is violated.

So far, we have considered only position measurements on a two-

particle system. But we can also consider many other measurements,
of momentum, for example. Given ψ in Q, we directly obtain
predictions for positions. But Dirac’s transformation theory enables
us to pass to the complementary momentum space, which gives
direct predictions for momentum measurements. If the wave function
is tightly entangled with respect to momentum, measuring the
momentum of particle 1 would immediately tell us the momentum of
particle 2. And this despite the fact that before any measurements
are made there is considerable uncertainty about the momenta of the
particles. However, what is certain is that they are entangled, or
correlated. This brings us to the EPR paradox.

THE EPR PARADOX

The nub of the Einstein-Podolsky-Rosen (EPR) paradox, formulated

in 1935 by Einstein and collaborators Boris Podolsky and Nathan
Rosen, is that two particles can be in a state in which they are
perfectly correlated (entangled) as regards both their position and
their momentum. The actual example of such a state that EPR found
is rather unrealistic, but in 1952 David Bohm, an American
theoretical physicist who later worked in London for many years,
proposed a much more readily realized state using spin, the intrinsic
angular momentum associated with quantum particles. Alain Aspect
performed his experiments on such a system. What puzzled EPR
about their state was that if the position of one particle was
measured, the position of the other particle could be immediately
established with certainty because of the perfect correlation. Since
the second particle, being far away, could not be physically a ected
by the measurement, but it was known for certain where it would be
found, EPR concluded that it must have had this de nite property
before the measurement on the rst particle.

But, it could just as well have been decided to measure

momentum. The measurement of one momentum will then
instantaneously determine the other momentum with certainty. By
the same argument as before, the particle must have possessed that
momentum before the measurement on the rst particle. Finally, the
choice between momentum or position measurement is a matter of
our whim, about which the second particle can know nothing. The
only conclusion to draw is that the second particle must have
possessed de nite position and momentum before any measurements
were made at all. However, according to the fundamental rules of
quantum mechanics, as exempli ed in the Heisenberg uncertainty
principle, a quantum particle cannot possess de nite momentum and
position simultaneously. EPR concluded there must be something
wrong – quantum mechanics must be incomplete.

Niels Bohr actually answered EPR quite easily, though not to

everyone’s satisfaction. His essential point was that quantum
mechanics predicts results made in a de nite experimental context.
We must not think that the two-particle system exists in its own
right, with de nite properties and independent of the rest of the
world. To make position or momentum measurements, we must set
up di erent instruments in the laboratory. Then the total system,
consisting of the quantum system and the measuring system, is
di erent in the two cases. Nature arranges for things to come out
di erently in the two cases. Nature is holistic: it is not for us to
dictate what Nature is or does. Quantum mechanics is merely a set of
rules that brings order into our observations. Einstein never found an
answer to this extreme operationalism of Bohr, and remained deeply
dissatis ed.

I feel sure that Bohr got closer to the truth than Einstein.
However, Bohr too adopted a stance that I believe is ultimately
untenable. He insisted that it was wrong to attempt to describe the
instruments used in quantum experiments within the framework of
quantum theory. The classical world of instruments, space and time
must be presupposed if we are ever to talk about quantum
experiments and communicate meaningfully with one another. Just
as Schrödinger made his Kantian appeal to space and time as
necessary forms of thought, Bohr made an equally Kantian appeal to
macroscopic objects that behave classically. Without them, he
argued, scienti c discourse would be impossible. He is right in that,
but in the nal chapters I shall argue that it may be possible to
achieve a quantum understanding of macroscopic instruments and
their interaction with microscopic systems. Here it will help to
consider why Einstein thought the way he did.

Referring to their demonstration that distant measurement on the

rst system, ‘which does not disturb the second in any way’,
nevertheless seems to a ect it drastically, EPR commented that ‘No
reasonable de nition of reality could be expected to permit this.’
These words show what is at stake – it is the atomistic picture of
reality. Despite the sophistication of all his work, in both relativity
and quantum mechanics, Einstein retained a naive atomistic
philosophy. There are space and time, and distinct autonomous
things moving in them. This is the picture of the world that underlies
the EPR analysis. In 1949 Einstein said he believed in a ‘world of
things existing as real objects’. This is his creed in seven words. But
what are ‘real objects’?

To look at this question, we rst accept that distinct identi able

particles can exist. Imagine three of them. There are two possible
realities. In the Machian view, the properties of the system are
exhausted by the masses of the particles and their separations, but
the separations are mutual properties. Apart from the masses, the
particles have no attributes that are exclusively their own. They – in
the form of a triangle – are a single thing. In the Newtonian view,
the particles exist in absolute space and time. These external
elements lend the particles attributes – position, momentum, angular
momentum – denied in the Machian view. The particles become
three things. Absolute space and time are an essential part of
atomism.

The lent properties are the building blocks of both classical and
quantum mechanics. Classically, each particle has a unique set of
them, de ning the state of each particle at any instant. This is the
ideal to which realists like Einstein aspire. The lent properties also
occur in quantum mechanics. They are generally not the state itself,
but superpositions of them are. If a quantum system is considered in
isolation from the instruments used to study it, its basic elements still
derive from a Newtonian ontology. This is what misled EPR into
thinking they could outwit Bohr. Einstein’s defeat by Bohr is a clear
hint that we shall only understand quantum mechanics when we
comprehend Mach’s ‘overpowering unity of the All’.

BELL’S INEQUALITIES

Strong con rmation for quantum mechanics being holistic in a very

deep sense was obtained in the 1960s, when John Bell, a British
physicist from Belfast, achieved a signi cant sharpening of the EPR
paradox. The essence of the original paradox is the existence of
correlations between pairs of quantities – pairs of positions or pairs
of momenta – that are always veri ed if one correlation or the other
is tested. By itself, some degree of correlation between the two
particles is not mysterious. The EPR-type correlated states are
generally created from known uncorrelated states of two particles
that are then allowed to interact. Even in classical physics,
interaction under such circumstances is bound to lead to
correlations. Bell posed a sharper question than EPR: is the extent of
the quantum correlations compatible with the idea that, before any
measurement is made, the system being considered already possesses
all the de nite properties that could be established by all the
measurements that, when performed separately, always lead to a
de nite result?
 Bell’s question perfectly re ects Einstein’s ‘robust realism’ – that
the two-particle system ought to consist of two separate entities that
possess de nite properties before any measurements are made.
Assuming this, Bell proceeded to derive certain inequalities, justly
famous, that impose upper limits on the degree of the correlations
that such ‘classical’ entities could exhibit (tighter correlations would
simply be a logical impossibility). He also showed that quantum
mechanics can violate these inequalities: the quantum world can be
more tightly correlated than any conceivable ‘classical world’.
Aspect’s experiments speci cally tested the Bell inequalities and
triumphantly con rmed the quantum predictions. The only way in
which the atomized world after which Einstein hankered can be
saved is by a physical interaction that has so far completely escaped
detection and is, moreover, propagated faster than light. Einstein
could hardly have taken comfort from this straw. Far better, it seems
to me, is to seek understanding of the Here in Mach’s All. I shall give
some indication of what I mean by this after we have considered the
next topic.

THE MANY-WORLDS INTERPRETATION

In 1957, Hugh Everett, a student of John Wheeler at Princeton,

proposed a novel interpretation of quantum mechanics. Its
implications are startling, but for over a decade it attracted little
interest until Bryce DeWitt drew wide attention to it, especially by
his coinage many worlds to describe the main idea. Everett had used
the sober title ‘Relative state formulation of quantum mechanics’.
One well-known physicist was prompted to call it the ‘best-kept
secret in physics’. So far as I know, Everett published no other
scienti c paper. He was already working for the Weapons Systems
Evaluation Group at the Pentagon when his paper was published. He
was apparently a chain smoker, and died in his early fties.

Everett noted that in quantum mechanics ‘there are two

fundamentally di erent ways in which the state function can
change’: through continuous causal evolution and through the
notorious collapse at a measurement. He aimed to eliminate this
dichotomy, and show that the very phenomenon that collapse had
been introduced to explain – our invariable observation of only one
of many di erent possibilities that quantum mechanics seems to
allow – is actually predicted by pure wave mechanics. Collapse is
redundant.

The basis of Everett’s interpretation is the endemic phenomenon

of entanglement. By its very nature, entanglement can arise only in
composite systems – those that consist of two or more parts. In fact,
an essential element of the many-worlds interpretation as it is now
almost universally understood is that the universe can and must be
divided into at least two parts – an observing part and an observed
part. However, Everett himself looked forward to the application of
his ideas in the context of uni ed eld theories, ‘where there is no
question of ever isolating observers and object systems. They are all
represented in a single structure, the eld.’ That is ultimately the kind
of situation that we must consider, but for the moment we shall look
at the familiar form of the interpretation.

The simplest two-particle system can be used to explain a

quantum measurement. The core idea is all that counts. One particle,
called the pointer, is used to establish the location of the other
particle, called the object. Figure 43 shows things with which we are
already familiar. At an initial time tset-up, the pointer (horizontal axis)
and object (vertical axis) are not entangled. For any of the small
range of possible pointer positions, the object has identical ranges of
possible positions, shown schematically by points 1 to 6 on the left.
Determination of the pointer position in this state would tell us
nothing about the object. But the interactions of the particles are so
arranged that by the later time tmeasurement the wave function, passing
through the interaction region, has ‘swung round’ into the position
shown on the right. Remembering how points in Q translate into
positions in space, we see that the object still has its original range of
positions 1 to 6, but that the new pointer positions are strongly
correlated with them.
 Figure 43 The initially unentangled state at tset-up, shown
schematically at the bottom left by the vertical column of positions
1, 2, 3, 4, 5, 6, and the rst probability density above it, evolves into
the entangled state at tmeasurement, indicated on the right by the
inclined numbers 1, 2, 3, 4, 5, 6 and the upper right probability
density.
 If we operate detectors at tmeasurement to nd the pointer’s position
(by letting it strike an emulsion or even, in principle, by acting
ourselves as observers and placing our eyes at the appropriate
places), the standard quantum rules immediately tell us the object’s
position, for di erent pointer positions are now correlated with
di erent object positions. Quantum measurement consists of two
stages: the creation from an unentangled state of a strongly
entangled state (creating the conditions of a so-called good
measurement) followed by the exploitation of the correlations in that
state (using the determination of the pointer position to deduce –
measure – the object’s position). The existence of such correlations
has now been wonderfully well con rmed by experiments. If
detectors have been used to nd the pointer and are then used to
locate the object directly, the quantum correlations predicted are
invariably con rmed. Measurement theory is really veri cation of
the correlations associated with entanglement. Personally, I think
that the term ‘measurement’ has generated misunderstanding, and
that it would be better simply to speak of veri cation of correlations.

The measurement problem of quantum mechanics is this: how does

the entangled state of many possibilities collapse down to just one,
and when does it happen? Is it when the pointer strikes the
emulsion, or when the human observer sees a mark on the emulsion?
I won’t go into all the complications, which depend on how much of
the world we wish to describe quantum mechanically. It leads to a
vicious in nite regress. You can go on asking quantum mechanics
again and again to say when collapse occurs, but it never gives an
answer. The di erent possibilities already represented at tset-up in the
di erent positions of the object system can never be eradicated, and
simply ‘infect’ the rest of the world – rst the pointer, then the
emulsion, then the retina of the experimentalist’s eye, nally his or
her conscious state. All that the Copenhagen interpretation can say is
that collapse occurs at the latest in the perceptions of the
experimentalist. When it happens no one can say – it can only be
said that if collapse does not happen we cannot explain the observed
phenomena.

But must it happen? Everett came up with a simple – with

hindsight obvious – alternative. Collapse does not happen at all: the
multiple possibilities represented in the entangled state continue to
coexist. In each possibility the observer, in di erent incarnations,
sees something di erent, but what is seen is de nite in each case.
Each incarnation of the observer sees one of the possible outcomes
that the Copenhagen interpretation assumes is created by collapse.
The implications of this are startling. A single atomic particle – the
object particle in Figure 43 – can, by becoming entangled with rst
the pointer and then the emulsion, and nally the conscious
observer, split that observer (indeed the universe) into many
di erent incarnations. In his paper in 1970 that at last brought
Everett’s idea to wide notice, Bryce DeWitt wrote:

I still recall vividly the shock I experienced on rst

encountering this multiworld concept. The idea of 10100+
slightly imperfect copies of oneself all constantly splitting
 into further copies, which ultimately become unrecognizable,
is not easy to reconcile with common sense. Here is
schizophrenia with a vengeance.

Everett’s proposal raises two questions. If many worlds do exist,

why do we see only one and not all? Why do we not feel the world
splitting? Everett answered both by an important property of
quantum mechanics called linearity, or the superposition principle. It
means that two processes can take place simultaneously without
a ecting each other. Consider, for example, Young’s explanation of
interference between two wave sources. Each source, when active
alone, gives rise to a certain wave pattern. If both sources are active,
the processes they generate could disturb each other drastically. But
this does not happen. The wave pattern when both sources are active
is found simply by adding the two wave patterns together. The total
e ect is very di erent from either of the individual processes, but in
a real sense each continues una ected by the presence of the other.
This is by no means always the case; in so-called non-linear wave
processes, the wave pattern from two or more sources cannot be
found by simple addition of the patterns from the separate sources
acting alone. However, quantum mechanics is linear, so the much
simpler situation occurs.

As a result, quantum processes can be regarded as being made up

of many individual subprocesses taking place independently of one
another. In Figure 43 the total process of the wave function
‘swinging round’ and becoming entangled is represented
symbolically by the arrows as six individual subprocesses (or
branches, to use Everett’s terminology). In all of them, the pointer
starts in the same position but ends in a di erent position. Everett
makes the key assumption that conscious awareness is always
associated with the branches, not the process as a whole. Each
subprocess is, so to speak, aware only of itself. There is a beautiful
logic to this, since each subprocess is fully described by the quantum
laws. There is nothing within the branch as such to indicate that it
alone does not constitute the entire history of the universe. It carries
on in blithe ignorance of the other branches, which are ‘parallel
worlds’ of which it sees nothing. The branches can nevertheless be
very complicated. An impressive part of Everett’s paper demonstrates
how an observer (modelled by an inanimate computer) within one
such branch could well have the experience of being all alone in such
a multiworld, doing quantum experiments and nding that the
quantum statistical predictions are veri ed.

Any scienti c theory must establish a postulate of psychophysical

parallelism: it is necessary to say what elements of the physical theory
correspond to actual conscious experience. Given our current meagre
understanding of consciousness, we have considerable freedom in the
choice we make. Everett exploited the linearity of quantum
mechanics to make his particular choice. It leads, however, to what
is now widely seen as a serious technical problem.

Right at the start, Everett stated that ‘The wave function is taken
as the basic physical entity with no a priori interpretation.’ He aimed
to show that the interpretation of the theory emerges from ‘an
investigation of the logical structure of the theory’. This aim, coupled
with his insistence that the wave function is the only thing that
exists, creates the di culty, since the logical structure of the theory
is generally reckoned to be represented by Dirac’s transformation
theory. According to it, any quantum state can indeed be regarded as
made up of other states – branches in an Everett-type ‘many-worlds’
picture. The di culty is that this representation is not unique. There
are many di erent ways in which one and the same state, formed
from the same two ‘observer’ and ‘object’ systems, can be
represented as being made up of other states. We can, for example,
use position states, but we can equally well use momentum states.

The fact is that quantum mechanics is doubly inde nite. First, if

states of a de nite kind are chosen, any state of a composite system
is a unique sum of states of its subsystems. For position states, this is
shown in Figures 40 to 43. The probability distribution is spread out
over a huge range of possibilities in which one particle has one
de nite position and the other particle has another de nite position.
Positions are always paired together in this way. Everett resolved the
apparent con ict between our experience of a unique world and this
multiplicity of possibilities by associating a separate and autonomous
experience with each. However, he did not address the second
inde niteness: the states shown as positions in Figures 40 to 43 could
equally well be represented by, for example, momentum states. Then
pairs of momentum states result. Depending on the representation,
di erent sets of parallel worlds are obtained: ‘position histories’ in
the one case, ‘momentum histories’ in the other. One quantum
evolution yields not only many histories but also many families of
di erent kinds of history.

It was surprisingly long before this di culty was clearly

recognized as the preferred-basis problem: a de nite kind of history
will be obtained only if there exists some distinguished, or preferred,
choice of the basis, by which is meant the kind of states used in the
representation. The preferred basis problem is the EPR paradox in a
di erent guise. Everett may have instinctively assumed that the
position basis is somehow naturally singled out, but there is little
evidence in his paper to con rm this.

The rst question that must be addressed is surely this: what is

real? Everett took the wave function to be the only physical entity.
The price for this wave-function monism is the preferred-basis
problem. Because the wave functions of composite systems can be
represented in so many ways, the application of Everett’s ideas to
di erent kinds of representation suggests that one and the same
wave function contains not only many histories, but also many
di erent kinds of history. It leads to a ‘many-many-worlds’
interpretation. Some accept this, but I feel there is a more attractive
alternative.

A DUALISTIC PICTURE

The purists among the quantum ‘founding fathers’, above all Dirac
and Heisenberg, saw a close parallel between the representation of
one and the same quantum state in many ways and the possibility of
putting many di erent coordinate systems on one and the same
space-time. In relativity, this corresponds to splitting space-time into
space and time in di erent ways. After Einstein’s great triumph, no
physicist would dream of saying that this could be done in one way
only. Similarly, Dirac and Heisenberg argued, there is nothing in
quantum theory to suggest that there is a preferred way to represent
quantum states. However, the parallel may not be accurate.

First, in classical relativity, space-time represents all reality – the

complete universe. In contrast, a quantum state by itself has no
de nite meaning until the strategic decision – say, to measure
position or momentum – has been taken. The state acquires its full
meaning only in conjunction with actual measuring apparatus
outside the system. The system must interact with the apparatus to
reveal its latent potentialities. At present, its interaction with an
apparatus – essentially the rest of the universe – is not fully
understood. The quantum state by itself is only part of the story. It
may be premature to draw conclusions about the quantum universe
from incomplete quantum descriptions of subsystems of it.

Second, quantum mechanics as presently formulated needs an

external framework. Indeed, the most basic observables, those for
position, momentum and angular momentum, all correspond to the
‘lent’ properties mentioned in the discussion of the EPR paradox.
They could not exist without the framework of absolute space, and
Mach’s principle suggests strongly it is determined by the
instantaneous con gurations of the universe. Time, moreover, plays
an essential role in quantum mechanics yet stands quite outside the
description of the quantum state. But we saw in Chapter 6 that time
is really just a shorthand for the position of everything in the
universe, so the con gurations of the universe can be expected to
play an essential and direct role in a quantum description of the
universe. I cannot see how we can hope to understand the external
framework of current quantum theory unless we put them into the
foundations of quantum cosmology. This is what leads me to the
dualistic picture of Platonia, the collection of all possible
con gurations of the universe, and the completely di erent wave
function, conceived of as ‘mist’ over Platonia. In the language of
Everett’s theory, this introduces a preferred basis. In answer to the
question ‘what is real?’, I answer ‘con gurations’. My book is the
attempt to show that they explain both time and the quantum – as
di erent sides of the same coin.
CHAPTER15
The Rules of Creation

THE END OF CHANGE

In this chapter I am going to go into a little detail about how wave

mechanics works. This means looking at two equations Schrödinger
discovered in 1926 which, Dirac remarked, explained all of
chemistry and most of physics. You will need to absorb enough to
understand the bearing of the rst part of the book on the structure
of quantum cosmology. That is the goal; I hope you will nd it is
worth the e ort. I believe it will show us how creation works. No
theory can ever explain why anything is – that is the supreme
mystery. But theory may be able to tell us why one thing rather
than another is created and experienced. What is more, I believe
that in every instant we experience creation directly. Creation did
not happen in a Big Bang. Creation is here and now, and we can
understand the rules that govern it. Schrödinger thought he had
found the secret of the quantum prescriptions. Properly understood,
what he found were the rules of creation.

Let us get down to business. We shall be considering how the

wave function ψ changes. In quantum mechanics, this is all that
does change. Forget any idea about the particles themselves moving.
The space Q of possible con gurations, or structures, is given once
and for all: it is a timeless con guration space. The instantaneous
position of the system is one point of its Q. Evolution in classical
Newtonian mechanics is like a bright spot moving, as time passes,
over the landscape of Q. I have argued that this is the wrong way to
think about time. There is neither a passing time nor a moving spot,
just a timeless path through the landscape, the track taken by the
moving spot in the ction in which there is time.

In quantum mechanics with time, which we are considering now,

there is no track at all. Instead, Q is covered by the mists I have
been using to illustrate the notion of wave functions and the
probabilities associated with them. The red and green mists evolve
in a tightly interlocked fashion, while the blue mist, calculated from
the other two, describes the change of the probability. All that
happens as time passes is that the patterns of mist change. The mists
come and go, changing constantly over a landscape that itself never
changes.

One of the equations that Schrödinger found governs this

process. If ψ is known everywhere in Q at a certain time, you know
what ψ will be slightly later. From this new value, you can go on
another small step in time, and another, and so on arbitrarily far
into the future. The role played here by the red and green mists, the
two primary components of ψ, is quite interesting: the way the red
mist varies in space determines the rate of change of the green mist
in time, and vice versa. The two components play a kind of tennis.
This equation is sometimes called the time-dependent Schrödinger
equation because time features in it. This is not in fact the rst
equation that Schrödinger discovered.

The rst one he found is now usually called the stationary or

time-independent Schrödinger equation. This determines what
happens in certain special cases in which the two components of ψ,
the red and green ‘mists’, oscillate regularly, the increase of one
matching the decrease of the other. This has the consequence, as we
have already seen for a momentum eigenstate, that the blue mist
(the probability density) has a frozen value – it is independent of
time (though its value generally changes over Q). Such a state is
called a stationary state. This explains the name given to the second
equation – its solutions are stationary states. The standard view is
that the time-dependent equation is the fundamental equation of
quantum mechanics; the stationary equation is seen as a special case
derived from it. This corresponds to an overall scheme in which
some state of ψ is created at some time and then evolves until a
measurement is made.

There are intriguing hints that in the quantum mechanics of the

universe the roles of these two equations are reversed. The
stationary equation (or something like it) may be the fundamental
equation, from which the time-dependent equation is derived only
as an approximation. We think it is fundamental because we have
been fooled by circumstances that make it valid for the description
of the phenomena we nd around us. However, these phenomena
deceive us greatly when it comes to the overall story of the
universe. In particular, they lead us to believe time exists when it
does not.

That this is likely to be so follows from an important property of

the two Schrödinger equations. For any quantum system, we can use
the time-independent equation to nd all the stationary states it can
have. Each of these states corresponds to a de nite energy, and in
each of them the red and green mists oscillate with the same xed
frequency while the blue mist remains constant. These solutions are
also solutions of the time-dependent equation, though they are
special, being stationary. I have mentioned linearity in quantum
mechanics. Here, linearity means that two or more solutions of the
time-dependent Schrödinger equation can be simply added together
to give another solution. If the special stationary solutions are
added, something signi cant results. In each solution, considered
separately, the red and green mists oscillate at a xed frequency
while the blue mist remains constant. However, when we add two
such solutions with di erent frequencies, they interfere: the added
intensities of the red and green mists no longer oscillate regularly.
More signi cantly, the blue mist varies in time.

Now this is very characteristic – indeed, it is the essence of

quantum evolution. All solutions of the time-dependent equation
can be found by adding stationary solutions with di erent
frequencies. Each stationary solution on its own has regular
oscillations of its red and green mists, but a constant – in fact static
– distribution of its blue mist. But as soon as stationary states with
di erent energies, and hence frequencies, are added together,
irregular oscillations commence – in particular in the blue mist, the
touchstone of true change. All true change in quantum mechanics
comes from interference between stationary states with di erent
energies. In a system described by a stationary state, no change takes
place.

The italics are called for. We have reached the critical point. The
suggestion is that the universe as a whole is described by a single,
stationary, indeed static state. Why should this – with its implication
that nothing happens – be so? This is where we start to make
contact with the earlier part of the book. Time and change come to
an end when Machian classical dynamics meets quantum mechanics.
We have seen that a Machian universe should have only one value
of the energy: zero. We also know (Box 2) that a quantum theory
can be obtained by quantizing a corresponding classical theory. In
fact, it is easy to show that whereas quantizing Newtonian
dynamics, with its external framework of space and time, leads to
the time-dependent Schrödinger equation, quantizing the simple
Machian model considered in Chapter 7 leads to a quantum theory
in which the basic equation is not the time-dependent but the
stationary Schrödinger equation.

If the Machian approach to classical dynamics is correct,

quantum cosmology will have no dynamics. It will be timeless. It
must also be frameless.

CREATION AND THE SCHRÖDINGER EQUATION

Before I can explain how this can be achieved, I must tell you what
the Schrödinger equation is like and what it can do. I believe it is
even more remarkable than physicists realize. This is where – if I am
right – we are getting near the secret of creation.

When Schrödinger created wave mechanics, Bohr’s was the only

existing model of the atom. It suggested that atoms could exist in
stationary states, each with a xed energy, photons being emitted
when the atom jumped between them. Schrödinger’s great aim was
to explain how the stationary states arise and the jumps occur. De
Broglie’s proposal suggested strongly that a stationary state should
be described by a wave function that oscillated rapidly in time with
xed frequency, though its amplitude might vary in space. As a rst
step Schrödinger therefore looked for an equation for the variation
in space.

It is ironic that only later did he nd the time-dependent

equation from which, strictly speaking, he should have derived this
equation. However, he had luck and was guided by good intuition.
Although it is easy for mathematicians, I shall not go into the details
of how Schrödinger found his equations or how to get from one to
the other. Box 13 gives the minimum about the stationary equation
needed to understand the thrust of the story.
BOX 13 How Creation Works

You can think of the Schrödinger wave function in a stationary state

as follows. At each point of the con guration space Q, imagine a
child swinging a ball in a vertical circle on a string of length ,
which remains constant. As the ball whirls, its height above or
below the centre of the circle changes continuously. The height is an
image of the red mist, which is sometimes positive (above the
centre), sometimes negative (below it). The distance sideways – to
the right (positive) or the left (negative) – is an image of the green
mist. The square of is the image of the constant intensity of the
blue mist. A stationary state is like having children swinging such
balls at the same rate everywhere in Q, all perfectly in phase – they
all reach the top of the circle together. The only thing not perfectly
uniform is the string length, , which can change from point to
point in Q. In a momentum eigenstate, is the same everywhere. It
is a very special state, but in a more general stationary state does
vary over Q. The stationary Schrödinger equation governs its
variation.
 It does this by imposing a condition at each point of Q. The sum
of two numbers, calculated in de nite ways, must equal a third. The
rst number is the most interesting but the most di cult to nd.
Take a quantum system of three bodies. Its con guration space Q
has nine dimensions. Each point in Q corresponds to a position of
the three bodies in absolute space. Imagine holding two bodies
xed, and moving the third along a line in absolute space. This will
move you along a line in Q. Suppose that along it you plot , the
string length, as a curve above the line. At each point, this curve
will have a certain curvature. At some places it will curve strongly,
towards or away from the line, at others weakly. In the calculus, the
curvature is the second derivative.

At each point of Q there are nine such curvatures because Q has

nine dimensions, one for each of the three directions in absolute
space in which each particle can move. The rst number in the
Schrödinger condition is the sum of these nine curvatures after each
has been multiplied by the mass of the particle for which it has been
calculated. I shall call this the curvature number.

The second number is much easier to nd. Recall that any

con guration of bodies has an associated potential energy. The
con guration (and the nature of the bodies, their masses, etc.)
determines it uniquely. For gravity, this was explained in Figure 17.
The second number, which I shall call the potential number, is found
simply by multiplying the potential by
 The third number is also easy to nd. If ω is the frequency of the
state (the number of ‘rotations of the balls’ in a second), then, by the
quantum rules, the energy of the state is E = hω, where h is
Planck’s constant. This is the relationship Einstein found between
the energy and frequency of a photon. The third number, which I
shall call the energy number, is then found by multiplying the energy
E by .

The condition imposed by the stationary Schrödinger equation is

then

Curvature number + Potential number = Energy number

(Planck’s constant also occurs in the rst number, to ensure that all
three numbers have the same physical nature.)

However, nding this condition, which must hold everywhere in

Q, was only half the story. Schrödinger thought that an atom in a
stationary state was like a violin string vibrating in resonance.
Because its two ends are xed, the amplitude at the ends is zero. He
therefore imposed on not only the above condition, but also the
condition that it should tend to zero at large distances. It was this
requirement that enabled him to make the huge discovery that
convinced him – and very soon everyone else – that he had found
the secret of Bohr’s quantum prescriptions.

This hinges on an extremely interesting property of the

stationary Schrödinger equation. As yet E is a xed but unknown
number. It may be smaller or greater than the potential V, which
varies over Q. The interesting thing is that the above condition
forces to do very di erent things depending on the value of E – V.
Where it is greater than zero, oscillates. As Schrödinger said rather
quaintly, ‘it does not get out of control’. However, where E – V is
less than zero, the condition forces an entirely di erent behaviour
on . It must either tend rapidly to zero or else grow rapidly –
exponentially in fact – to in nity. The latter would be a disaster.
Schrödinger therefore commented that things become tricky and
must be handled delicately. Indeed, he showed that it is only in
exceptional cases, for special values of E, that does not ‘explode’
but instead subsides to zero at in nity. These are the cases he was
looking for. Well-behaved solutions exist for only certain values of
E, which are discrete (separated from each other) if E is less than
zero.

The well-behaved solutions are called eigenfunctions, and the

corresponding values of E are called (energy) eigenvalues. It is a
fundamental property of quantum mechanics that any system
always has at least one eigenfunction. The eigenfunction of any
system that has the lowest value of its energy eigenvalue (there is
often only one such eigenfunction) is called the ground state. In
general, there are also eigenfunctions with higher energies, called
excited states. Finally, if E is large enough for E – V to be positive
everywhere, the eigenfunctions oscillate everywhere, though more
rapidly where the potential is lowest. The negative eigenvalues E
form the discrete spectrum, and the corresponding states are called
bound states because for them has an appreciable value only over a
nite region. The remaining states, with E greater than zero, are
called unbound states, and their energy eigenvalues form the
continuum spectrum.

Schrödinger won the 1933 Nobel Prize for Physics mainly for his
wave-mechanical calculation for the hydrogen atom. He found that
the energy eigenvalues of its stationary states are precisely the
energies of the allowed states in Bohr’s model. This was a huge
advance, since Schrödinger’s formalism had an inner unity and
consistency to it completely lacking in the older model. Brilliant
successes of the new wave mechanics, many achieved by
Schrödinger himself, soon came ooding in, leaving no doubt about
the great fruitfulness of the new scheme.

In Chapter 14 I described how molecules appear in the

Schrödinger picture: as immense collections of all the con gurations
they could conceivably have, with the blue mist of the quantum
probability strongly concentrated on the most probable
con gurations. These most probable con gurations, generally
clustered around a single point in Q, are the ones represented by the
ball-and-strut models. I can now begin to make good my claim that
Schrödinger found the laws of creation. His stationary equation
determines the structures – indeed, creates the structures – of all
these amazing atoms and molecules that constitute so much of the
matter in the universe, our own bodies included. The equation does
it by determining which structures are probable. But I mean
creation not only in this sense of the structure of atoms and
molecules, but in an even deeper one. The full explanation is still to
come, but we are getting closer to our quarry.

QUANTUM MECHANICS HOVERING IN NOTHING

We must now see if we can dispense not only with time but also
with absolute space in quantum mechanics. In a timeless system the
energy E is zero, and the condition in Box 13 says simply that at
every point of Q the sum of the curvature number and the potential
number is zero. The potential number is already in the form we
need. For any possible relative con guration, the potential has a
unique value: it depends on nothing else. To nd the potential
number, we simply calculate the potential V for each con guration
and then multiply by , getting This part of the calculations is
pleasingly self-contained because V depends only on the relative
con guration. Each structure has its own potential irrespective of
how we imagine the structure to be embedded in space.

However, a lack of ‘self-containment’ shows up in the curvature

number. To nd it, we must know how varies from position to
position in the con guration space Q. This is not a self-contained
process in Schrödinger’s equation because the points of his Q are
de ned by the particles’ positions in absolute space, which is used
crucially in Q, making it hybrid. The all-important curvatures of
are ultimately determined by position di erences in absolute space.
As a result, in standard quantum mechanics the orientations are in
general entangled with the relative data that specify the particle
separations. Now, besides positions, momenta and energy there is
another very important quantity in quantum mechanics – angular
momentum, which, being an action, always has discrete
eigenvalues. It owes its existence in quantum mechanics to absolute
space. We have not yet escaped from Newton’s framework.

We are now coming to another critical point. We have seen that

in classical physics the action is a kind of ‘distance’ between two
con gurations that are nearly but not exactly the same. Absolute
space is an auxiliary device that makes it possible to de ne such
‘distances’. This is why angular momentum exists in classical and
quantum physics. However, in Chapter 7 we found an alternative
de nition of ‘distances’ that works in the purely relative
con guration space – in Platonia – and owes nothing to absolute
space. They are de ned by the best-matching procedure, which uses
relative con gurations and nothing else. In classical physics, this
makes it possible to create a purely relative and hence self-
contained dynamics. We also found that a sophisticated form of best
matching lies at the heart of general relativity. Best matching would
appear to be a basic rule of the world.

It is therefore very tempting to see whether it can be applied in

quantum mechanics. What we would like to do is establish rules for
operating on wave functions de ned solely on the relative
con guration space. For example, for three bodies we would want
to eliminate the six dimensions associated with their position and
orientation in absolute space, and work just with the sides of the
triangle. We shall then have a wave function de ned on a three-
dimensional Platonia. For that, we shall want to calculate a
curvature number and a potential number. The latter will present no
di culty, since it will be the same as in ordinary quantum
mechanics. The di culty is in the curvature number. What, after all,
is curvature? For any given curve, it is the rate at which its slope
changes. But the key thing about a rate of change is that it is with
respect to something. That something is all-important. It is a kind of
‘distance’. The ordinary quantum-mechanical ‘distance’ is simply
distance in absolute space (times the mass of the particle
considered). To eliminate absolute space in classical physics, we
replaced it by the Machian best-matching distance. There is no
reason why we should not do the same in quantum physics.

This is where the unfolding of quantum mechanics on

con guration space is so important. To retain that essential property
of it – the huge step that Schrödinger took – we must pass from his
hybrid Q to Platonia. If we are to succeed in formulating quantum
mechanics in the new arena, there must be ‘distances’ in it. But that
is precisely what the best-matching idea was developed to provide.
Exactly the same ‘distances’ needed to realize Mach’s principle in
classical physics can be used in a version of wave mechanics for a
universe without absolute space. All we have to do is measure
curvatures with respect to the Machian distances created on
Platonia by best matching. We then add curvatures measured in as
many mutually perpendicular directions as there are dimensions in
that timeless arena, and set the sum equal to minus the potential
number.

In fact, it is quite easy to see that the wave functions that satisfy
the Schrödinger conditions in this Machian case are precisely the
eigenfunctions of ordinary quantum mechanics for which the
angular-momentum eigenvalues are zero. This exactly matches our
result in classical mechanics – that the best-matching condition
leads to solutions identical to the Newtonian solutions with angular
momentum zero. We have already seen why they must be static
solutions.

The picture that emerges is very simple. The quantum

counterpart of Machian classical dynamics is a static wave function
ψ on Platonia. The rules that govern its variation from point to point
in Platonia involve only the potential and the best-matching
‘distance’. Both are ‘topographic features’ of the timeless arena.
Surveyors sent to map it would nd them. They would see that the
mists of Platonia respect its topography. It determines where the
mist collects.
CHAPTER 16
‘That Damned Equation’

HISTORY AND QUANTUM COSMOLOGY

The year 1980 was another turning point in my life. It was when
Bruno Bertotti and I thought we might have found a new theory of
gravitation, only to learn that the two ideas on which we had based
it were already an integral part of Einstein’s theory. Karel Kuchař’s
intervention rounded o our work but also brought it to an end. It
was something of an anticlimax. Bruno became increasingly
involved in experiments using spacecraft, aimed at detecting the
gravitational waves predicted by Einstein’s theory. For a year or two
I actually stopped doing physics and became politically active in the
newly founded Social Democratic Party (the SDP). However, the old
interests soon revived. Margaret Thatcher’s decisive general election
victory in 1983 hastened the process.

Two things occupied me through the 1980s. First, I wrote the

book from which I quoted the comments about Kepler. It had always
been my ambition to write about absolute and relative motion, and
in 1984 I signed a contract with Cambridge University Press for a
book of four hundred pages covering the period from Newton to
Einstein and including an account of my work with Bruno. When I
embarked upon it, it occurred to me that I ought to nd out why
Newton had said what he had. What had given him the idea of
absolute space? Might it not be an idea to look at what Galileo had
said? I made a wonderful mistake by asking those questions. Before
I knew what was happening, my research into Galileo dragged me
ever further into past history, through the Copernican revolution to
the work of Ptolemy and all the way back to the pre-Socratic
philosophers. By reading the actual works of scientists such as
Ptolemy, Kepler and Galileo, I found that the early history of
mechanics and astronomy was far more interesting than any account
of it I could nd by the professional historians of science. They had
missed all sorts of fascinating things, and their histories were quite
inadequate. Inspired by Kepler’s comment that the ways by which
men discover things in the heavens are almost as interesting as the
things themselves, I started to write about all the early work. I spent
from 1985 to 1988 writing a completely unplanned book: The
Discovery of Dynamics. My sympathetic and understanding editor at
Cambridge, Simon Capelin, agreed to publish it as the rst of a two-
volume work. The second volume was to be the book originally
proposed and should have been completed a year or two later.
However, that got badly delayed by a parallel development that
turned my interest to physics that does not yet exist at the same
time as I was working backward to the early history.
 As I mentioned earlier, Bruno and I had been completely
concerned with classical physics. We had wanted to show that Mach
had been right and that his ideas could lead to new classical physics;
we had given not a moment’s thought to any quantum implications
they might have. Quantum cosmology was a world beyond our ken.
It is strange what sparks a desire to work on something. My lack of
interest in quantum gravity was particularly odd, since it was the
early work done in that eld which, through the remark by Dirac,
quoted in the Preface, had set me on my long trek. It was the same
work that had led to the work of Baierlein, Sharp and Wheeler that
Bruno and I had come to see as the implementation of Mach’s ideas
within general relativity. Not even working with Karel Kuchař, one
of the world’s leading experts in quantum gravity, provided the
stimulus I needed. Perhaps it all seemed too daunting. I needed the
example and encouragement that came from a new friend, Lee
Smolin.

I rst met Lee a few weeks before I travelled to Salt Lake City in
the autumn of 1980. It was quite a dramatic time for me since I had
just narrowly escaped death through an insidious appendix that had
burst without giving me any pain. My only symptoms were
tiredness, slight sickness and the merest hint of stomach pain.
Luckily my vigilant doctor sent me to hospital as a precaution. An X-
ray proved di cult to interpret, and after quite lengthy deliberation
the doctors decided to open me up. They found that any further
delay could have been fatal. Seeing my state, the surgeon apparently
commented that ‘this must be a very brave man’, believing I must
have been in agony. In fact, I had been cheerfully reading The Times
without any discomfort only half an hour before the operation. The
day after I came back from hospital still convalescing, two American
physicists visiting Oxford phoned to say that they had heard from
Roger Penrose about my interest in Mach’s principle. Could they
come and see me? They came the next day, and I greeted them in
my dressing gown.

One was Lee, then a young postdoc. The meeting changed both
of our lives signi cantly. He proved very receptive to the ideas of
Leibniz and Mach to which I introduced him, while he encouraged
me to see what application they might have to the problem to which
he had decided to devote himself – quantum gravity. We met several
times in the next few years, and collaborated on an attempt to
formulate Leibniz’s philosophical system, his ‘monadology’, in
mathematical form. I think we made some real progress. Lee has
written about his view of things in his The Life of the Cosmos. Certain
aspects of our work together were decisive in my own elaboration of
the notion of time capsules and my conviction that the ultimate and
only truly real things are the instants of time. As far as I am aware,
Leibnizian ideas o er the only genuine alternative to Cartesian-
Newtonian materialism which is capable of expression in
mathematical form. What especially attracts me to them is the
importance, indeed primary status, given to structure and
distinguishing attributes, and the insistence that the world does not
consist of in nitely many essentially identical things – atoms
moving in space – but is in reality a collection of in nitely many
things, each constructed according to a common principle yet all
di erent from one another. Space and time emerge from the way in
which these ultimate entities mirror each other. I feel sure that this
idea has the potential to turn physics inside out – to make the
interestingly structured appear probable rather than improbable.
Before he became a poet, T. S. Eliot studied philosophy. He
remarked, ‘In Leibniz there are possibilities.’

In 1988, when I had nished my book on the discovery of

dynamics, I spent three weeks with Lee at Yale, and began to think
seriously how one might make sense of the embryonic form of
quantum gravity that had been developed from about the time of
Einstein’s death in 1955, leading to the publication of the Wheeler-
DeWitt equation in 1967. During the next four years, Lee and I had
many discussions. Although we eventually followed di erent paths –
Lee is reluctant to give up time as a primary element in physics –
the ideas I want to describe in the nal part of the book crystallized
during those discussions. For me, their attraction stems from the
inherent plausibility of Platonia as the arena of the universe and the
implication of Schrödinger’s breathtaking step into a rather similar
con guration space. As I see it now, the issue is simple.

A SIMPLE-MINDED APPROACH
You can play di erent games in one and the same arena. You can
also adjust the rules of a game as played in one arena so that it can
be played in a di erent arena. Both general relativity and quantum
mechanics are complex and highly developed theories. In the forms
in which they were originally put forward, they seem to be
incompatible. What I found to my surprise was that it does seem to
be possible to marry the two in Platonia. The structures of both
theories, stripped of their inessentials, mesh. What if Schrödinger,
immediately after he had created wave mechanics, had returned to
his Machian paper of only a year earlier and asked himself how
Machian wave mechanics should be formulated? His Machian paper
implicitly required Platonia to be the arena of the universe, while
any wave mechanics simply had to be formulated on a con guration
space. Such is Platonia, though it is not quite the hybrid Newtonian
Q he had used. But the structure of Machian wave mechanics would
surely have been immediately obvious to him, especially if he had
taken to heart Mach’s comments on time. As a summary of the
previous chapter, here are the steps to Machian wave mechanics in
their inevitable simplicity.

For a system of N particles, the Schrödinger wave function in the

Newtonian case will in general change if the relative con guration
is changed, if the position of its centre of mass is changed, if its
orientation is changed, and if the time is changed. Mathematicians
call these things the arguments of the wave function. They constitute
its arena. To see what really counts, we can write the wave function
in the symbolic way that mathematicians do:

ψ (relative con guration, centre of mass, orientation, (1)

time)

But if the N particles are the complete universe, there cannot be any
variation with change of centre of mass, orientation or time for the
simple reason that these things do not exist. The Machian wave
function of the universe has to be simply

ψ (relative con guration) (2)

Note the grander ψ. This is the wave function of the universe. It has
found its home in Platonia.

I have met distinguished theoretical physicists who complain of

having tried to understand canonical quantum gravity, the
formalism through which the Wheeler-DeWitt equation was found,
and have given up, daunted by the formalism and its seemingly
arcane complexity. But, as far as I can see, the most important part
boils down simply to the passage from the hybrid (1) to the holistic
(2).

‘THAT DAMNED EQUATION’

This is a bold claim, but the fact is that it still remains the most
straightforward way to understand the Wheeler-DeWitt equation. To
conclude Part 4, I shall say something about this remarkable
equation and the manner of its conception, which unlike the hapless
Tristram Shandy’s was inevitable, being rooted in the structure of
general relativity. You may nd this section a little di cult, which
is why I have just given the simple argument by which I arrive at its
conclusion. Just read over any parts you nd tough.

That there was a deep problem of time in a quantum description

of gravity became apparent at the end of the 1950s in the work of
Dirac and Arnowitt, Deser and Misner (ADM) described in Chapter
11. The existence of the problem was – and still is – mainly
attributed to general covariance. The argument goes as follows. The
coordinates laid down on space-time are arbitrary. Since the
coordinates include one used to label space-time in the time
direction and all coordinates can be changed at whim, there is
clearly no distinguished label of time. This is what leads to the
plethora of paths when a single space-time is represented as
histories in Platonia. However, the real root of the problem lies in
the deep structure of general relativity that we considered in the
same chapter.

Indeed, as Dirac and ADM got to grips with the dynamics of

general relativity, the problem began to take on a more concrete
shape. The rst fact to emerge clearly was the nature of the ‘things
that change’. This was very important, since it is the ‘things that
change’ that must be quantized. They turned out to be 3-spaces –
everything in the universe on one simultaneity hypersurface,
including the geometrical relationships that hold within it. These
are the analogue of particle positions in elementary quantum
mechanics. As I have mentioned, Dirac was quite startled by this
discovery – it clearly surprised him that dynamics should distinguish
three-dimensional structures in a theory of four-dimensional space-
time. I am surprised how few theoreticians have taken on board
Dirac’s comments. Many carry on talking about the quantization of
space-time rather than space (and the things within it). It is as if
Dirac and ADM had never done their work. Theoreticians are loath
to dismantle the space-time concept that Minkowski introduced. I
am not suggesting anything that he did is wrong, but it may be
necessary to accommodate his insight to the quantum world in
unexpected ways. One way or another, something drastic must be
done.

As explained in Chapter 11, in general relativity four-

dimensional space-time is constructed out of three-dimensional
spaces. It turns out that their geometry – the way in which they are
curved – is described by three numbers at each point of space. This
fact of there being three numbers acquired a signi cance for
quantum gravity a bit like the Trinity has for devout Christians.
Intriguingly, the issue at stake is somewhat similar – is this trinity
one and indivisible? Is one member of the trinity di erent in nature
from the other two? The reason why the three numbers at each
space point turned into such an issue is because it seems to be in
con ict with a fact of quantum theory that I need to explain brie y.

I mentioned in Chapter 12 the ‘zoo’ of quantum particles, which

are excitations of associated elds. The typical example is the
photon – the particle conjectured by Einstein and associated with
Maxwell’s electromagnetic eld. An important property of particles
is rest mass. Some have it, others do not. The massless particles
must travel at the speed of light – as the massless photon does. In
contrast, electrons have mass and can travel at any speeds less than
the speed of light.

Now, massless particles are described by fewer variables

(numbers) than you might suppose. Quantum mechanically, a
photon with mass would be associated with vibrations, or
oscillations, in three directions: along the direction of its motion
(longitudinal vibrations) and along two mutually perpendicular
directions at right angles to it (transverse vibrations). However, for
the massless photon the longitudinal vibrations are ‘frozen out’ by
the e ects of relativity, and the only physical vibrations are the two
transverse ones. These are called the two true degrees of freedom.
They correspond to the two independent polarizations of light. This
remark may make these rather abstract things a bit more real for the
non-physicist. Humans cannot register the polarization of light, but
bees can and use it for orientation.

There are many similarities between Maxwell’s theory of the

electromagnetic eld and Einstein’s theory of space-time. During the
1950s this led several people – the American physicist Richard
Feynman was the most famous, and he was followed by Steven
Weinberg (another Nobel Laureate and author of The First Three
Minutes) – to conjecture that, just as the electromagnetic eld has its
massless photon, the gravitational eld must have an analogous
massless particle, the graviton. It was automatically assumed that the
graviton – and with it the gravitational eld – would also have just
two true degrees of freedom.

From 1955 to about 1970, much work was done along these
lines in studies of a space-time which is almost at and therefore
very like Minkowski space (I did my own Ph.D. in this eld). In this
case, the parallel between Einstein’s gravitational eld and
Maxwell’s electromagnetic eld becomes very close, and a
moderately successful theory (experimental veri cation is at present
out of the question, gravity being so weak) was constructed for it.
Within this theory it is certainly possible to talk about gravitons;
like photons, they have only two degrees of freedom. However,
Dirac and ADM had set their sights on a signi cantly more
ambitious goal – a quantum theory of gravity valid in all cases. Here
things did not match up. The expected two true degrees of freedom
did not tally with the three found from the analysis of general
relativity as a dynamical theory – as geometrodynamics.

Within the purely classical theory, the origin of the mismatch is

clear: it is the criss-cross best-matching construction of space-time
that I illustrated with the help of Tristan and Isolde. However, the
discrepancy between the quantum expectations of well-behaved
massless particles with two polarizations and the intricate
interstreaming reality of relativity rapidly became the central
dilemma of quantum gravity. Forty years on, it has still not yet been
resolved to everyone’s satisfaction – it is that intractable. This is
perhaps not surprising, for the issue at stake is the fabric of the
world. Does it exist in something like that great invisible framework
that took possession of Newton’s imagination, or is the world self-
supporting? Do we swim in nothing? Nobody has yet been able to
make quantum theory function without a framework. In fact, many
people do not realize that the framework is a potential problem –
Dirac’s transformation theory is in truth the story of acrobatics in a
framework, and for physicists nurtured on Dirac’s The Principles of
Quantum Mechanics the acrobatics is quantum theory. Acrobatics
must be precise – if the trapeze is not where it should be, death can
result. Such are the exigencies that led the early researchers to posit
a graviton with just two true degrees of freedom.

An intriguing way to achieve this was suggested by Baierlein,

Sharp and Wheeler’s paper in 1962 and its enigmatic hint that ‘time’
was somehow carried within space. This was taken literally,
especially since it seemed to solve another problem of quantum
acrobatics. In real acrobatics, not only location but also timing is of
the essence. Nobody knew how to do quantum mechanics without
an independent time external to the quantum degrees of freedom.
But such a time appeared to have gone missing in gravity. Instead of
time and two true degrees of freedom, there appeared to be no time
but three degrees of freedom; these, moreover, were suspect. The
count was all too suggestive – and many people came to the same
conclusion: there is a time, but it is hidden in the three degrees of
freedom.

According to this insight, the basic framework of quantum

mechanics could be preserved, but the time it so urgently needed
would be taken from the ‘world’ to which it was to be applied.
Putting it in very gurative terms, one-third of space would become
time, while the remaining two-thirds would become two true
quantum degrees of freedom. Because time was to be extracted from
space, from within the very thing that changes, the time that was to
be found was called intrinsic time. The notion of intrinsic time was –
and is – a breathtaking idea. But there was a price to be paid, and
there was also a closely related problem to be overcome: which
third of space is to be time?

The problem was that no clear choice could be made. Any and
all 3-spaces can appear in the relations that summarize so
beautifully the true essence of general relativity. What is more, any
choice would ultimately amount to the introduction of distinguished
coordinates on space-time. But this would run counter to the whole
spirit of relativity theory, the essence of which was seen to be the
complete equivalence of all coordinates. So if a choice were made,
the price would be the loss of this equivalence. The price and the
problem are one and the same. They presented the quantum
theoreticians with a head-on collision between the basic principles
of their two most fundamental theories – the need for a de nite
time in quantum mechanics and the denial of a de nite time in
general relativity. At an international meeting on quantum gravity
held at Oxford in 1980, Karel Kuchař, concluding his review of the
subject, stated that the problem of ‘quantum geometrodynamics is
not a technical one, but a conceptual one. It consists in the
diametrically opposite ways in which relativity and quantum mechanics
view the concept of time’. I have added the italics. I was there to hear
the talk, and Kuchaf’s comment made a deep impression on me.

The search for the third of space that would become time has
been like The Hunting of the Snark, Lewis Carroll’s mythical beast
that no one could nd. Since the idea of intrinsic time was rst
clearly formulated about thirty- ve years ago, the beast has not
been found. Karel has done more than anyone else to try to track it
down. If he cannot nd it, I feel that comes quite close to a non-
existence proof. My own belief is that the idea is based on an
incorrect notion of time. It is a mythical beast invoked in vain to
solve a titanic struggle. It does not surprise me that a special time
has not been found lurking in the tapestry of space-time. All I see in
that tapestry are change and di erences – and the di erences are
measured democratically. The idea of a special intrinsic time to be
extracted out of space, or out of any part of space-time or its
contents, violates the democratic theory of emphemeris time that
lies at the heart of general relativity.
 If we look at the Newtonian parallel of the notion, it seems
strange. In a world of three particles, it is like saying that one of the
sides of the triangle they form is time while the other two are true
degrees of freedom. Such an attempt to nd time breaks up the
unity of the universe. No astronomer observing a triple-star system
would begin to think like that. The key property of astronomical
ephemeris time is that all change contributes to the measure of
duration. There has to be a di erent way to think about time.

I believe it was found, perhaps unintentionally, by Bryce DeWitt

in 1967. John Wheeler had strongly urged him to nd the
fundamental equation of quantum gravity. It was Wheeler’s high
priority to nd the Schrödinger equation of geometrodynamics.
What the theory of intrinsic time should yield is a time-dependent
Schrödinger equation that – in gurative language – evolves a wave
function for ‘two-thirds of space’ with respect to a ‘time’ constituted
by the remaining ‘one-third of space’. Balking at the invidious task
of selecting which third should be ‘time’, DeWitt fell back on a very
general formalism developed fteen years earlier by Dirac that
made it possible to avoid having to make a choice.

Dirac’s method makes it possible to treat all parts of space on an

equal footing, and simply defers to later the problem of time.
DeWitt used Dirac’s method to write the fascinating equation that,
as Kuchaf noted, he himself calls ‘that damned equation’, John
Wheeler usually calls the ‘Einstein-Schrödinger equation’ and
everyone else calls the ‘Wheeler-DeWitt equation’. But what is this
equation, and what does it tell us about the nature of time?

The most direct and naive interpretation is that it is a stationary

Schrödinger equation for one xed value (zero) of the energy of the
universe. This, if true, is remarkable, for the Wheeler-DeWitt
equation must, by its nature, be the fundamental equation of the
universe. I pointed out in the discussion of the structure of
molecules that the ‘ball-and-strut’ models are only approximations
to the quantum description, being merely the most probable
con gurations. The Wheeler-DeWitt equation is telling us, in its
most direct interpretation, that the universe in its entirety is like
some huge molecule in a stationary state and that the di erent
possible con gurations of this ‘monster molecule’ are the Instants of
time. Quantum cosmology becomes the ultimate extension of the
theory of atomic structure, and simultaneously subsumes time.

We can go on to ask what this tells us about time. The

implications are as profound as they can be. Time does not exist.
There is just the furniture of the world that we call instants of time.
Something as nal as this should not be seen as unexpected. I see it
as the only simple and plausible outcome of the epic struggle
between the basic principles of quantum mechanics and general
relativity. For the one – in its standard form at least – needs a
de nite time, but the other denies it. How can theories with such
diametrically opposed claims coexist peacefully? They are like
children squabbling over a toy called time. Isn’t the most e ective
way to resolve such squabbles to remove the toy? We have already
seen that there is a well-de ned sense in which classical general
relativity is timeless. That is, I believe, the deepest truth that can be
read from its magical tapestry. The question then is whether we can
understand quantum mechanics and the existence of history without
time. That is what the rest of the book is about.
PART 5
History in the Timeless Universe

If things simply are, how can history be? If quantum cosmology is

merely a static mist that enshrouds eternal Platonia, whence the
manifest appearance of motion and our conviction history is real?
This is the great question. I am not going to give any summary of
Part 5: read on, please. The king sher just about to set o in ight
is a symbol for the task. Explaining how we see it in motion in a
timeless world is no more of a problem than explaining why we are
convinced that Henry VIII had six wives.
CHAPTER 17
The Philosophy of Timelessness

You should by now recognize the connection between the picture

that emerges from the simplest interpretation of the Wheeler-DeWitt
equation and the timeless world I sketched in Part 1. I outlined
there, using the notion of the time capsule, how the seemingly dead
and static Platonia might correspond to the vibrant living world we
experience in every instant. In this nal part of the book, I want to
explain the arguments from physics that led me to the notion of
time capsules, and also to show that the structure of quantum
cosmology may well cause the wave function of the universe to
‘seek out’ time capsules. This is the story of how physics brings the
Platonic forms to life. I start with some general comments.

I believe in a timeless universe for the childlike reason that time

cannot be seen – the emperor has no clothes. I believe that the
universe is static and is described by something like the Wheeler-
DeWitt equation. I would like you to accept this as a working
hypothesis, so we can see where it leads. As I said earlier, I believe
that it leads to the rules of creation. Let me now explain why.

According to many accounts, in both mainstream science and

religion, the universe either has existed for ever or was created in
the distant past. Creation in a primordial reball is now orthodox
science – the Big Bang. But why is it supposed that the universe was
created in the past rather than newly created in every instant that is
experienced? No two instants are identical. The things we nd in
one are not exactly the same as the things we nd in another. What,
then, is the justi cation for saying that something was created in the
past and that its existence has continued into the present?

The most obvious reason is the apparent persistence of objects

and living beings. If pressed, though, we acknowledge that they
never remain exactly the same. Even rocks weather slowly.
However, enough properties remain unchanged for us to say that
the same things do continue to exist. Indeed, human existence is
inconceivable without a signi cant degree of stability in the world.
No doubt the baby’s recognition of the continually reappearing
smiling face of its mother soon implants the notion of persistence.
But if we want to think rationally and as philosophers about these
matters, we ought to cultivate a degree of detachment. We must
practise Cartesian doubt and, just once at least, question all our
preconceptions.

I am not persuaded that the people who ought to be best at this –

theoretical physicists – do achieve full freedom of thought. Many
are passionately committed to an objectively existing external
world. They hate anything that smacks of solipsism or creationism.
This explains the controversies, virulent at times, about the reality
of atoms that took place a century ago, and the equally impassioned
debates today about the meaning of quantum mechanics (in many
ways a continuation of the debate about atoms). For scientists
committed to realism, atoms that remain the same in themselves
and merely move in space and time are very welcome. Atoms, space
and time are the things that either existed for ever or else came into
being with the Big Bang.

However, the elds introduced by Faraday and Maxwell now

provide the basis of quantum eld theory, which is currently the
deepest known form of quantum theory, and such elds are in
perpetual ux. And within classical physics Einstein made space and
time equally uid and transient. Today there is only one scienti c
justi cation for saying that the universe was created in the past: the
hypothesis of lawful dynamical evolution from some past, into the
present, and on into an as yet unexperienced future. If an initial
state uniquely determines a subsequent state of the same generic
kind which di ers only in detail, it is reasonable to speak of initial
creation and subsequent evolution.

But this view must be challenged. It belongs to a mindset that

holds the world either to be classical in its entirety, or to have
quantum objects within the old classical framework of space and
time. How slow we are to move out of old quarters! All the evidence
indicates that anything dynamical must obey the rules of quantum
mechanics even if it appears classical to our senses. But Einstein
made space dynamical – that is the lesson of geometrodynamics
taught us in detail by Dirac; by Arnowitt, Deser and Misner (ADM);
and by Baierlein, Sharpe and Wheeler (BSW). When space submits
to the quantum, as it surely must, the last vestige of a created but
persisting framework is lost. Moreover, the transition from the
classical world we see to the quantum world that underlies it is
xed in its broad outlines. All we need do is put together the two
things that go into quantization – a classical theory and the rules to
quantize it – and see what comes out.

The central insight is this. A classical theory that treats time in a

Machian manner can allow the universe only one value of its
energy. But then its quantum theory is singular – it can only have
one energy eigen-value. Since quantum dynamics of necessity has
more than one energy eigenvalue, quantum dynamics of the
universe is impossible. There can only be quantum statics. It’s as
simple as that!

In Part 1 I mentioned the dichotomy in physics between laws

and initial conditions. Most equations in physics do not by
themselves give complete information, they only put limits on what
is possible. To arrive at some de nite prediction, further conditions
are necessary. Neither Newton’s nor Einstein’s equations tell us why
the universe has its present form. They have to be augmented by
information about a past state. We could invoke a deity in the way
Einstein was wont, who goes through two steps in creating the
universe. First, laws are chosen, then an initial condition is added.
Many people have wondered whether this is a permanent condition
of physics.
 The stationary Schrödinger equation is quite di erent in this
respect. It obviously cannot have initial conditions, since it is a
timeless equation. It does not require boundary conditions, either.
Let me explain what this means. There are many equations in
physics which describe how quantities vary in space without there
being any change in time. Such equations can have many di erent
solutions, and to nd the one that is applicable in a speci c case,
mathematicians often stipulate the actual values the solution must
have at the boundary of some region. This stipulation is what is
called a boundary condition. Boundary conditions have the same kind
of importance as initial conditions. However, as explained in Box
13, the stationary Schrödinger equation requires no such conditions.
Instead, there is just a general condition on the way the wave
function behaves. It must be continuous (not make any jumps), it
must have only one value at each point and it must remain nite
everywhere. As we saw, the condition of remaining nite – of not
rushing o to in nity – is very powerful. It was what unlocked the
quantum treasure chest. In fact, the rst two conditions are also
very powerful and lead to many important results. To distinguish
these conditions from normal initial or boundary conditions, let me
call them conditions of being well behaved. Mathematicians may
regard this as somewhat arti cial, since the condition of remaining
nite does actually enforce a de nite kind of behaviour at
boundaries. It is therefore in some sense equivalent to a boundary
condition. However, I prefer not to think of it in that way, since it is
very general and can be formulated in a completely timeless
fashion. It avoids all particular speci cation, which must always be
arbitrary.

Now, my suggestion is this. There are no laws of nature, just one

law of the universe. There is no dichotomy in it – there is no
distinction between the law and supplementary initial or boundary
conditions. Just one, all-embracing static equation. We can call it
the universal equation. Its solutions (which may be one or many)
must merely be well behaved, in the sense explained in the previous
paragraph. It is an equation that creates structure as a rst principle,
just as the ordinary stationary Schrödinger equation creates atomic
and molecular structure. This is because it attaches a ranking – a
greater or lesser probability – to each conceivable static
con guration of the universe.

I explained in connection with Figure 40 how the density of the

blue mist can be used to create a collection of con gurations in a
bag, a heap even, from which the most probable atomic
con gurations can be drawn at random. Con gurations – which are
structures – are created as more or less de nite potentialities to the
extent that the stationary Schrödinger equation tells us to put more
or less into the heap. Like the individual structures within it, the
heap is static. It is carefully laid up in a Platonic palace, which,
since probabilities play such a mysterious role in quantum
mechanics, is a kind of ‘antechamber of Being’.

Now I can start to make good my deeper claims about

Schrödinger and creation. We have to forget all previous physics
and approach things with an open mind. First, we look at what the
Machian time-independent Schrödinger equation is and what it
does. It is completely self-contained. For a system of three bodies it
just works on triangles and masses, and nothing else. In a timeless
fashion, it associates a probability with each triangle. This is
tantamount to giving them a ranking. It is particularly suggestive
that this ranking is determined by the triangles themselves –
nothing else is involved. The probabilities for the triangles emerge
from a comprehensive testing and comparison programme. The
equation ‘looks’ at all possible wave functions that could exist on
Platonia and throws out all those that do not ‘resonate’ properly.
Those that are left have to be nely tuned, otherwise they will
satisfy neither the equation nor the condition of being well behaved.
And it is not just the wave function that resonates. We can say that
the triangles that get the greatest probability are the ones that
‘resonate best with their peers’, since the triangles alone determine
how the probability is distributed. This is what the rationality of
best matching in classical dynamics translates into in quantum
cosmology. There is a perfect, circle-closing, rational explanation for
all the relative probabilities.

I do believe that what we have here are putative rules of

creation, or perhaps we should say of being. Considered purely as an
intellectual exercise, this quantum-mechanical determination of
probabilities for relative con gurations is no odder than the
classical-dynamical determination of curves in con guration space.
The aim of science is to nd rational and economic explanations of
observed phenomena, not to prejudge the issue. Each hypothetical
scheme should be judged on its merits. There should be a clear
statement of the phenomena that are to be explained, the
conceptual entities that are to be employed and the mechanism that
is to yield the explanation.

The rst aim is to create a realist (non-solipsistic) cosmology in

which there are sentient beings whose primary awareness is of
structured instants of time as de ned earlier in the book. These
instants are like subjective snapshots, and may be called atoms of
perceptual existence. Each snapshot holds together in an
indissoluble unity everything that we would want to call the actual
facts of which we are aware in an instant of time. These include not
only the things we see, feel and hear, but also our awareness of
them, our memories and our interpretation of everything. The fact
that many di erent things are known at once is regarded (by me at
least) as the most remarkable – and de ning – property of instants
of time. I do not believe that science (or religion) will ever explain
why we experience instants, but perhaps it can explain the structure
we nd within them.

The scheme is realist because the structure of an external,

objectively existing real thing is being proposed as the explanation
of the structure experienced within a perceptual instant. What we
experience in subjective instants re ects, through psychophysical
parallelism, physical structure in external things: con gurations of
the universe. Their actual nature is a matter for ongoing research.
The notion has been illustrated by con gurations of mass points in
Euclidean space, by island-type distributions of elds of Faraday-
Maxwell type in Euclidean space, and by closed Riemannian 3-
geometries (which may also have elds de ned on them). It is at the
last level that I believe satisfactory explanations can in principle be
obtained for many of the known facts of physics and cosmology.
However, some further development, very possibly associated with
the notions of superstrings and supersymmetry, may well be needed
to explain the actual cocktail of forces and particles that pervades
the universe.

What is important about relative con gurations is that they are

intrinsically de ned – they are self-contained things – and that the
rule that de nes one thing simultaneously de nes many. Moreover,
they can all be arranged systematically in a relative con guration
space: Platonia, as I have called it.

Classical physics before general relativity ‘explained’ the world

by assuming it to be a four-dimensional history of such relative
con gurations located in a rigid external framework of absolute
space and time. Such a world is supposed to have evolved from
certain initial conditions to the state we now observe by means of
the laws of classical dynamics, in which the framework of space and
time play a signi cant role. These laws provide all the explanation
of which classical physics is in principle capable. In Part 2 I showed
how the external framework can be dispensed with. It does not need
to be invoked to formulate the laws of dynamics, nor even to
visualize how things are located in space and time. Schrödinger’s
Kantian appeal to space and time as the ineluctable forms of
thought was unnecessary. We can form a clear conception of
structured things that stand alone. We have seen how this is also
true of general relativity, in which space-time is ‘constructed’ by
tting together 3-spaces in a very re ned and sophisticated way.

So, then, what does the Wheeler-DeWitt equation tell us can

happen in a rational universe? The answer is ironic. Nothing! The
quantum universe just is. It is static. What a denouement. This is a
message that needs to be shouted from the rooftops. But how can
this seemingly bleak message reverberate around a static universe?
How can we bring dead leaves to life? The poet Shelley called on
the wild west wind to carry his thoughts over the universe. What
can play the role of the wind in static quantum Platonia?
CHAPTER 18
Static Dynamics and Time Capsules

DYNAMICS WITHOUT DYNAMICS

DeWitt already clearly saw the problem posed at the end of the last
chapter – the crass contradiction between a static quantum universe
and our direct experience of time and motion – and hinted at its
solution in 1967. Quantum correlations must do the job. Somehow
they must bring the world alive. I shall not go into the details of
DeWitt’s arguments, since he saw them only as a rst step. However,
the key idea of all that follows is contained in his paper. It is that the
static probability density obtained by solving the stationary
Schrödinger equation for one xed energy can exhibit the
correlations expected in a world that does evolve – classically or
quantum mechanically – in time. We can have the appearance of
dynamics without any actual dynamics.

It may surprise you, but it was about fteen years before

physicists, and then only a few, started to take this idea seriously.
The truth is that most scientists tend to work on concrete problems
within well-established programmes: few can a ord the luxury of
trying to create a new way of looking at the universe. A particular
problem in everything to do with quantum gravity is that direct
experimental testing is at present quite impossible because the scales
at which observable e ects are expected are so small.

Something like a regular research programme to recover the

appearance of time from a timeless world probably began with an
in uential paper by Don Page (a frequent collaborator of Stephen
Hawking) and William Wootters in 1983. This was followed by
several papers that concentrated on an obvious problem. In ordinary
laboratory physics, the fundamental equation used to describe
quantum phenomena is the time-dependent Schrödinger equation. It
undoubtedly holds to an extraordinarily good accuracy for all
ordinary physics: we could not even begin to understand, for
example, the radiation of atoms without this equation. But if the
universe as a whole is described by a stationary Schrödinger
equation and time does not exist at all, how does a Schrödinger
equation with time arise? This question seems to have been rst
addressed by the Russians V. Lapchinskii and V. Rubakov, but a
paper in 1985 by the American Tom Banks did more to catch the
imagination of physicists. This was followed in 1986 by a paper
treating the same problem by Stephen Hawking and his student
Jonathan Halliwell. Further papers on the subject appeared in the
following years. The whole associated research programme has
become known as the semiclassical approach, for a reason I shall
explain later. The basic idea is easy to grasp.

Imagine yourself on a wide sandy beach on which the receding

tide has left a static pattern of waves. As you are a free agent,
nothing can stop you from laying out a rectangular grid on the beach
and calling the direction along one axis ‘space’ and that along the
perpendicular axis ‘time’. For each value of the ‘time coordinate’, you
can examine the wave pattern along the one-dimensional line of
‘space’ at that ‘time’. When you move to the neighbouring line on the
beach corresponding to ‘space’ at a slightly later ‘time’, you will nd
that the wave pattern has changed. Simply by laying out your grid
and calling one direction ‘space’ and the other ‘time’, you have
transformed – in your mind’s eye – a two-dimensional static picture
into wave dynamics in one dimension. This can be done with wave
patterns in spaces of any dimension N. One direction can always be
called ‘time’, and this automatically creates ‘evolution’ in the
remaining N – 1 dimensions.

Of course, if the original wave pattern is ‘choppy’ and has not

been created by some rule, the choice of the ‘time’ direction will be
arbitrary. Any choice will create the impression of evolution in the
remaining N – 1 dimensions, but it will not obey any de nite and
simple law. In the semiclassical approach, there are two decisive
di erences from the arbitrary situation. First, the static wave pattern
is the solution of a de nite equation. Second, it is a somewhat
special solution – called a semiclassical solution – in that it exhibits a
more or less regular wave pattern. This assumption will be
considered later. However, if the wave pattern satis es the
assumption, it automatically selects a direction that it is natural to
call time. With respect to this direction, a genuine appearance of
dynamics arises in a static situation (Box 14). The result is this. Two
static wave patterns (in a space of arbitrarily many dimensions) can,
under the appropriate conditions, be interpreted as an evolution in
time of the kind expected in accordance with the time-dependent
Schrödinger equation. The appearance of time and evolution can
arise from timelessness.

BOX 14 The Semiclassical Approach

This box provides some necessary details about the semiclassical

approach. It is important here that the quantum wave function is not
one wave pattern but two (the red and green ‘mists’). I mentioned
the ‘tennis’ played between them – the rate of change in time of the
red mist is determined by the curvatures of the green mist, and vice
versa. This leads to the characteristic form of a momentum
eigenstate, in which both mists have perfectly regular wave
behaviour but with wave crests displaced relative to each other by a
quarter of a wavelength. If the red crests are a quarter of a
wavelength ahead of the green crests, the waves propagate in one
direction and the momentum is in that direction. If the red crests are
a quarter of a wavelength behind, the waves travel in the opposite
direction and the momentum is reversed. We can call this phase
locking. In a momentum eigenstate, there is perfect phase locking.

The semiclassical approach shows how two approximately phase-

locked static waves can mimic evolution described by the time-
dependent Schrödinger equation. In Figure 44 each of the two-
dimensional wave patterns is nearly sinusoidal, and they are
approximately phase-locked. These waves, being solutions of the
stationary Schrödinger equation, are static – they do not move. But
there is nothing to stop us (as in the example of the waves on the
beach) from calling the direction along the axis perpendicular to the
wave crests ‘time’ and the direction along the crests ‘space’.

Figure 44 Two nearly sinusoidal wave patterns.

The key step now is to divide the total pattern of each wave into
a regular part, corresponding to an imagined perfectly sinusoidal
behaviour, and a remainder that is the di erence between it and the
actual (nearly sinusoidal) behaviour. Call this the di erence pattern
(there is one for each mist). If the condition of approximate phase
locking holds, it turns out that the di erence patterns satisfy with
respect to our ‘space’ and ‘time’ an equation of the same form as the
time-dependent Schrödinger equation, except for the appearance of
one additional term. This term will have less and less importance,
though, the more closely the assumptions of the semiclassical
approach are satis ed.

In fact, the semiclassical approach o ers the prospect of an

explanation of time – in all its manifestations. It begins with a
uni ed concept of things. Each point of Platonia is one distinct
logically possible structure – it is one thing. The rules that make the
structures make everything. Platonia is entire and eternal. No place
in it is di erent from any other place, considered as something that
is logically possible. But each structure is still a distinct individual.
We see before us a true landscape whose every point is marked of
necessity by individuality. It has striking topographic features. So
there is a landscape, but nothing of a quite di erent nature that one
might call time.

There is, though, one quite di erent element: a wave function.

Schrödinger’s enigmatic ψ covers Platonia. Mist hovers over the
eternal landscape. The static mist is a well-behaved solution – an
eigenfunction – of the Wheeler-DeWitt equation. There is nothing
here an unsuspecting bystander could say looks like time. You have
seen mist on a landscape. Did it enter your head that such a thing
could explain time? But it can, in principle. The static wave function,
simply by its well-behaved response to the landscape it nds, may be
induced into a regular wave-like pattern. If so, time can ‘emerge’
from timelessness. We shall see how the wave function enables the
logically possible structures to interact – in a very real sense – with
each other, thereby helping each other into an actual existence that
seems to be deeply marked by time.

WHY DO WE THINK THE UNIVERSE IS EXPANDING?

This ‘marking with time’ brings us to the tricky part in the

semiclassical approach. It is what led me to the notion of time
capsules. This is a point at which my ideas part company from
(comparative) orthodoxy. Two closely related di culties convinced
me that a radical step was needed. The rst arises from a signi cant
di erence between the two Schrödinger equations. The complex
time-dependent equation is actually two equations for two separate
components – the red and the green mist. They play a kind of ‘tennis’
which tightly couples their behaviour and creates phase locking in
any semiclassical solution. In contrast, the stationary equation is
usually a real equation which does not couple the two components of
the wave function.

The existence of two separate yet almost perfectly matched wave

patterns is crucial in the semiclassical approach. The waves must be
parallel, and the wave crests displaced by a quarter of a wavelength.
In standard quantum mechanics this is a valid assumption. Indeed, it
is imposed because the true primary equation is the time-dependent
Schrödinger equation. The secondary stationary equation is just a
short cut to tell us the distribution of the blue mist without having to
nd the red and green mists rst. But they are there, and they are of
necessity phase-locked.
 But quantum cosmology gives us only the Wheeler-DeWitt
equation. It is the primary equation, but as it stands it will give only
a blue mist. We cannot assume some deeper equation hiding behind
it that will give phase-locked red and green mists. The truth is that
this part of the semiclassical approach assumes something that
should be derived. Luckily, this di culty threatens to undermine
only that part of the semiclassical approach in which the speci c
structure of the time-dependent Schrödinger equation is recovered.
The broad picture in which ‘time’ emerges from timelessness is not
threatened. In fact, complex numbers, which appear in my account
in the guise of the red and green mists, are so deeply ingrained in
quantum mechanics that I feel fairly con dent that this problem will
be sorted out. What is needed is some independent argument which
enforces the appearance of a complex wave function and a coupling
between its components. That would then ensure the necessary phase
locking.

Nevertheless, we must take care not to introduce inadvertently

into quantum cosmology assumptions that may be valid only in
ordinary quantum mechanics. This brings me to the second di culty
with the semiclassical approach. It concerns motion and our
conviction that we experience it, and simultaneously the issue of
where our sense of the passage of time comes from. To understand
the answer to this question is to understand time. It is all very well
for me to speak about static wave patterns in a mist that hangs over
Platonia. Such patterns will indeed, where they are su ciently
regular, de ne unambiguously a direction that may be called ‘time’.
But even if the wave pattern is rather regular, we could not look at it
and say that it distinguishes a direction of time. The one direction at
right angles to the wave crests will look the same whichever way we
face. There will be no signs set up on the distant horizons saying PAST
and FUTURE. This is the issue we must now address.

It will be helpful to think about what it is that determines which

way quantum wave packets move. Quantum mechanics is very
di erent from classical mechanics in this respect. A classical initial
condition consists of an initial position and an initial velocity. You
know which way a particle will move because it is speci ed in the
velocity. However, in quantum mechanics the initial condition is
simply the values of the two components of the wave function, the
red and green mists, everywhere at the initial time. Data like this
seem to correspond to giving only the position in classical
mechanics. Yet wave packets move under the rules Schrödinger
prescribed.

In fact, the way in which a wave packet will move is coded in the
relative positioning of the crests and troughs of the red and green
mists. We see this most clearly in momentum eigenstates. If the red
crests are ahead of the green crests, they go one way, but if the crest
positioning is reversed, they go the other way.

As we have seen, states very like momentum eigenstates play a

crucial role in the semiclassical approach. In all the original papers,
these states also played another role – they were used to model
situations corresponding to either expanding or contracting
universes. All physicists and astronomers are convinced that we live
in an expanding universe. There is certainly very good evidence in
many di erent forms to support this view. The formalism of
quantum cosmology must be capable of re ecting this aspect of the
observed universe. There must be something that codes expansion or
contraction of the universe or, rather (in the timeless interpretation),
codes the observed evidence that leads us to say the universe is
expanding.

All the models of Platonia that we have considered include a

dimension that we may call the ‘size’ of the universe. In fact, instead
of representing Triangle Land by means of the sides of triangles, we
could equally well – and more appropriately here – use two angles
(the third is found by subtracting their sum from 180°) and the area
of the triangles. The area is one direction, or dimension, in Triangle
Land. Expansion or contraction of the universe then corresponds to
motion along the line of increasing or decreasing size. The size
dimension begins at the point of zero size – what I have called Alpha,
or the centre of Platonia – and then proceeds all the way to in nity.

In the semiclassical approach, it was rather reasonably assumed

that the regular wave pattern needed for ‘time’ to emerge from
timelessness would develop along the direction of increasing or
decreasing size. This is a fair working hypothesis. What worried me
was the way in which expanding and contracting universes were
modelled – by analogy with momentum eigenstates in ordinary
quantum mechanics. Expansion or contraction were supposed to be
coded in the relative positions of wave crests.

It is certainly possible to imagine two static wave patterns – our

red and green mists – whose crests are perpendicular to lines that
seem to emanate from Alpha. This was done by nearly all the
researchers who used the semiclassical approach, and they assumed
that one relative positioning of the ‘red’ and ‘green’ crests would
model a universe expanding out of the Big Bang, while the opposite
positioning would model a universe headed for the Big Crunch (the
name given to one possible fate of the universe, in which it
recollapses to a state of in nite density and zero size). Thus,
momentum-like semiclassical states were used to achieve three
di erent things at once: the emergence of ‘time’, the recovery of the
time-dependent Schrödinger equation, and modelling expanding and
contracting universes. I believe that only the rst is soundly based. I
have some concern about the second. I think the third is de nitely
wrong.

The point is that the position of the ‘green crests’ ahead of or

behind the ‘red crests’ by itself has no signi cance. In ordinary
quantum mechanics the wave function depends not only on the
spatial position but also on the time. What really moves wave
packets is the relation of the time dependence to the space
dependence. It is not the case that if in some wave packet the green
crests are ahead of the red crests then the wave packet is bound to
move one way. This happens only because the time-dependent
Schrödinger equation is written in a particular form. But this is a
pure convention. All observed phenomena are described just as well
by an alternative choice, analogous to changing ends in tennis. The
two choices are identical in their consequences. They only di er in
the relative positions of the red and green crests, but this is o set by
reversing the time dependence. The real physics is unchanged.
Without the time dependence, the positions of the crests cannot
determine the direction of motion.

But this presents us with a real dilemma in static quantum

cosmology, in which there is no external time and no time
dependence to determine which way wave packets move. There is
simply no motion or change at all. We have to nd a di erent
explanation for why we think there is motion in the world and that
the universe expands.

One thing is clear: the origin of our belief that the universe is
expanding cannot be coded in the relative positioning of the crests of
the two waves, for the designations ‘red’ and ‘green’ are purely
conventional. The ‘colours’ could be swapped, and nothing
observable would change. The argument that mere static positioning
of crests can correspond to what we call expansion of the universe is
a chimera. This was clearly recognized in 1986 by my German
physicist friend Dieter Zeh, who commented that it has meaning only
if an absolute time exists. It really is necessary to think very
di erently about these things if time is abolished once and for all as
an independent element of reality.
THE IDEA OF TIME CAPSULES: THE KINGFISHER

From 1988 to 1991 I was absorbed by this issue. I became more and
more convinced that a decisive new idea was needed, but for a long
time could nd no answer that satis ed me. I formulated the
problem this way. I imagined myself watching some phenomena
involving motion in a very essential and vital way – a display of
acrobatics, say, or the ight of a king sher. I then imagined being
struck dead instantaneously and my ‘soul’ being carried down to a
kind of Plato’s cave. Here I would nd omniscient mathematicians
examining a model of Platonia all covered with these red, green and
blue quantum mists that I have asked you to conjure up in your
mind’s eye. They are examining the solution of the Wheeler-DeWitt
equation corresponding to the universe in which I had just been
taken from life. I then asked myself this: what precise thing in that
mysterious pattern of mists blanketing Platonia corresponds to my
being aware of seeing the king sher in ight? Where – in a timeless
static world – is the appearance of motion coded? Where can I see
the king sher’s colours ashing in the sunlight?

As we have noted, in standard quantum mechanics the

information about wave-packet motion is coded in the relative
positioning of the red and green mists. This was the questionable
assumption taken over in the semiclassical approach. However, there
is much more to quantum mechanics than just the wave function at
one instant (the pattern of red and green mists). We have already
seen how time is needed if such relationships are to be translated
into wave-packet motion. But even that is not enough, for the wave
function acquires de nite meaning only through prescriptions about
the measurements that will be made on the system. These take the
form of statements about the positions and construction of measuring
instruments that behave classically and are external to the quantum
system.

It is obvious that in quantum cosmology the whole superstructure

of an external time, and of measuring instruments outside the
considered system, must go. The instruments must be subsumed into
the quantum system (which becomes the complete universe), and we
must get to grips with a static wave function. Does this leave any
scope for making a connection between actual experiences and the
bare bones of embryonic quantum gravity as found by DeWitt?

I believe it does. Is not our most primitive experience always that

we seem to nd ourselves, in any instant, surrounded by objects in
de nite positions? Each experienced instant is thus of the nature of
an observation, a discovery, even – we establish where we are.
Moreover, what we observe is always a collection, or totality, of
things. We see many things at once. In fact, most humans, indeed
nearly all animals, have a wonderfully developed spatial awareness.
In writing this book I have relied heavily on you possessing this gift –
time and again I have asked you to imagine con gurations of the
universe as entities. They are all the places in Platonia.

When, therefore, I nd myself in Plato’s cave and see his demesne

of Platonia laid out before me, I can, using my vivid memory of the
king sher ashing between the banks of the stream where I stood,
identify the instant in which death took me. By ‘identify the instant’,
I mean recognize the con guration of riverbank, sunlight and
shadow, rippled water and king sher’s wings – all frozen in the
position I last witnessed. As always, I insist that instant of time simply
means con guration of the universe. This part of the problem of
nding a connection between the psychical experience and the
model of physical reality is relatively straightforward. There is little
or no problem in the representation of position.

The real problem, then, is in the representation of motion. We

seem to have exhausted all the resources of static quantum
cosmology simply to put everything into place on the riverbank.
Quantum mechanics does permit us to gain total information about
position, but only at the expense of total loss of information about
motion. We seem to have nothing left over to enable the king sher
to y. This is the crux of the matter. Classical physics presupposes
both positions and motions, matching our experience that we see
both at once. But quantum mechanics – in its present standard form
– has this curious halving of the accessible data.

So how can we let the king sher y? As few things delight me

more than a king sher in ight, this is a matter of some interest to
me. The answer that suddenly came to me in the summer of 1991
(which, of course, is a place in Platonia, not a time) was that the
ight of the king sher is ultimately an illusion, though it rests on
something that is very special and just as real as we take ight to be.
It is ight without ight. Let me return to the imagery of the blue
mist that shimmers over Platonia. It is easy to locate the instant of
my death – I see the point in that great con guration space in which
I stand on the bank of the stream. Now let me make an assumption
in the hallowed tradition of Boltzmann: only the probable is
experienced. The blue mist measures probability. Therefore, in
accordance with the tradition, the blue mist must shine brightly at
the point in Platonia in which I see the king sher frozen in ight
above the water. I experienced the scene, so it must have a high
probability. But there is still no motion.

I do not think there can be any. But there can be something else.
As I mentioned in Part 1, nobody really knows what it is in our
brains that corresponds to conscious experience. I make no pretence
to any expertise here, but it is well known that much processing goes
on in the brain and, employing normal temporal language, we can
con dently assert that what we seem to experience in one instant is
the product of the processing of data coming from a nite span of
time.

This is all I need. It enables me to make the working conjecture

that I outline in Part 1 – that when we think we see motion at some
instant, the underlying reality is that our brain at that instant
contains data corresponding to several di erent positions of the
object perceived to be in motion. My brain contains, at any one
instant, several ‘snapshots’ at once. The brain, through the way in
which it presents data to consciousness, somehow ‘plays the movie’
for me.

Down in Plato’s cave, thanks to the perfect representation of

everything that is, I can look more closely at the point in the model
of Platonia that contains me at the point of death. I can look into my
brain and see the state of all its neurones. And what do I see? I see,
coded in the neuronal patterns, six or seven snapshots of the
king sher just as they occurred in the ight I thought I saw. This
brain con guration, with its simultaneous coding of several
snapshots, nevertheless belongs to just one point of Platonia. Near it
are other points representing con gurations in which the correct
sequence of snapshots that give a king sher in ight is not present.
Either some of the snapshots are not there, or they are jumbled up in
the wrong order. There are in nitely many possiblilities, and they
are all there. They must be, since there is a place in Platonia for
everything that is logically possible.

Now, at all the corresponding points the blue mist will have a
certain intensity, for in principle the laws of quantum mechanics
allow the mist to seep into all the nooks and crannies of Platonia.
Indeed, the rst quantum commandment is that all possibilities must
be explored. But the laws that mandate exploration also say that the
blue mist will be very unevenly distributed. In some places it will be
so faint as to be almost invisible, even with the acuity of vision we
acquire in Plato’s cave for things mathematical. There will also be
points where it shines with the steely blue brilliance of Sirius – or the
king sher’s wings. And again my conjecture is this: the blue mist is
concentrated and particularly intense at the precise point in Platonia
in which my brain does contain those perfectly coordinated
‘snapshots’ of the king sher and I am conscious of seeing the bird in
ight.

As I explained in Chapter 2, a time capsule, as I de ne it, is in

itself perfectly static – it is, after all, one of Plato’s forms. However, it
is so highly structured that it creates the impression of motion. In the
chapters that follow, we shall see if there is any hope that static
quantum cosmology will concentrate the wave function of the
universe on time capsules. As logical possibilities, they are certainly
out there in Platonia. But will ψ nd them?
CHAPTER 19
Latent Histories and Wave Packets

SMOOTH WAVES AND CHOPPY SEAS

All interpretations of quantum mechanics face two main issues. First,

the theory implies the existence of far more ‘furniture’ in the world
than we see. I have suggested that the ‘missing furniture’ is simply
other instants of time that we cannot see because we experience only
one at a time. The other issue is why our experiences suggest so
strongly a macroscopic universe with a unique, almost classical
history. In the very process of creating wave mechanics, Schrödinger
found a most interesting connection between quantum and classical
physics that cast a great deal of light on this problem. The
interpretation he based on it was soon seen to be untenable, but it is
full of possibilities and continues to play an important role. It is the
starting point of other interpretations, including the one I advocate,
so I should like to say something about it.

In the 1820s and 1830s, William Rowan Hamilton, whom we

have already met, established a fascinating and beautiful connection
between the two great paradigms of physical thought of his time –
the wave theory of light and the Newtonian dynamics of particles.
Cornelius Lanczos, a friend of Einstein and author of the ne book
The Variational Principles of Mechanics, opens his chapter on these
things with a quotation from Exodus: ‘Put o thy shoes from o thy
feet, for the place whereon thou standest is holy ground.’ Let me
quote Lanczos – he is not exaggerating:

We have done considerable mountain climbing. Now we are

in the rare ed atmosphere of theories of excessive beauty
and we are nearing a high plateau on which geometry,
optics, mechanics, and wave mechanics meet on common
ground. Only concentrated thinking, and a considerable
amount of re-creation, will reveal the full beauty of our
subject in which the last word has not yet been spoken. We
start with … Hamilton’s own investigations in the realm of
geometrical optics and mechanics. The combination of these
two approaches leads to de Broglie’s and Schrödinger’s great
discoveries, and we come to the end of our journey.

The italics are mine. Lanczos’s account does end with

Schrödinger’s discoveries, but I think it can be taken one step
further. By the way, do not worry about the call for ‘concentrated
thinking’. If you have got this far, you will not fail now.

Hamilton made several separate discoveries, but the most

fundamental result is simple and easy to visualize. Two characteristic
situations are encountered in wave theory – ‘choppy’ waves, as on a
squally sea, and regular wave patterns. Hamilton was studying the
connection between Kepler’s early theory of light rays and the more
modern wave theory introduced by Young and Fresnel. Hamilton
assumed that light passing through lenses took the form of very
regular, almost plane waves of one frequency (Figure 45).

In optics, many phenomena can be explained by such waves. To

do this, we need to know how the wave crests are bent and how the
wave intensity, which is measured by the square of the wave
amplitude (Figure 45), varies. In general, when the wave is not very
regular, the ways in which the wave crests bend and the amplitude
varies are interconnected, and it is not possible to separate their
behaviour. However, as the behaviour gets more regular, the
amplitude changes less and simultaneously ceases to a ect the
bending of the wave crests. Hamilton found the equation that
governs the disposition of the wave crests in this case. Now known as
the eikonal equation, it is the foundation of all optical instruments –
microscopes, telescopes – and also electron microscopes. Indeed,
numerous e ects in optics are fully explained by the bending of the
wave crests. However, other phenomena, above all the di raction
and spreading of light when it passes through a small ori ce, can be
explained only by the full wave theory. In these phenomena the
regular pattern of wave crests is broken up.
 Figure 45 An example of a regular wave pattern, showing wave
crests and the lines that run at right angles to them. Such patterns
are characterized by two independent quantities – the wavelength
and the amplitude (the maximum height of the wave).

We shall stick to phenomena in which the wave crests remain

regular. Lines that run at right angles to such wave crests can be
de ned; they are easy to visualize (Figure 45). Hamilton’s work in
the 1820s showed that these lines correspond to the older idea of
light rays, and that there are two seemingly quite di erent ways of
explaining the behaviour of light and the functioning of optical
instruments. In the older, more primitive way, light is composed of
tiny particles (corpuscles) that travel along straight lines in empty
space, but are bent in air, water and optical instruments (made of
glass). The theory of light corpuscles works because the paths they
take, along Kepler’s light rays, coincide with the lines that run at
right angles to the wave crests. This is the second of Hamilton’s great
discoveries: if light is a wave phenomenon, there are nevertheless
many occasions in which it can be conceived as tiny particles that
travel along these rays.

This insight led to the distinction between wave optics and

geometrical optics, which uses light rays. Innumerable experiments
show that only wave optics, in which light is described by waves, can
explain certain phenomena. The earlier theory of light rays simply
fails under these circumstances. Equally, there are many cases in
which geometrical optics, with its Keplerian light rays, functions
perfectly well. We see here the typical situation that arises when a
new theory supplants an old one. The new theory invariably uses
very di erent concepts – it ‘inhabits a di erent world’ – yet it can
explain why the old theory worked as well as it did and why it is
that it fails where it does. Where the wave pattern becomes irregular,
geometrical optics ceases to be valid.

Geometrical optics shows how theories that explain many

phenomena impressively and simply can still give a misleading
picture. As my daughter learned on those frosty nights, this had
happened in ancient astronomy. Ptolemy’s epicycles gave a
beautifully simple and successful theory of planetary motion, but
were made redundant when Copernicus made the Earth mobile.
Geometrical optics is another classic example of a ‘right yet wrong’
theory. In fact, with its confrontation and reconciliation of seemingly
di erent worlds (particles and waves), it is one long, ongoing saga. It
started with Kepler’s optics, continued with the rival optical theories
of Newton (particles) and Huygens, Euler, Young and Fresnel (wave
theory), and reached a rst peak with Hamilton. It burst into life
again in 1905 with Einstein’s notion of the light quantum, then went
through another remarkable transformation in Schrödinger’s 1926
discovery of wave mechanics. I believe this saga has not yet run its
course, as I will explain in the next chapter.

Now we come to Hamilton’s next discovery – the explanation of

Fermat’s principle of least time, the idea that did more than anything
else to foster the development of the principle of least action.

HISTORY WITHOUT HISTORY

Figure 46 shows the wave crests of a light wave in a medium in

which the speed of light is the same in all directions but varies from
point to point, causing the wave crests to bend. The speed of light is
less where the crests are closer together. Obviously, if some particle
wanted to get from A to F in the least time, travelling always at the
local speed of light, it would follow the curve ABCDEF. The
individual segments of this curve are always perpendicular to the
wave crests, and any deviation would result in a longer travel time.
But this is also exactly the route a light ray would follow, cutting the
wave crests at right angles. This was another great discovery that
Hamilton made – that when geometrical optics holds, the wave
theory of light can explain both Fermat’s principle of least time and
Kepler’s light rays. The rays follow the lines of least travel time, and
these are simultaneously the lines that always run perpendicular to
the wave crests.

One of the most interesting things about geometrical optics is a

connection it establishes with particles in Newtonian mechanics. The
characteristic property of a moving particle is that it traces out a
path through space. When regular wave patterns are present, wave
theory creates similar one-dimensional tracks without any particles
being present at all – the tracks of the light rays. Of course, in a strict
wave theory the rays are not really ‘there’, but they are present as
theoretical constructs. And many phenomena can be explained rather
well by assuming that particles really are there. As John Wheeler
would say, one has ‘particles without particles’, or even ‘histories
without histories’.
 Figure 46 The explanation of Fermat’s principle of least time under
the conditions of a regular wave pattern, so that geometrical optics
holds.

In fact, work that Hamilton did about ten years after his optical
discoveries shows how apt such a ‘Wheelerism’ is. As we saw in Part
2, classical physics is the story of paths in con guration spaces. They
are Newtonian histories. Hamilton thought about what would
happen if for them only one value of the energy is allowed, and
made a remarkable discovery. He found that just as light rays, which
are paths, arise from the wave theory of light when there is a regular
wave pattern, the paths of Newtonian dynamical systems can arise in
a similar fashion. I need to spell this out.
 Working entirely within the framework of Newtonian dynamics,
Hamilton introduced something he called the principal function. All
you need to know about this function is that it is like the mists on
con guration space: at each point of the con guration space, it has a
value (intensity), the variation of which is governed by a de nite
equation. Hamilton showed that when, as can happen, the intensity
forms a regular wave pattern, the family of paths that run at right
angles to its crests are Newtonian histories which all have the same
energy. They are not all the histories that have that energy, but they
are a large family of them. Each regular wave pattern gives rise to a
di erent family. Hamilton also found that the equation that governs
the disposition of the wave crests, which in turn determine the
Newtonian histories, has the same basic form as the analogous
eikonal equation in optics. But whereas that equation operates in
ordinary three-dimensional space, this new equation operates in a
multidimensional con guration space.

Many physicists have wondered how the beautiful variational

principles of classical physics arise. Hamilton’s work suggests an
explanation. If the principle that underlies the world is some kind of
wave phenomenon, then, wherever the wave falls into a regular
pattern, paths that look like classical dynamical histories will emerge
naturally. For this reason, waves that exhibit regular behaviour are
called semiclassical. This is because of the close connection between
such wave patterns and classical Newtonian physics. It also explains
the name of the programme discussed in the previous chapter.
 All the things that this book has been about are now beginning to
come together. A review of the essential points may help. We started
with Newton’s three-dimensional absolute space and the ow of
absolute time. History is created by particles moving in that arena.
Then we considered Platonia, a space with a huge number of
dimensions, each point of which corresponds to one relative
con guration of all the particles in the Newtonian arena. The great
advantage of the concept of a con guration space, of which Platonia
is an example, is that all possible histories can be imagined as paths.
There are two ways of looking at the single Newtonian history that
was believed to describe our universe. The rst is as a spot of light
that wanders along one path through Platonia as time ows. The
spot is the image of a moving present. In the alternative view, there
is neither time nor moving spot. There is simply the timeless path,
which we can imagine highlighted by paint. Newtonian physics
allows many paths. Why just one should be highlighted is a mystery.
We have also seen that only those Newtonian paths with zero energy
and angular momentum arise naturally in Platonia.

Hamilton’s studies opened up a new way to think about such

paths. It works if the energy has one xed value, which may be zero,
and introduces a kind of mist that covers the con guration space
with, in general, variable intensity. In those regions in which the
mist happens to fall into a pattern with regular wave crests, there
automatically arise a whole family of paths which all look like
Newtonian histories. They are the paths that run at right angles to
the wave crests. If you were some god come on a visit to the
con guration space and could see these wave crests laid out over its
landscape, you could start at some point and follow the unique path
through the point that the wave crests determine. You would nd
yourself walking along a Newtonian history. However, your starting
point, and the path that goes with it, would have to be chosen
arbitrarily, because precisely when the pattern of wave crests
becomes regular, the wave intensity (determined by the square of the
wave amplitude) becomes uniform. There would be nothing in the
wave intensity to suggest that you should go to one point or another.

Hamilton’s work opens up a way to reconcile contradictory

pictures of the world. Quantum mechanics and the Wheeler-DeWitt
equation suggest that reality is a static mist that covers Platonia. But
all our personal experience and evidence we nd throughout the
universe speak to us with great insistence of the existence of a past –
history – and a eeting present. The paths that can be followed
anywhere in Platonia where the mist does form a regular wave
pattern can be seen as histories, present at least as latent
possibilities.

I feel sure that the mystery of our deep sense and awareness of
history can be unravelled from the timeless mists of Platonia through
the latent histories that Hamilton showed can be there. But just how
is the connection to be made? In the remainder of this chapter I shall
explain Schrödinger’s valiant, illuminating, but unsuccessful attempt
to manufacture a unique history out of Hamilton’s many latent
histories. Then, in the next chapter, I shall consider the alternative –
that all histories are present.

AIRY NOTHING AND A LOCAL HABITATION

When Schrödinger discovered wave mechanics he was well aware of

Hamilton’s work, since de Broglie had used the deep and curious
connection between wave theory and particle mechanics in his own
proposal. De Broglie’s genius was to suggest that Hamilton’s
principal function was not just an auxiliary mathematical construct
but a real physical wave eld that actually guided a particle by
forcing it to run perpendicular to the wave crests. Schrödinger
sought to exploit Hamilton’s work somewhat di erently. His instinct
was to interpret the wave function as some real physical thing – say,
charge density. Of course, this could not be concentrated at a point,
since its behaviour was governed by a wave equation, and waves are
by nature spread out. Nevertheless, Schrödinger initially believed
that his wave theory would permit relatively concentrated
distributions to hold together inde nitely and move like a particle.
His work led to the very fruitful notion of wave packets. These can
be constructed using the most regular wave patterns of all – plane
waves like the example in Figure 45. A plane wave has a direction of
propagation and a de nite wavelength. All the lines that run
perpendicular to the wave crests are then latent, or potential, particle
‘trajectories’.
 Because the Schrödinger equation has the vital property of
linearity mentioned earlier, we can always add two or more solutions
and get another. In particular, we can add plane waves. Although
each separate solution is a regular wave throughout space, when the
solutions are added the interference between them can create
surprising patterns. This makes possible the beautiful construction of
Schrödinger’s wave packets (Box 15).

BOX 15 Static Wave Packets

A wave with its latent classical histories perpendicular to the wave

crests is shown at the top of Figure 47. Using the linearity, we add an
identical wave with crests inclined by 5° to the original wave. The
lower part of the computer-generated diagram shows the resulting
probability density (blue mist). The superposition of the inclined
waves has a dramatic e ect. Ridges parallel to the bisector of the
angle between them (i.e. nearly perpendicular to the original wave
elds) appear, and start to ‘highlight’ the latent histories. In fact,
these emergent ridges are the interference fringes that show up in
the two-slit experiment (Box 11), in which two nearly plane waves
are superimposed at a small angle, and also in Young’s illustration of
interference (Figure 22).

Much more dramatic things happen if we add many waves,

especially if they all have a crest (are in phase) at the same point. At
that point all the waves add constructively, and a ‘spike’ of
probability density begins to form. At other points the waves
sometimes add constructively, though to a lesser extent, and
sometimes destructively. Wave patterns like those shown in Figure
48 are obtained.

Figure 47 If two inclined but otherwise identical plane waves like

the one at the top are added, the gure at the bottom is obtained.
The ridges run along the direction of the light rays’ in the original
plane waves. (The top gure shows the amplitude, the bottom the
 square of the added waves, since in quantum mechanics that
measures the probability density.)

Figure 48 brings to mind a passage in A Midsummer Night’s Dream

that has haunted poets for centures:

And, as imagination bodies forth

The forms of things unknown, the poet’s pen
Turns them to shapes, and gives to airy nothing
A local habitation, and a name.

The intersection of two wave elds does not result in any

distinguished point, just a eld of parallel ridges. There is no ‘local
habitation’. But if the crests of three or more waves intersect at a
common point – so that the waves are in phase there – and their
amplitudes are varied appropriately, then a point becomes
distinguished. A localized ‘blob’ is formed. As Schrödinger realized
with growing excitement in the winter of 1925/6, this begins to look
like a particle.
 Figure 48 These wave patterns are obtained (from the bottom
upward) by adding increasing numbers of plane waves oriented
within a small range of directions. All waves have a crest where the
‘spike’ rises from the ‘choppy’ pattern. Their amplitudes also vary in
a range, since otherwise ‘ridges’ like those in Figure 47 are obtained.
 The pièce de résistance is nally achieved if the waves of
di erent wavelengths move and do so with di erent speeds. This
often happens in nature. In most media – above all in vacuum – light
waves all propagate with the same speed. However, in some media
the waves of di erent wavelengths travel at di erent speeds. Since
waves of di erent wavelengths have di erent colours, this can give
rise to beautiful chromatic e ects. In quantum mechanics, the waves
associated with ordinary matter particles like electrons, protons and
neutrons always propagate at di erent speeds, depending on their
wavelengths. The relationship between the wavelength and the speed
of propagation is called their dispersion relation.

Figure 49 has been constructed using such a dispersion relation.

The initial ‘spike’ (wave packet) at the bottom is the superposition of
waves of di erent angles in a small range of wavelengths. The
dispersion relation makes each wave in the superposition move at a
di erent speed. At the initial time, the waves are all in phase at the
position of the ‘spike’, but the position at which all the waves are in
phase moves as the waves move. The ‘spike’ moves! Its positions are
shown at three times (earliest at the bottom, last at the top). This
wave packet disperses quite rapidly because relatively few waves
have been used in its construction. In theoretical quantum
mechanics, one often constructs so-called Gaussian wave packets,
which contain in nitely many waves all perfectly matched to
produce a concentrated wave packet. These persist for longer.
 It is a remarkable fact about waves in general and quantum
mechanics in particular that the wave packet moves with a de nite
speed, which is known as the group velocity and is determined by the
dispersion law. It is quite di erent from the velocities of any of the
individual waves that form the packet. Only when there is no
dispersion and all the waves travel at the same speed is the velocity
of the packet the same as the speed of propagation of the waves.
These remarkable purely mathematical facts about superposition of
waves were well known to Schrödinger at the time he made his great
discoveries – one of which was that this beautiful mathematics
seemed to be manifested in nature.

SCHRÖDINGER’S HEROIC FAILURE

This led him to propose the wave-packet interpretation of quantum

mechanics. His main concern was to show how a theory based on
waves could nevertheless create particle-like formations. A potential
strength of his proposal was that particle-like behaviour could be
expected only above a certain scale. Over short enough distances,
within atoms or in colliding wave packets, the full wave theory
would have to be used, but in many circumstances it seemed that
particles should be present. With total clarity, which shines through
his marvellous second paper on wave mechanics, he saw that if
particles are associated with waves, then in atomic physics we must
expect an exact parallel with geometrical optics. There will be many
circumstances in which ordinary Newtonian particles seem to be
present, but in the interior of atoms, for example, where the
potential changes rapidly, we shall have to use the full wave theory.
Schrödinger’s second paper contains wonderful insights.

Figure 49 A moving wave packet obtained by adding plane waves

having slightly di erent orientations, wavelengths and propagation
velocities. The initially sharply peaked packet disperses quite
quickly, as shown in the two upper gures.
 Unfortunately, his idea soon ran into di culties. He had been
aware of one from the start. For a single particle, the con guration
space is ordinary space, and the idea that the wave function
represents charge density makes sense. But he was well aware that
his wave function was really de ned for a system of particles and
therefore had a di erent value for each con guration of them. I
highlighted this earlier by imagining ‘wave-function meters’ in a
room which showed the e ect of moving individual atoms in models
of molecules. It is di cult to see how the wave function can be
associated with the charge density of a single particle in space.

Another problem actually killed the proposal. Although wave

packets do travel as if they were a particle, they do spread. This was
the one e ect that Schrödinger failed to grasp initially. He actually
did detailed and beautiful calculations for one special case – the two-
dimensional harmonic oscillator, or conical pendulum. If a lead bob
suspended on a weightless thread is pulled to one side and released,
it will swing backwards and forwards like an ordinary pendulum.
However, if it is given a sidewise jolt as well it will trace out an
ellipse.

Schrödinger was able to show that for the quantum states

corresponding to large ellipses it is possible to form wave packets
that do not spread at all – the wave packets track round the ellipse
for ever. This was a truly lovely piece of work, but misleading.
Murphy’s law tripped up Schrödinger. The harmonic oscillator is
exceptional and is essentially the only system for which wave packets
hold together inde nitely. In all other cases they spread, doing so
rapidly for atomic particles. This doomed the idea of explaining
particle-like behaviour by the persistence of wave packets, as
Heisenberg noted with some satisfaction. (Most of the founding
fathers of quantum mechanics defended their own particular
directions with great fervour. Schrödinger hated quantum jumps, and
found the extreme abstraction of Heisenberg’s matrix mechanics
‘positively repulsive’.)

However, the notion of a wave packet is beautiful and

transparent, and has been widely and e ectively used. This has
tended to make people think that Schrödinger’s original idea was still
to a large degree right, and does explain why classical particle-like
behaviour (restricted in its accuracy only by the Heisenberg
uncertainty relation) is so often observed, especially in macroscopic
bodies. There is one great di culty, though. We can construct wave
packets with strongly expressed particle-like properties, but we have
to superimpose many di erent semiclassical solutions in just the
right way. There must be a relatively small range of directions and
wavelengths, adjustment of the wave amplitudes and, above all,
coincidence of the phases of all waves at one point. Nothing in the
formalism of quantum mechanics explains how this miraculous pre-
established harmony should occur in nature. A single semiclassical
solution might well arise spontaneously and naturally. But that will
be associated with a whole family of classical trajectories, which
exist only as formal constructions – they are at best latent histories.
Quantum mechanics generally gives a wave function spread out in a
uniform regular manner. Even if by some miracle we could
‘manufacture’ some wave packet, it would inevitably spread. Some
further decisive idea is needed to explain how a universe described
by quantum mechanics appears so classical and unique.
CHAPTER 20
The Creation of Records

HISTORY AND RECORDS

In Newtonian physics the notion of history is clear cut. It is a path, a

unique sequence of states, through a con guration space. This
picture is undermined in relativity and severely threatened in
quantum mechanics, since the wave function in principle covers the
complete con guration space. Almost all interpretations of quantum
mechanics seek to recover a notion of history by creating or
identifying in some way paths through the con guration space which
are then candidates for the unique history that we seem to
experience. This is a di cult and delicate exercise, since such paths
simply do not belong to the basic quantum concepts. The methods
used are quite varied, but they come in four main categories: the
basic equations of quantum mechanics are modi ed (by ad hoc
collapse of the wave function in the Copenhagen interpretation and
by spontaneous physical collapse in some other interpretations); the
equations are not changed but very special solutions are constructed
(as Schrödinger attempted); extra elements are added to the quantum
formalism (in so-called hidden-variable theories); or the equations
and their solutions are accepted in full but it is asserted that the
solutions in reality represent many parallel histories (Everett’s many-
worlds interpretation). None of these approaches is free of severe
problems, some of which I have mentioned.

I suspect that the main di culties arise because an important

aspect of history has been ignored. Even if history is a unique
succession of instants, modelled by a path in con guration space, it
can be studied only through records, since historians are not present
in the past. This aspect of history is not captured at all by a path. All
the solutions of a Newtonian system correspond to unique paths, but
they very seldom resemble the one history we do experience, in
which records of earlier instants are contained in the present instant.
This simply does not happen in general in Newtonian physics, which
has no inbuilt mechanism to ensure that records are created. It is a
story of innumerable histories but virtually no records of them. (1
discussed this at the end of Chapter 1.)

In thinking about history, I believe we should reverse the

priorities. Up to now the priority has been to achieve successions of
states and to assume that records will somehow form. But nothing in
the mechanisms that create successions ensures that records of them
will be created. Now a record is a con guration with a special
structure. Quantum mechanics, by its very construction, makes
statements about con gurations: some are more probable than
others. This is especially apparent in the quantum mechanics of the
stationary states of atoms and molecules. It determines their
characteristic structures. In contrast, there is no way that quantum
mechanics can be naturally made to make statements about histories.
It is just not that kind of theory.

It is also interesting that classical physics makes only one crude

distinction. Either a history is possible because it satis es the
relevant laws, or it is impossible because it does not. The possible
continuous curves in the con guration space are divided into a tiny
fraction that are allowed and the hugely preponderant fraction that
are not. It is yes or no. Quantum mechanics is much more re ned: all
con gurations are allowed, but some are more probable than others.
By its very nature, quantum mechanics selects special con gurations
– those that are the most probable. This opens up the possibility that
records, which are special con gurations by virtue of their structure,
are somehow selected by quantum mechanics. This is the possibility I
want to explore in this and the following chapter. The aim is to show
that quantum mechanics could create a powerful impression of
history by direct selection of special con gurations that happen to be
time capsules and therefore appear to be records of history. There
will be a sense in which the history is there, but the time capsule,
which appears to be its record, will be the more fundamental
concept.

THE CREATION OF RECORDS: FIRST MECHANISM

In the same conference in Oxford in 1980 at which Karel Kuchař

spoke about time in quantum gravity, John Bell gave a talk entitled
‘Quantum mechanics for cosmologists’. Among other things, he
considered how records arise. This led him to describe a
cosmological interpretation of quantum mechanics in which there
are records of histories but no actual histories. Perhaps not
surprisingly he rejected this as too implausible, but his account of
how records arise is most illuminating. I shall reproduce it here in
somewhat di erent terms, and then use it to propose an
interpretation that is quite close though not identical to his, since
Bell still assumed that the wave function of the universe would
evolve with time. If this assumption is removed, as I believe it must
be, Bell’s interpretation becomes less implausible.

Bell illustrated how records are created in quantum mechanics by

showing how elementary particles make tracks in detection devices.
The essential principles had already been published, by Nevill Mott
in 1929 and Heisenberg in 1930. As far as I am concerned, their
work is more or less the interpretation of quantum mechanics, but
surprisingly few people know about it.

It was stimulated by the Russian physicist George Gamow’s

theory of radioactive decay, put forward in 1928, in which alpha
particles escape from radium nuclei by a process called tunnelling.
The only detail we need to know is that Gamow represented an
escaping alpha particle by means of an expanding, spherical wave
function surrounding a radium nucleus. In accordance with the
standard quantum interpretation, there is then a uniform density of
the probability for nding the alpha particle all round the nucleus. In
my pictorial analogy, blue mist spreads uniformly from the nucleus.
 In those days, alpha particles were observed in devices called
Wilson cloud chambers through their interaction with atoms, which
they ionize by dislodging electrons, leaving the previously neutral
atoms positively charged. The alpha particles invariably ionize atoms
that lie more or less along a straight line emanating from the
radioactive source. The excess positive charge of the ionized atoms
stimulates vapour condensation around them, making the tracks
visible. If we take Gamow’s theory literally, there is something
deeply mysterious about these tracks. If there really is a blue
probability mist spreading out spherically all round the radium atom,
why are atoms not ionized at random all over the chamber, wherever
the blue mist permeates? How come they are ionized only along one
line?

Standard quantum mechanics gives two answers, one much

cruder than the other. In the crude answer (which is nevertheless
very interesting, so I shall take a few pages to discuss it), only the
alpha particle is treated in quantum-mechanical terms: the atoms of
the cloud chamber are treated as classical external measuring
instruments. They are used to ‘measure the position’ of the alpha
particle, this being done by the ionization of an atom at some
position. In accordance with the standard rules, any position
measurement yields a unique position, after which the wave function
will be concentrated at that position. The rest of the wave function
will be instantaneously destroyed.
 Now, atoms actually have a nite diameter, of about 10–8
centimetres. So the ionization of an atom is not a perfect position
measurement, and this has important consequences for the alpha-
particle tracks. It is helpful to think in terms of the blue mist. Before
the measuring ionization happens, the blue mist is expanding
outwards uniformly in all directions. When the rst ionization
occurs, it is as if a spherical shell has suddenly been placed round the
atom. At one point on the shell there is a small hole through which
the wave function can pass. This is the point at which the ionized
atom is situated. It is only here that the wave function is not totally
destroyed and can continue streaming on outwards. In fact, it does so
in the form of a jet, which can be very narrow and accurately
directed, especially if the alpha particle has a high energy.

At this point it is worth saying something about the di raction of

light. If monochromatic light (light of one wavelength) encounters an
opaque screen with an opening, the result depends on its size. If the
opening is large compared with the wavelength of the light, the
screen cuts o all the light except at the opening and a more or less
perfect ‘pencil’ of light – a beam – passes through. The width of the
luminous pencil is equal to the width of the opening. However, if the
opening is made smaller, di raction comes into play and the beam of
light spreads out, becoming very di use for a tiny opening.
Di raction e ects are more pronounced for red light, with its longer
wavelength, than violet light. Like light, alpha particles have an
associated wavelength, which is very short for the ones produced in
radioactive decay. Although ionization of the atom creates e ectively
a very small ‘opening’, the ‘jet of wave function’ that survives the
wave-function collapse is narrow and concentrated in a cone with a
very small opening angle (much less than a degree). The wave-
function jet continues through the cloud chamber like a searchlight
beam.

To simplify things, imagine that the cloud-chamber atoms are

concentrated on uniformly spaced, spherical concentric shells
surrounding the radium atom. The rst ionization (quantum
measurement and collapse) happens when the alpha particle’s
spherical waves reach the rst shell. On the second shell, the alpha
particle can ionize atoms only where its wave function has non-
vanishing value. The atoms that can be ionized are located in the
small spot that is ‘lit up’ by the ‘beam’ and hence lie rather
accurately on the line joining the radium atom to the ‘opening’ in the
rst shell. The spot still contains many hundreds or thousands of
atoms, any one of which can now be ionized. A second position
‘measurement’ of the alpha particle is about to be made.

The quantum measurement laws now tell us that one and only
one of the atoms will be ionized. It is selected by pure chance – it
can be anywhere in the spot. Once again, the entire wave function
that ‘bathes’ the other atoms is instantly destroyed, and a new
narrow beam continues outward from the second ionized atom. The
same process of ionization, collapse and ‘jet formation’ is repeated at
each successive shell. For an alpha particle with su cient energy,
this may happen hundreds or even thousands of times. A track is
formed. It has some important features.

First, although it is nearly straight, there are small de ections at

nearly all ionizations. It should not be supposed that the de ection
occurs where the kink in the track suggests it did. This subtlety is
illustrated in Figure 50. At each ionization and collapse a new cone
of the wave function is created. It is not unlil the next ionization
occurs that any actual de ection angle is selected. Until then, the
complete cone of de ection angles is potentially present. As
Heisenberg put it in a famous remark, the track is created solely by
the fact that we observe the particle.

Second, quantum mechanics makes no predictions about the

individual de ection angles. It merely predicts their statistical
distribution, according to a law found by Max Born a few months
after Schrödinger had created wave mechanics. Its form is
determined by the structure of the atoms on which the scattering
(de ection) of the alpha particle occurs. It is normally veri ed by
making experiments with many di erent alpha particles, the
statistical distribution being built up by the repetition of many
experiments over time. However, in principle it is possible to test the
statistical predictions on a single track, especially if it contains
thousands of ionizations.
 Figure 50 The creation of an alpha-particle track by successive
ionizations. After each ionization a wave-function beam spreads out,
but it is not until the next ionization occurs that the ‘kink’ is created.

Third, at each ionization the alpha particle loses a fraction of its

energy, typically about one part in ten thousand. Since the energy is
related to the particle’s wavelength, it becomes progressively longer
along the track. Just as di raction e ects are more pronounced for
red than for violet light, this means that the de ection angles get
progressively larger along the track. The nature of the track changes
along its length – it starts to show quite large zigzags.

Bell comments on this rst account of track formation that it ‘may

seem very crude. Yet in an important sense it is an accurate model of
all applications of quantum mechanics.’ Before we consider the
second – in nitely more illuminating – account, we need to draw
some conclusions and start to develop new ways of thinking about
things, above all history.
THE PREREQUISITES OF HISTORY

The central question of this fth part of the book is this: whence
history?

What light does Bell’s rst account cast on this question? What
are the essential elements that go into the creation of history? Bell’s
analysis promises to give us real answers to these questions, since an
alpha-particle track can truly be seen as prototypical history. All the
elements are there – a unique succession of events, a coherent story
and qualitative change as it progresses. It even models birth – when
the particle escapes from the radium atom – and demise – when it
nally comes to rest. It literally staggers to its death. The laws that
govern the unfolding of history are beautifully transparent. They
combine, in an intriguing way, causal development – the forward
thrust of the track – with unpredictable twists and turns governed
only by probability. History is created by what looks like a curious
mixture of classical and quantum mechanics – the continuous track
and the twists and turns, respectively.

Three distinct factors together create history in this rst account.

First, the alpha particle emerges from the radium atom in a state that
matches geometrical optics. Its wave function propagates outward in
perfectly spherical waves of an extremely regular shape and with a
very high frequency and short wavelength. This is a perfect example
of a semiclassical solution. Hamilton’s ‘light rays’ are the tracks that
run radially outward from the radium atom, always perpendicular to
the wave-function crests. Each of these tracks is a good simpli ed
model of the one solitary track that eventually emerges.

I mentioned the ongoing saga of geometrical optics. Schrödinger

attempted to create history by superimposing many slightly di erent
semiclassical solutions in a wave packet that mimicked particle
motion. We can now see that this attempt was doomed to failure,
mainly because it attempted to create particle tracks using the
quantum-mechanical properties of just one particle in isolation. The
interaction of the particle with the environment played no role in
Schrödinger’s attempt, but is crucial in the account just given. We
cannot begin to think of a track being formed without the atoms
waiting to be ionized. Geometrical optics still plays a vital role
because the very special semiclassical state ensures that sharply
de ned beams are created by the process of ionization and collapse.

We no longer need many semiclassical solutions: one

semiclassical solution is now su cient to create one history.
Nevertheless, at least one semiclassical solution remains – and will
remain – the prerequisite for history. The core mathematical fact
discovered by Hamilton keeps reappearing and being used in
di erent ways. I feel sure that this is the true deep origin of history –
we have already seen alpha-particle tracks form before our eyes.
Watch a little longer, and even Henry VIII and his six wives will
appear.

The second element in Bell’s account is collapse: crude, but

e ective. Little more needs to be said except that it is hard to believe
that nature can behave so oddly. However, Bell’s down-to-earth
account does show up the arti ciality of the quantum measurement
rules. These are formulated for individual observables, and insist that
measurement invariably results in the nding of a single eigenvalue
of a chosen observable. But in the case of the alpha particle ionizing
an atom, no pure measurement results – there is simultaneous
measurement of both position and momentum (both with imperfect
accuracy, so that the uncertainty relation is not violated).

The third element in the creation of history is low entropy: the

initial state of the system is highly special. The alpha particle, which
could be anywhere, is inside the radioactive nucleus; the countless
billions of cloud-chamber atoms, which could be in innumerable
di erent excited states, are all in their ground states. The only reason
we are not amazed by such order is our familiarity with the special.
What we have known from childhood ceases to surprise us. But even
the experiencing of coherent thoughts is most improbable. Among all
possible worlds, the dull, disordered, incoherent states are
overwhelmingly preponderant, while the ordered states form a
miniscule fraction. But such states, sheer implausibility, must be
presupposed if history is to be made manifest – at least it is in the
normal view of things.

The initial ordered state creates history and a stable canvas on

which it can be painted. The special position of the alpha particle
gives rise to its semiclassical state. The thousand or so atoms it
ionizes stand out as a vivid track on the un-ionized billions.
Photographed before dispersal, the track becomes a record of history.
If a large proportion of the atoms were already ionized, such a track
could hardly form, let alone stand out. We might claim that history
had unfolded, but there would be no evidence of it.

Records are all we have. We have seen one account of their

creation. Except for quantum collapse, it does not seem outlandish.
But Bell gives a second, fully quantum account in which the
monstrously multidimensional con guration space of the cloud
chamber is vital. This story of history is amazing. The next section
prepares for it.

THE IMPROBABILITY OF HISTORY

The cloud chamber is treated schematically as a collection of

hydrogen atoms, each consisting of a nucleus – a single proton – and
an electron. We ignore the fact (here not an issue) that all protons
are identical, and so are all electrons. It is also reasonable to assume
that the protons are at xed points, and to treat only the electrons
and the alpha particle quantum mechanically. The coordinates of
each electron can be three mutually perpendicular distances from its
proton. A real cloud chamber may contain 1027 atoms. It is daunting
to contemplate a space with 3 × 1027 (+ 3 for the alpha particle)
dimensions, but we must do our best if we are to get a true feeling
for what is going on in quantum mechanics.
 The really important thing here is that each con guration point
represents one totality of all electron positions in the chamber. If we
keep all the electrons xed except one, which we move, it explores
just three of the dimensions. In a much more modest way, there is an
analogy here with our existence on the Earth: we live in three
dimensions, but are normally restricted to the Earth’s two-
dimensional surface and do not normally move far in the third
dimension. For the electron, the unexplored dimensions are not one
but 3 × 1027.

We can now think about representing an ionization track. The

electron of a hydrogen atom has a characteristic probability
distribution of diameter 10−8 centimetres around its proton. In
quantum mechanics it is di cult to be certain about anything, but if
we nd a proton with no electron near it, this can indicate ionization
– the electron has been torn away by the alpha particle. Imagine that
we nd a state of the chamber in which 1000 protons have no
electrons near them; that these 1000 electron-less protons all lie
more or less on a line between the decayed radium nucleus and the
alpha particle; and that the statistics of the kinks along the line
match Born’s predictions for small-angle scattering. Naturally we
should say that this is an alpha-particle track. It has all the
appearances of recording quantum evolution with intermittent
collapse. This state of the chamber, interpreted as an ionization
track, is a perfect time capsule. Purely mathematically, it is a single
point in a space. But the one point stands for a distribution of a huge
number of electrons. As such, it is extraordinarily special – it is like a
snapshot of history itself. If it could think, it would say, ‘I am the
track of an alpha particle moving in space and time through a cloud
chamber.’

If the con guration space has innumerable dimensions, how

much vaster is the number of its points. The overwhelming – hugely
overwhelming – majority of the distributions they represent
correspond to nothing interesting or striking. Sprinkled very thinly
through this immense space are the distributions in which 1000
proton nuclei have no electrons near them. There are an incredible
number of such distributions, but they are still much more thinly
distributed than the stars in the sky. Within this already very thin
company with 1000 ionizations are those for which the ionizations
are all more or less on the line between the radium nucleus and its
escaped alpha particle. But still these are not yet alpha-particle
tracks. There is one more sieve – the scattering angles of the kinks
must match Born’s statistical distribution.

This piling of improbability upon improbability may seem

pedantic, but I do want to bring home the sheer improbability of
history. What immense creative power makes it? In addition, I am
preparing the next step in the story of geometrical optics. For this, as
I suggested earlier, it is helpful to start thinking of historical records
as exceptional, specially structured points in con guration space:
time capsules. Of course, if you look hard enough you can nd not
only them but all sorts of other things – pictures of Marilyn Monroe,
more or less anything you like – but all such ‘interesting pictures’ are
terribly thinly distributed. It is amazing that anything ‘ferrets them
out’. But causal quantum mechanics coupled with the incongruous
collapse mechanism and a benign low-entropy environment can do
the trick.

Before taking the next step, jettisoning collapse, we can add some
re nements. In the collapse picture, we can not only mark (with
‘paint’) the con guration point that is the time capsule of the
complete track. We can imagine a snapshot taken when only, say,
557 atoms have been ionized. The con guration point captured by it
will also be a time capsule, and we can mark it too. If we mark in
this manner all the stages – from no ionizations to all ionizations –
all the corresponding time capsules will be di erent points in the
con guration space. That is because they tell di erent stories, some
of which only reach, say, the track’s ‘adolescence’ or ‘middle age’.
Di erent con guration points necessarily represent di erent stories.
However, they are joined up more or less continuously in a path,
which represents an unfolding process.

If, like the god I imagined come to look at Platonia and its mists,
we could ‘see’ the con guration space and the wave function
sweeping over it, then in Bell’s ‘crude’ account we should see a patch
of wave function jigging its way along a track. The points along it
are the complete cloudchamber con gurations with successively
more ionizations. This con guration track is quite unlike the track
that represents a history in Newtonian dynamics. For a single alpha
particle, that is a track in three-dimensional space and the points
along it, de ned by three numbers, cannot possibly record history. In
contrast, each of the points traced out in the big con guration space
looks like a history of the three-dimensional track up to some point
along it. An analogy may help. Doting parents take daily snapshots of
their child and stick them day by day into a progress book. The
progress book after each successive day is like each successive point
along the track in the big con guration space: it is the complete
history of the child up to that date. Similarly, a point along the track
does not show the alpha particle at an instant of time, but its history
up to that time.

THE CREATION OF RECORDS: SECOND MECHANISM

If experiments as in Bell’s rst account are repeated many times, a

similar but di erent track will be photographed each time. Because
quantum mechanics deals in probabilities, some tracks may well be
more probable than others. Now imagine recording an alpha-particle
track by ‘marking’ the corresponding con guration point with ‘paint’.
All con guration points that have been ‘illuminated’ in any of the
experimental runs will be touched with paint, some many times.
Because the instant of radioactive decay cannot be predicted,
photographs taken at random will catch tracks of all ‘ages’ – birth,
adolescence, middle age, old age. Eventually, many di erent points
will have been touched by paint. A rich structure will have been
highlighted. Perhaps the best way to picture this is as innumerable
laments, all emanating from the small region in the con guration
space that represents the alpha particle trapped in the radium
nucleus while all the cloud-chamber atoms are in their ground states.

It would be quite wrong to suppose that these laments are so

numerous that they ll the con guration space. That comes from
confusion with ordinary three-dimensional space. It is always
dangerous to take analogies too literally, but if we are going to try to
use images, it is better to think of the structure that is formed in the
con guration space by the points that have been ‘touched with paint’
as being more like strands of a spider’s web spun out in the reaches
of interstellar space with huge gaps between them. Such a structure
is then a record of innumerable experiments interpreted in the rst
‘crude’ way.

One more comment. So far, we have considered only single

tracks. But in modern experiments a single particle colliding with a
detector particle can create many secondary particles. These also
make tracks simultaneously in the detector. A single quantum event
gives rise to many tracks. If a magnetic eld is applied the tracks are
curved by di erent amounts depending on the particle masses,
charges and energies. Beautiful patterns, representing quite
complicated histories, are created (Figure 51). This multitrack
process in ordinary space is still represented by one track in
con guration space. History, no matter how complicated, is always
represented by a single con guration path; records of that history,
which may be very detailed and more or less pictorial (actual
snapshots), can readily be represented by a single con guration
point. A library containing all the histories of the world ever written
is just one point in the appropriate con guration space.

We now come to the more sophisticated account of alpha-particle

interaction with a cloud chamber. The entire process is treated
quantum mechanically – as wave-function evolution in a space of
around 1027 dimensions. Initially, before the alpha particle escapes,
the wave function (of all the electrons and the alpha particle) is
restricted to a rather small con guration region. In the crude
collapse picture, alpha-particle escape and track formation is
represented as a ‘ nger’ of wave function that suddenly emerges
from it and rushes through the con guration space like a rocket
shooting through the sky.

Figure 51 Multiple tracks of elementary particles created by a single

quantum event. The swirls and curved tracks arise from the e ect of
 a magnetic eld on the charges of the particles created.

In the new picture, with everything treated quantum

mechanically and no collapse, an immense number of wave-function
‘ ngers’ emerge almost at once and race in a multitude of directions
across the con guration space. Each follows more or less one of the
tracks of the scenario with collapse. All the tracks are traced out
simultaneously. It is like one of those spectacular reworks that
explodes and shoots out a blazing shower in all directions. This is
what we should observe if we could see the wave function bursting
out from its original con nes into the great open spaces of Platonia.

It is not easy to explain why it behaves like this, but let me try.
The most important thing is that a con guration space is not some
blank open space like Newton’s absolute space, but a kind of
landscape with a rich topography. Think of the wave function
pouring forth like oodwater sweeping over a rocky terrain, whose
features de ect the water. It will help if you look again at Triangle
Land (Figures 3 and 4). It is bounded by sheets and ribs, and is the
con guration space for just three particles. The con guration space
for 1027 particles is immensely more complicated. Things like the
ribs and sheets that appear as boundaries of Triangle Land occur as
internal topography in Platonia, which is traversed by all kinds of
structures. The rules that govern the evolution of the wave function
force it to respond to this rich topography. The wave-function
laments are directed by salient features in the landscape.
 Now that we have some idea of how the ‘ rework explodes’, we
can think about its interpretation. The problem is that we never see
con guration space. That is a ‘God’s-eye’ view denied to our senses –
but fortunately not to our imaginations. We also never see a solitary
alpha particle making many tracks at once: all we ever see is one
track. How is this accounted for in the second scenario? By the same
device as before – by collapse. In the rst scenario, the alpha particle
was in many di erent places in its con guration space
simultaneously before we forced it to show itself in one region. This
was done by making it interact with an atom. This, most
mysteriously, triggered collapse, which was repeated again and
again.

In the second scenario, the complete system is, after a time

su cient for the ionization of 1000 atoms, potentially present at
many di erent places in its huge con guration space. The wave
function is spread out over a very large area, though concentrated
within it, in tiny regions. All the points within any of these regions is
like a snapshot of an ionization track, all di ering very slightly (and
hence represented by di erent points within a small region). There is
an exact parallel between the alpha particle in the rst scenario
being at many di erent places before the rst collapse-inducing
ionization and the state now envisaged for the complete system of
cloud chamber and alpha particle. It too is in many di erent ‘places’
at once.
 We can now collapse this much larger system by making a
‘measurement’ on it to see where it is. This is often done simply by
taking a photograph of the chamber. It catches the chamber in just
one of its many possible ‘places’. And what do we nd? A chamber
con guration showing just one ionization track, corresponding to
one of the points within one of the tiny regions on which the wave-
function mist is concentrated. We have collapsed the wave function,
but this time onto a complete track, not onto one position of one
particle.

If such experiments are repeated many times, the tracks obtained

are found to be essentially the same as the tracks in the rst
scenario. There are in principle small di erences, which come about
because the evolution is not quite the same in the two cases – in the
latter case the tracks can interfere to some extent, but in general the
nal results are more or less the same despite the very di erent
theoretical descriptions.

The reason for this is that seeds of the many di erent tracks –
di erent histories – are already contained in the initial wave
function. A concentrated wave function necessarily spreads, and if
this happens in a large enough con guration space under low-
entropy conditions it can excite many di erent con gurations that
embody records of many di erent histories. There is a snowball
e ect. We start with many small snowballs, the di erent possibilities
for the alpha particle at the beginning of the process. Each possibility
then becomes associated – entangled – with a di erent track. This is
rather like many di erent snowballs picking up snow. Subject always
to a pervasive quantum uncertainty, a fuzziness at the edges, these
are Everett’s many worlds. The distinctness of these di erent worlds,
the di erent histories, is determined by the extent to which part of
the system (the alpha particle in this case) is in the semiclassical
(geometrical-optics) regime.

It is the near perfection of the initial semiclassical state of the

alpha particle that creates such sharply de ned histories and ensures
that two such di erent scenarios give more or less the same results.
This is ultimately the reason why the notorious Heisenberg cut – the
position at which we suppose the quantum world to end and the
external, non-quantum world of classical measuring instruments to
begin – can be shifted in such a bewildering manner. As Bell
remarks, for practical purposes it does not matter much where we
place the cut to determine where collapse occurs, since the end
results are much the same. In either case, the appearance of history
is created by interaction between the semiclassical part and the
remaining, fully quantum system. The resulting correlation forces the
quantum system into a very special state.

It is really almost miraculous how the classical histories, latent as

very abstract entities within a semiclassical state of the alpha particle
when it is considered in isolation, force the wave function of the
remainder of the system (the cloud chamber) to seek out with
extraordinary precision tiny regions of its vast con guration space.
When these regions – or, rather, the points within them – are
examined, they turn out to represent con gurations that are
snapshots of tracks. They are records of histories.

So this is the next twist in the saga. First Hamilton found families
of classical, particle-like histories as ‘light rays’ in a regular
(semiclassical) wave eld. Then Schrödinger tried to mimic particle
tracks by superposing many slightly di erent semiclassical solutions
to create just one wave packet – the model of a single particle. It was
rather hard and contrived work for a meagre – but still very beautiful
– result. However, it immediately slipped through his ngers. But
then Heisenberg and Mott showed that quantum mechanics could
work far more e ectively as the creator of history than Schrödinger
had ever dreamed. Now one single semiclassical solution generates
(before the nal collapse) many histories. Instead of Schrödinger’s
contrived

Many semiclassical solutions → One history

we have natural organic growth:

One semiclassical solution → Many records of histories

CHAPTER21
The Many-Instants Interpretation

MANY HISTORIES IN ONE UNIVERSE

The story goes on. We have put only the cloud chamber into the
quantum mill – can we put the universe, ourselves included, in too?
That will require us to contemplate the ultimate con guration
space, the universe’s.

You can surely see where this is leading. Now the snowballs can
grow to include us and our conscious minds, each in di erent
incarnations. They must be di erent, because they see di erent
tracks; that makes them di erent. These similar incarnations seeing
di erent things necessarily belong to di erent points in the
universal con guration space. The pyrotechnics of wave-function
explosion out of a small region of Platonia – the decay of one
radioactive nucleus – has sprinkled ery droplets of wave function
at precise locations all over the landscape. (What an awful mixing of
metaphors – snowballs and sparks! But perhaps they may be
allowed to survive editing. The snowballs are in the con guration
space, the sparks in the wave function. This is a dualistic picture.)
 And now to the great Everettian di erence: collapse is no longer
necessary. Nothing collapses at all. What we took to be collapse is
more like waking up in the morning and nding that the sun is
shining. But it could have been cloudy, or cloudy and raining, or
clear and frosty, or blowing a howling gale, or even literally raining
cats and dogs. When we lay down to sleep in bed – when we set up
the alpha-particle experiment – we knew not what we should wake
to. What we take to be wave-function collapse is merely nding that
this ine able self-sentient something that we call ourselves is in one
point of the con guration space rather than another. When we
observe the outcome of an experiment, we are not watching things
unfold in three-dimensional space. Something quite di erent is
happening. We are nding ourselves to be at one place in the
universal con guration space rather than another. All observation,
which is simultaneously the experiencing of one instant of time, is
ultimately a (partial) locating of ourselves in Platonia. Each of our
instants is a self-sentient part of a Platonic form.

The coherence of this picture hangs on the ability of the

universal wave function to seek out time capsules in Platonia that
tell a story of organic growth. All stories are in Platonia, some
bizarre beyond the dreams of Hieronymus Bosch or modern
surrealists. The history that we experience may have its horrors, but
it is extraordinarily coherent and self-consistent. The rst task of
science is to save the appearances. So, rst and foremost, we need to
nd a rational explanation for the habitual miraculous experiencing
of time capsules, these freighters of history. This is where the
probability density of the wave function, its shimmering blue mist,
plays such a crucial role. Because apparent records of all histories –
and a mind-numbing multitude of non-histories – are present in
Platonia, we shall not have an explanation of the appearances
worthy of the name unless the blue mist shines brightly over time
capsules of the kind we know so well from direct experience. And it
should not shine brightly anywhere else. We shall then have a
theory that does truly save the appearances. Bell’s analysis hints that
universal quantum cosmology might be that theory.

It is time to take stock once more. First, we muster the

interpretations of quantum mechanics. How do they look in the light
of Bell’s analysis? What appearances do they save and how well do
they do it? There are two minimum requirements of an
interpretation – it must explain why we see just one world
(Einstein’s Moon problem) and it must explain why we think it has a
history. The latter is the harder task. However, it may be important
not to ask for too much. To save the appearances, we do not have to
create a unique history: we need only explain why there seems to be
a unique history. That was Everett’s insight. If we can stand back
from our parochial prejudices, a theory which can achieve that is
already little short of miraculous.

Except for many-worlds variants, all the interpretations strive for

the severe criterion of only one history. They were created for that
and all achieve it by brute force. History is created by repeated
strangling of the wave function (Copenhagen and physical collapse)
or by adding incongruous extras: the so-called hidden variables. The
German doppelt gemoppelt means messing things up by doing them
twice over. In their anxiety to recover a unique history, the
proponents of these interpretations crudely impose one history on a
theory that can already create many histories which are autonomous
– and hence each unique in our experience – by a beautiful natural
mechanism. Hamilton’s discovery makes it inescapable that histories
are latent in the quantum formalism. It is just a matter of coaxing
them out into the open.

BELL’S ‘MANY-WORLDS’ INTERPRETATION

From his discussion of alpha-particle tracks, Bell turned to a

remarkable cosmological interpretation of quantum mechanics. It
makes essential use of the notion of time capsules and is therefore
very similar to the interpretation I shall present in the nal chapter.
Bell saw it as a way of retaining Everett’s idea that the wave
function never collapses without proliferating worlds.

Bell claimed that the really novel element in Everett’s theory had
not been identi ed. This was ‘a repudiation of the concept of the
“past”, which could be considered in the same liberating tradition as
Einstein’s repudiation of absolute simultaneity’. Obviously,
something exciting is in prospect, and Bell does not disappoint. He
looked for the quantum property that enabled Everett to make his
many-worlds idea plausible, and pointed out that the accumulation
of mutually consistent records is a vital part of it. This recognition
had led Bell to his analysis of the formation of alpha-particle tracks,
which have the obvious interpretation that they are records of
alpha-particle motion. He showed that ‘record formation’ is a
characteristic quantum property. At least under cloud-chamber
conditions, the wave function concentrates itself at con guration
points that can be called records. Although Bell did not use my
term, such points are manifestly time capsules. He noted that
Everett’s interpretation could not even be formulated were it not for
the wave function’s propensity to nd them.

He then attacked head-on the conventional notion of history

inherited from classical physics as a continuous path through
con guration space. This might make sense if, god-like, we could
see all time and the con guration space with history highlighted as
a path in it by a ‘thread’ or ‘paint’. But our only access to the past is
through records. As Bell says, ‘We have no access to the past. We
have only our “memories” and “records”. But these memories and
records are in fact present phenomena.’ Our only evidence for the
past is through present records. If we have them, the actual
existence of the past is immaterial. It will make no di erence to
what we know. Hence ‘there is no need whatever to link successive
con gurations of the world into a continuous trajectory’.

His ‘Everettian’ interpretation is this: time exists, and the

universal wave function ψ evolves in it without ever collapsing.
Because ψ has the propensity to seek out time capsules, it will
generally be concentrated on them. Real events are actualized as
follows. At each instant of time, ψ associates a de nite probability
(the intensity of the blue mist in my analogy) with each
con guration. At any instant, just one event is actualized at random
in accordance with its relative probability. The higher the
probability, the greater the chance of actualization. Since time
capsules have the highest probabilities, they will generally be
selected.

Sentient beings within them will possess memories and records

that convince them they are the product of history. But this will be
an illusion. In reality, the points realized at successive instants of
time are chosen randomly and jump around in a wildly
unpredictable manner in the con guration space. The sentient
beings within the actualized points have memories of quite di erent
histories. It is all very bizarre, though within each randomly
selected time capsule the memories and records tell a most
consistent story. Bell rejected his ‘many-worlds’ interpretation as too
absurd:

Everett’s replacement of the past by memories is a radical

solipsism – extending to the temporal dimension the
replacement of everything outside my head by my
impressions, of ordinary solipsism or positivism. Solipsism
cannot be refuted. But if such a theory were taken seriously
it would hardly be possible to take anything else seriously.
So much for the social implications. It is always interesting
 to nd that solipsists and positivists, when they have
children, have life insurance.

This is all very entertaining – and I too have children and life
insurance – but these are just the kind of ad hominem quips that
were tossed at Copernicus and Galileo. I do believe that Bell came
close to a viable cosmological interpretation of quantum mechanics,
and should have kept faith with his title (‘Quantum mechanics for
cosmologists’). But he left the cosmologists with nothing. Later he
gave warm support to one of the theories in which wave-function
collapse is a real physical process. In it, the propensity of the
quantum-mechanical wave function to nd time capsules plays no
role. History is created by a succession of actually realized states. It
is there with or without any record of it.

From the way Bell wrote in 1980, either he was unaware of the
Wheeler-DeWitt equation and the possibility that the universal wave
function is static, or he dismissed this without mention. It would be
interesting to know how he would have reacted to the idea – he
seems to have had a somewhat Newtonian notion of time. Sadly, he
died several years ago, so we cannot ask him. I regret this especially
since his 1980 proposal is very close to mine in two of its three
main elements. He may have believed in time, but his emphasis on
memories and records and their rather natural occurrence in the
quantum context are valuable support for me. So are his views on
ontology and psychophysical parallelism. This is the third common
element.
 In discussing Everett’s theory, I mentioned the so-called
preferred-basis problem. This arises from transformation theory: a
quantum state simultaneously encodes information about mutually
exclusive properties. Viewed one way, it gives probabilities for
particle positions; viewed another, it gives probabilities for their
momenta. It is impossible to extract this information simultaneously
and directly by, so to speak, ‘looking at the system’. We must let the
system interact with instruments. Depending on how the
instruments are arranged, we can extract information about either
the positions or the momenta, but not both at once. The ambiguity
becomes especially acute if the instruments are treated quantum
mechanically. We cannot say what state they are in or what they are
measuring.

Bell advocated a simple and robust answer to this in many of his

writings, including his 1980 paper: the complete system formed by
the particles and the instruments measuring them is always de ned
in the last resort by positions. In any quantum state, di erent sets of
positions are present simultaneously, but it is always positions that
are present. The di erent kinds of quantum measurement, giving
alternatively position-type and momentum-type outcomes, arise
because the same sets of positions of the measured system are made
to interact with characteristically di erent sets of instrument
positions. Everything is ultimately inferred from positions. This is
exactly my position. Platonia is the universal arena. To Bell’s
arguments – and gut conviction – for this standpoint I would add
the impossibility of obtaining a satisfactory theory of inertia and
time unless positions are fundamental.

Now, what did Bell regard as the physical counterpart of

psychological experience? Is it in the wave function, as Everett and
many others have assumed, or in matter con gurations? Bell, like
myself, opts for the latter: ‘It is ... from the xs [the con gurations],
rather than from ψ, that in this theory we suppose “observables” to
be constructed. It is in terms of the xs that we would de ne a
“psycho-physical parallelism” – if we were pressed to go so far.’
Although Bell does not spell out his parallelism too explicitly – he
does not seem to want to be ‘pressed’ too far – it is clear from the
way he makes memories and records responsible for our idea of the
past, rejecting any ‘thread’ connecting con gurations at di erent
times, that subjective awareness of both positions and motions of
objects must be derived from the structure in one instantaneous
con guration. The self-sentient con gurations must be time
capsules. Not only the king sher but also the appearance of its ight
must be in one con guration, for nothing else would be logically
consistent. The main lessons I draw from Bell’s paper are
incorporated in the many-instants interpretation that I favour.

THE MANY-INSTANTS INTERPRETATION

This is based on a conjecture that I shall try to justify in the next

chapter. Here I simply assume it. It is that the universe is described
by an equation of Wheeler-DeWitt type, which may have one or
many solutions, and that each of its well-behaved solutions
concentrates its probability density on time capsules. Bell showed
that this does happen if time exists, and if evolution is real and
commences from a low-entropy state. Since I deny time, I cannot
appeal to a special initial state. There is only one state and no
evolution. That is the problem for the next chapter; here I want to
describe the kind of state I conjecture and how it must change our
view of history.

Most important is a distinction between two di erent kinds of

variable. Bell showed how the alpha-particle semiclassical state
contains latent histories which then become entangled with the
cloud-chamber electrons. The electrons could be in a huge number
of di erent con gurations, but in the Mott-Heisenberg solution the
only con gurations with high probability are those that look like
alpha-particle tracks. Something similar must happen in cosmology,
but there is a di erence.

Imagine a swarm of 5000 bees. Its con guration space has

15,000 dimensions. However, from a distance we cannot see the
individual bees, only the overall position of the swarm and, say, its
size (radius). These are four dimensions of the con guration space.
In such situations a few of the con guration-space dimensions
describe the system’s large-scale properties, and the remaining,
much more numerous dimensions describe the ne details. The
corresponding large-scale and small-scale con guration spaces are
illustrated in Figure 52.
 Any point in Figure 52 represents a possible position of all the
bees. Horizontal motion from a point changes the swarm’s position
and size without changing the relative position of the bees within it.
Vertical displacement leaves the swarm’s position and size
unchanged but rearranges the bees. Since this can happen in so
many ways, each vertical point actually represents multitudinous
possibilities. Alas, we have only the vertical to represent them. Also,
to make even a moderately realistic model of the universe, the
horizontal positions should represent the positions of not just one
swarm but many. Imagine, say, 100 swarms. Each horizontal
position then represents one relative arrangement of their positions
and sizes as complete units. Di erent vertical positions having the
same horizontal position then correspond to all rearrangements of
the bees that leave the swarms as they are. This is very schematic,
but it is su cient to explain the scheme.

If the wave function of the universe is static, quantum cosmology

reduces to the question of how its values are distributed in Platonia.
For the moment, I shall simply give you my guess; arguments for it
come later. My guess is a special distribution closely similar to the
cloud-chamber one described by Bell. In most of Platonia the wave
function has extremely small values – the blue mist has negligible
intensity. However, in a few special regions, distributed over a large
area, the blue mist’s intensity is, relatively, hugely higher. These
regions correspond to some arrangement of the swarms, determined
by the horizontal position, and to the detailed positions within
them, determined by the vertical positions. It is in the probabilities
for these detailed positions that the blue mist is extraordinarily
selective. The probabilities for the horizontal positions are relatively
uniform over quite large regions. By themselves, they represent a
dull state of a airs. The situation is transformed by the
con gurations that specify the ne details within the swarms. At the
very rare con gurations where the blue mist shines brightly, the
ne details look like records of a history of the swarms as complete
units. They suggest that the swarms have moved in a classical
history from some past up to a present instant, in the position they
now occupy.

This is illustrated in Figure 52, in which the points X and Y in

Platonia have large-scale positions B and D. The ne details at X and
Y seem to represent records of how the swarms have moved from
earlier con gurations A and C along curves AB and CD in the large-
scale (horizontal) con guration space. They seem to be records of
these large-scale histories. The blue mist has a high intensity not
only at X and Y but also, for example, at P and Q, at which points
the ne details suggest they represent records up to the intermediate
stages E and F in the histories AB and CD.

By no means all details need represent history. Footprints in the

sand on a wide beach record the movements of people who have
walked on it, but over much of the beach there need be no apparent
records. Think again of the number of atoms in a pea. A tiny
fraction of them can easily record the pea’s history up to its current
present. The huge numbers we confront in physics explain why we
may have wrong ideas of what history actually is. We may have
jumped to a conclusion too quickly.

In the Newtonian picture, in which history is a curve in

con guration space, it is extremely hard to understand how records
arise. Even if a single curve is realized, any point on it could have
any number of histories passing through it. How can one
instantaneous con guration of particles suggest the motions that
they have? However, if we keep an open mind about the laws that
determine things, a fraction of a pea’s atoms may well seem to
record a history of its large-scale features. This does not mean that
all its atoms had a unique history. Without change in the pea’s
large-scale structure, the same large-scale history could be coded in
innumerable di erent ways by only a tiny fraction of its atoms. In
the imagery of Figure 52, there will be a whole cloud of points X in
the con guration space that correspond to the same large-scale
con guration and to the same history up to it. The di erent points
in the cloud simply code the same history in di erent ways. What is
more, for each point along the large-scale history AB there will be a
corresponding cloud of points that record the same history up to
that point in di erent ways. There will be a ‘tube’ of such points in
the con guration space. No continuous ‘thread’ joins up these points
in the tube into Newtonian histories. The points are more like sand
grains that ll a glass tube. Each grain tells its story independently
of its neighbours. In any section of the tube, the grains all tell
essentially the same story but in di erent ways, though some may
tell it with small variations.

Figure 52 The division of Platonia. The horizontal dimensions

represent the large-scale con guration space, and the single vertical
dimension represents the small-scale space. The ‘horizontal curves’
(AEB and CFD) represent histories of the large-scale features. Each
point like A represents, say, the overall position and size of a swarm
of bees. In contrast, each of the points Q, P, X, V on the vertical lines
represents the huge number of small-scale details.

I think this is the way to think about history in quantum stasis.

Could we but see the picture – all Platonia with its misty crannies –
we should see it as it is: the lawful de nite world for which Einstein,
like so many physicists, longed. But it is a timeless book full of
di erent stories that tell of time. Quantum mechanics can create the
appearance of multiple histories. However, will it in quantum
cosmology? Its conditions are not quite Mott and Heisenberg’s
conditions.
CHAPTER22
The Emergence of Time and its Arrow

CAUSALITY IN QUANTUM COSMOLOGY

John Bell’s account of time-capsule selection contains a very large

con guration space, time, the wave function and its equation (the
time-dependent Schrödinger equation) and a special initial state.
This last is most important. If quantum cosmology is static,
something else must replace it. We cannot impose an initial
condition in the past because there is no past. But we can try
something similar. Suppose that the universal con guration space
had only three dimensions and not the monstrous number I have so
often asked you to consider. We could then specify the wave
function on a two-dimensional plane in that three-dimensional
space, and use the equation satis ed by the wave function to nd it
at other points. This is like evolving a state in time except that the
evolution is in the third, spatial direction.

If we attempt this in ordinary quantum mechanics with the

stationary Schrödinger equation, which in some respects at least is
like the Wheeler-DeWitt equation, the wave function starts to
misbehave sooner or later. Either it becomes in nite, or it cannot be
evolved continuously, or some other disaster happens: it ceases to
be ‘well behaved’. The remarkable and exciting discovery that
Schrödinger made was that the hydrogen atom does have a very
special set of solutions that are well behaved everywhere and for
which therefore no disaster happens. These very special states
correspond exactly to the negative-energy states of the hydrogen
atom. He had explained what had hitherto been one of the deepest
mysteries of physics – the spectral lines of atoms and molecules.

My main interest here is the transformation of our notions about

causality that a solution of this kind could represent. The traditional
view is that what happens now was ‘caused’ by some state in the
past. There is always arbitrariness in this picture because the past
state is arbitrary. But suppose the world is described instead by a
solution of some Wheeler-De Witt equation that is everywhere well
behaved in Schrödinger’s sense. I have already pointed out that such
solutions are ultra-sensitive to the domain on which they are
de ned – otherwise they could not remain well behaved
everywhere. Such solutions present a kind of pre-established
harmony.

The Wheeler-DeWitt equation then constitutes the rules of a

game played in eternity. The wave function is the ball, Platonia is
the pitch. If a well-behaved solution exists, then only two things can
have conspired to create it: the rules of the game and the shape (the
topography) of the pitch. In contrast, Bell’s time capsules are
created by the rules, time, the topography and a special initial
condition. What a prize if we could create time capsules by the rules
and the shape of the pitch alone! Arbitrary, vertical causality
(through time) would then be replaced by timeless horizontal and
rational causation – across Platonia.

SOCCER IN THE MATTERHORN

It is possible. There are plenty of time capsules in Platonia. It is not

just time and the special initial conditions that enable the wave
function to nd time capsules. The rules of the game and, above all,
the pitch size and topography are most conducive to it. Indeed, the
con guration space is a prerequisite. As Nevill Mott remarked, ‘The
di culty that we have in picturing how it is that a spherical wave
can produce a straight track arises from our tendency to picture the
wave as existing in ordinary three-dimensional space, whereas we
are really dealing with wave functions in the multispace formed by
the co-ordinates both of the alpha-particle and of every atom in the
Wilson chamber.’ What interests me now is not so much the
dimensions as the pitch’s shape. What follows is speculation. Mine. I
am not aware that anyone else has made it (though Dieter Zeh
considered something rather similar). I have lectured several times
on the idea, and in 1994 published quite a long paper on it in the
journal Classical and Quantum Gravity. A problem with the idea is
that as yet it is purely qualitative. Physicists rightly want to see real
calculations (which, alas, are bound to be di cult), not mere
speculation, before they endorse an idea. But the more I think about
the idea, the more plausible, indeed almost inescapable, it appears
to be. It is about the origin of the arrow of time – and time itself.

The arrow of time, manifested in the ubiquity of time capsules, is

a colossal asymmetry. It is well-nigh inexplicable in time-symmetric
physics, the present rules. Since Boltzmann’s age, it has towered
there, an unsealed Everest. It can be described but not yet
explained.

This book has been one long, sustained e ort to shed redundant
concepts. We now are down to two: a static but well-behaved wave
function and the con guration space. The latter is Platonia, our
pitch. I look at it as a child might – what a lopsided thing it is!
However I turn in my mind the notion of ‘thing’, the space of all
things constructed according to one rule comes out asymmetric. All
the mathematical structures built by physicists to model the world
have this inherent asymmetry. One rule creates triangles, but they
are all di erent. No matter how you arrange them, their
con guration space falls out oddly. Have another look at Triangle
Land (Figures 3 and 4), which is just about the simplest Platonia
there can be. And what does it look like? An upturned Matterhorn.
Imagine trying to play football on that pitch.

The barest arena in which we can hope to represent appearances

is the set of possible things. If we banish things, we banish all the
world. So we have kept things and made them the instants of time.
As we experience them, they are invariably time capsules. This is
the principal contingent fact of existence: the wave function of the
universe, playing the great game in timelessness, seeks and nds
time capsules. What all-pervasive in uence can put such a rooted
bias into the game? The explanation seems to scream at us. Platonia
is a skewed continent.

My conjecture is this. The Wheeler-DeWitt equation of our

universe concentrates any of its well-behaved solutions on time
capsules. I suspect the same result would hold for many di erent
equations and con guration spaces. The inherent asymmetry of the
con guration space will always ‘funnel’ the wave function onto time
capsules.

I could ll up pages with hand-waving arguments for why this

should be so, but they would ba e the non-specialist and o end the
specialist. I shall attempt only to show that the ‘seeking out’ of time
capsules need not depend on time and a special initial condition.
Stationary equations may also do the trick. I may also mention that
if, as explained in the Notes, the universe can, in accordance with
my recent insights, be understood solely in terms of pure structure,
so that absolute distance plays no role, it will certainly be possible
to make the arguments of the remaining parts of this chapter more
precise and convincing.

TIMELESS DESCRIPTIONS OF DYNAMICS

In Part 2, we saw that Newtonian classical histories of the universe

can be described in a timeless fashion as ‘shortest’ curves
(geodesics) in con guration space. All histories that have the same
energy can be described this way. This fact is often helpful,
simplifying the solution of problems. In quantum mechanics
something similar happens.

Physicists often want to know what will happen if some atomic

particle is shot at a target of other particles. One way is to represent
it by a ‘cloud’ of wave function – a wave packet – that moves
towards the target. As Schrödinger showed, such a packet can be
formed from waves corresponding to a relatively small range of
momenta and energies, and it moves with a more or less de nite
velocity. When it reaches the target, it is ‘scattered’ – the wave
function ies o in many directions. Physicists use the time-
dependent Schrödinger equation to nd the probabilities for these
various directions. In this picture, the wave packet moves and is in
di erent positions at di erent times. This is rather like the
representation of history in Newtonian physics as an illuminated
spot moving along a curve in con guration space.

There is, however, an alternative method. A single static wave

covers the whole region – before, at, and after the target – traversed
by the wave packet in the rst picture. This one wave satis es the
stationary Schrödinger equation and corresponds to a particle with
the average momentum and energy of the packet. As far as the
target, the static wave is regular and plane, but at the target its
pattern gets broken up. The interesting thing is that if we examine
the pattern of the disrupted (but still static) wave in the region
behind the target, we can deduce from it the probabilities with
which the particle will be scattered in di erent directions in the rst
picture. I shall not go into details; su ce it to say that the one static
wave is a kind of record of all the successive wave-packet positions
in the rst description. This is closely analogous to the way in
classical physics in which the curve in the con guration space is a
summary of all positions of the illuminated spot taken to represent
the system at di erent times.

Interestingly, Max Born made his pioneering scattering

calculations in the newly created wave mechanics by the second
method. At that time Schrödinger had not even published his time-
dependent equation. All the great early discoveries in wave
mechanics, including Born’s statistical interpretation of the wave
function (which he came to by mulling over his scattering
calculations), were made before the supposedly more fundamental
and ‘correct’ time-dependent equation had been found. I nd this
suggestive. It strengthens my belief that all the physics of the
universe can be described by a timeless wave equation. In fact, Mott
also used the stationary equation to obtain the alpha-particle tracks.
That timeless equation can locate time capsules.

But ‘can’ is not ‘must’. The fact is that Mott used a special
technique, always followed in such calculations, that mimics the
wave-packet behaviour. The answer is to some extent simply
assumed rather than truly derived and demonstrated. This can be
done because the time-dependent and stationary Schrödinger
equations have di erent structures, the latter having an extra
freedom not present in the former. At each stage of his calculations,
Mott systematically exploited this extra degree by making a de nite
kind of choice. This choice was not imposed by the mathematics but
was made, probably instinctively, to match his temporal intuition.
In fact, Mott’s solution is not a proper solution at all but a kind of
bookkeeping record of how the real process would unfold in time. In
addition, the condition corresponding to low entropy was also
assumed rather than derived.

My conjecture seems to rest on a shaky basis. But there is more

than one way of looking at this. The arguments for a timeless
quantum universe are strong. The timelessness of the Wheeler-
DeWitt equation, found by well-tried quantization methods, re ects
the deepest structure of Einstein’s theory. Quite independently, we
never observe anything other than time capsules – the entire
observable universe is marked, at all epochs, by profound temporal
asymmetry. If we trust the equation, our observations tell us the
outcome of a mathematical calculation performed by the universe
itself using that equation. For that is what the contingent universe
must be: a solution of the equation. If what we observe – a profusion
of time capsules – is a representative fact, then the equation does
concentrate ψ on time capsules.

We can take Mott’s solution more seriously. Several points can

be made. Situations in which part of a quantum system is in the
semiclassical regime, so that Hamilton’s ‘light rays’ are present as
latent or even incipient classical histories, are rather common and
characteristic. The Heisenberg-Mott work then shows that such
latent histories will become entangled with the remaining quantum
variables, which must, in some way, re ect and carry information
about those histories. What is not clear is whether the histories will
exhibit a pronounced sense of direction – an arrow of time. That,
above all, is put into the Mott solution by hand.

Also relevant is the mathematically somewhat suspect procedure

known as successive approximation used to construct the Mott
solution. There is no global arena in which the cloud chamber
resides. Its atoms are e ectively located in empty Euclidean space,
and Mott could keep on adding approximations without worrying
about their behaviour far from the cloud chamber. He was not
constructing a genuine well-behaved solution, in which one must
ensure the behaviour is right everywhere, especially at in nity.
Instead, Mott used in nity as a kind of dustbin. This could not be
done in a realistic situation, as I would now like to show.

A QUANTUM ORIGIN OF THE UNIVERSE?

When Planck made the rst quantum discovery, he noted an

interesting fact. The speed of light, Newton’s gravitational constant,
and Planck’s constant clearly re ect fundamental properties of the
world. From them it is possible to derive the characteristic mass
length lplanck and time fplanck with approximate values
On atomic scales the Planck mass is huge, corresponding to about
1019 hydrogen atoms. In contrast, the Planck length and time are far
smaller than anything physicists can currently measure.

Much of current cosmology is concerned with the ‘interface’ of

quantum gravity and classical physics. The universe around us is
described by general relativity. This classical treatment is said to be
valid right back into the distant past, very close to the Big Bang. The
quantum phase of cosmology is supposed to become important only
at extraordinarily small scales, of the order of the Planck length, 10–
33 centimetres. Lights travel this distance in 10–43 seconds, and it is
argued that quantum gravity ‘comes into its own’ only in this almost
incomprehensibly early epoch.

All researchers agree that the nature of reality changes

qualitatively in this domain. Di erent laws must be used. Time
ceases to be an appropriate concept: things do not become, they are.
In a process often likened to radioactive decay, our classical
universe that emerges at the Big Bang is represented as somehow
‘springing’ out of timelessness, or even nothing. A mysterious
quantum birth creates the initial conditions that apply at the start of
the classical evolution. Our present universe is then the outcome of
the conditions created by quantum gravity. The dichotomy between
the laws of nature and initial conditions is thus resolved if the
quantum creation process can be uniquely determined.

Stephen Hawking has long been working on this problem, and

believes it can be solved by his so-called no-boundary proposal, a
mechanism which should lead to a unique prediction for the initial
conditions. His ‘imaginary-time’ mechanism, described in his A Brief
History of Time, seemed to have the potential to do this. However, it
has been widely criticized, and there are technical problems. The
most serious seems to be that even if the mechanism can be made to
work it will not produce unique initial conditions. Where Hawking
has led, many have followed, and numerous creation schemes have
been proposed.

My di culty with this approach is the division introduced

between the quantum and classical domains. One could almost get
the impression that the laws of nature actually change, and I am
sure that no theoretical physicist believes that. The approach is
adopted because the physical conditions are hugely di erent in the
two domains. In physics it is very common to use quite di erent
schemes if the conditions studied are di erent. No engineer would
use quantum mechanics to describe water ow in pipes, for
example. But the much more appropriate hydrodynamic equations
are consequences of the deeper quantum equations, and are valid in
the appropriate domain.

Cosmology may be di erent. Most physicists have a deeply

rooted notion of causality: explanations for the present must be
sought in the past (vertical causality, as I have called it). This
instinctive approach will be awed if the very concept of the past is
suspect. If quantum cosmology really is timeless, our notion of
causality may have to be changed radically. We cannot look to a
past to explain what we nd around us. The here and now arises not
from a past, but from the totality of things (horizontal causality).

Figure 53 A schematic representation of Platonia. All points in each

horizontal section represent con gurations of the universe with the
same volume but di erent curvatures and matter distributions in
them. According to the ideas of quantum creation, as yet unknown
laws of quantum gravity hold near Alpha, and in some rather
mysterious way give rise to conditions under which our universe –
 and with it, time – ‘spring out of Alpha’. The thread shown
ascending from Alpha represents the history of our universe that
results from the enigmatic quantum creation.

Figure 53, a schematic representation of Platonia, may help. This

is the skewed continent, as the cone shape makes clear. The
quantum-creation approaches imply that the enigmatic and as yet
unknown laws of quantum gravity create at the vertex – Alpha – a
‘spark in eternity’. The spark, in its turn, creates close to Alpha the
initial conditions of our actual universe. Time is born at the ‘spark’.
Our classical universe is the thread ascending through Platonia to
our present location.

Figure 54 shows what is often called the chronology of the

universe (the vertical axis is time, and the ‘quantum creation’ at
Alpha occurs at the bottom left). Here each horizontal section is one
point on the thread through Platonia in Figure 53: it is space at the
corresponding time. The characteristic structures in space at the
various cosmic epochs are shown: quarks in a soup near Alpha,
primordial hydrogen and helium after the rst three minutes,
incipient galaxies a few thousand years later, and so on, right
through to life on Earth at the present. Such is the thread ‘born in
the quantum spark’. The Everettian quantum cosmologists believe
that the one quantum spark creates many such threads, one of them
ours. But is there a ‘spark’ at Alpha? The laws of quantum gravity
hold not just near Alpha, but throughout Platonia – they are the
conjectured universal and ultimate laws. We have to ask what kind
of solutions they can have, and how the solutions are created. How
do threads, one or many, emerge?
 Figure 54 Chronology of the universe. Redrawn from A Short
History of the Universe by Joseph Silk (W. H. Freeman/Scienti c
American Library, 1994). The vertical axis represents time. Alpha,
the ‘quantum creation’, occurs at the bottom left.

Our entire experience tells us that the well-behaved solutions of

the stationary Schrödinger equation that describe the characteristic
structures of atoms, molecules and solids are determined by the
complete structure of the con guration spaces on which they are
de ned. They exhibit global sensitivity – their behaviour has to be
right everywhere. Since the governing equation does not contain the
time, this delicate ‘testing out of all possible behaviours’ takes place
in timelessness. If the Wheeler-DeWitt equation is like the stationary
Schrödinger equation, then Alpha, where time is allegedly born,
plays an important role, but it is not the locus at which some all-
decisive die is cast. Of course it is singular – Platonia abuts on
nothing at Alpha – but there are innumerable other special points
scattered all over Platonia. None are quite like Alpha but, together
with the overall shape of Platonia, they all have their role to play.
Quantum mechanics is nothing if not democratic. Solutions of the
Wheeler-DeWitt equation must be produced by a kind of dialogue
between every point in Platonia.

The picture suggested by such arguments is this. Platonia as a

whole determines how the static wave function ‘beds down’ on its
landscape. There will be regions of semiclassical behaviour with
respect to a large set of macroscopic quantities in which latent
classical histories are de ned. Just as the latent alpha-particle tracks
get correlated through the wave function with the chamber
electrons, ‘nudging’ the wave function onto time capsules, the same
thing can happen in Platonia.

VISION OF A TIMELESS UNIVERSE

Let me now give you my vision of quantum cosmology, contrasting

it with the chronology of the universe (Figure 54), that temporal
representation of the ‘thread’ in Figure 53. Every instant of time you
can conceive of is somewhere in Platonia. But the instants of time
can themselves be richly structured beyond imagination. All things
we see around us now in the universe are just parts of instants of
time. All over Platonia there exist instants of time in which Wagner
is composing Tristan and Isolde, astronauts are repairing the Hubble
Space Telescope, birds are building nests and I am baking bread.
The wave function of the universe nds its way to very few of them.
The structure of the wave function and the form of the laws of
nature – in which the tendency of gravity to clump matter is surely
vital – forces the blue mist to seek out the most special instants,
strung out along delicate threads. I think it is wrong of cosmologists
to call Figure 54 a chronology of the universe. It is the map of a
footpath in Platonia. The blue mist shines at instants containing
time capsules, all of which, in their di erent ways, tell stories of a
journey from Alpha along a ne thread of ‘history’ – a path winding
through Platonia. Time is in such instants since they re ect the story
of the path and, since the structure of Platonia in its totality forces
the universal wave function to ‘light up’ the paths, there is a sense
in which these instants re ect everything that is.

However, whereas alpha particles create, through their tracks, a

literal image of history, the time capsules of the real universe
embody their stories in a much subtler manner. This is inevitable
given the grandeur of the story – cosmology in its entirety. Consider,
for example, the Sun. Quantum mechanically, it will need to be
represented in a con guration space of, say, 1060 dimensions, but
vast stretches of it will be virtually devoid of wave function. The
mere fact that the Sun is roughly spherical and can be well modelled
by the laws of stellar structure sweeps most of the con guration
space clean of wave function. The particular abundances of the
chemical elements within the Sun have the same e ect, drastically
limiting the region of the solar con guration space in which the
blue mist is concentrated.

One con guration at which the blue mist does shine brightly will
be a characteristic distribution of all the particles in the Sun. To an
experienced astrophysicist, this distribution tells an immensely rich
story stretching back to the rst three minutes (in the standard
picture) when the primordial hydrogen and helium abundances
were established. The whole story of the cosmos that we call our
own is written in the distribution of the Sun’s particles: the
formation of galaxies and the earliest generations of stars; the
supernova explosion that triggered the formation of the Sun and the
solar system, and left the radioactivity that still powers so much
tectonic and volcanic activity on the Earth; and the Sun’s steady
burning of its nuclear fuel.

The decisive element in this picture is the seed – or rather, seeds

– from which these stories can all grow by the penetration of the
wave function into the nooks and crannies in Platonia where the
con gurations are coherent stories. The wave function can be
present there only if it is entangled with the latent histories of a
semiclassical wave function established at least somewhere in
Platonia. These are the Hamiltonian ‘light rays’ from which
everything must ‘grow’. Where are they likely to run, and what will
their properties be? This is where the shape of Platonia must
become decisive. The points in Platonia near Alpha containing
Wagner and king shers are simply not visited by the blue mist,
since Platonia as a whole lays out the latent histories in patterns
that do not get entangled with such points.

Modern classical cosmology gives some hints as to where the

latent histories might run. The simplest Big Bang cosmological
solutions of Einstein’s equations, rst discovered by the Russian
mathematician Alexander Friedmann in the early 1920s, have the
maximum degree of symmetry and therefore ascend the central line
in Figure 53. This is the history distinguished in cosmology by the
universe’s contents, despite the relativity of simultaneity. The
universe explodes out of the singular state of zero volume, expands
to a maximum volume and then recon-tracts to zero volume, gravity
having halted and then reversed the initial, very rapid expansion. As
shown schematically in Figure 55 (see p. 321), the universe ends in
a Big Crunch. In other cosmological models, which normally require
the universe to be spatially in nite, the expansion is so violent that
the expansion is never halted.

In reality the universe is not completely symmetric, and the path

out from Alpha is not exactly retraced, as shown in Figure 55, in
which the rays emanating from Alpha are a measure of the relative
‘irregularity’ of the spatial con gurations of the universe. Up the
vertical ray, the universe is perfectly smooth, but on the rays that
fan out at progressively larger angles the relative irregularity
increases. The diagram shows one classical history which, at one
end, starts in an almost perfectly smooth state but then becomes
more and more irregular, due to the formation of galaxies, stars,
black holes, planets and even human beings. This history reaches
maximum expansion, turns round and recontracts, becoming more
irregular all the time. It returns to the state of very small volume at
a di erent point of Platonia, since although the volumes are small in
both cases there are many additional variables that describe the
structure of the state. Thus Alpha, the ‘end of Platonia’, is not a true
point but actually a huge space of di erent possibilities, all with
vanishingly small volumes.

Regarded purely as a path through Platonia, we cannot say that

one end of this history is its beginning and the other is its end. I
have lapsed into conventional talk. Such a priori notions do not
belong in a timeless theory. Nevertheless, the two ends of the path
are very di erent in nature, and it is tempting to say that, if our
own existence is associated with such a path, its smooth end is what
we would call the past and the irregular end the future. This would
be very much in the tradition, initiated by Boltzmann, of suggesting
that our sense of the forward ow of time, its arrow, is grounded
solely in the increase in disorder that virtually all classical
trajectories must exhibit if they pass through an exceptionally
ordered region. Normally there are trajectories that both enter and
leave such regions, and the entropic explanation of the arrow of
time suggests that time will seem to ow forward in both directions
out of these regions. In the present example, the exceptional region
is on the frontier of Platonia, so the path truly ends there. This is
beginning to look quite promising as the basis for a total
explanation of time, but several conditions must be met. Before we
address them, it is worth saying a little about gravity and
thermodynamics.

The central conclusion of standard thermodynamics with an

external time is that, if the low entropy of the world and its habitual
increase are to be explained, the universe must presently be
evolving out of a statistically most unlikely state. In systems in
which gravity does not act, the unlikely state is generally one that is
structured, while the likely state is characterized by a bland
uniformity. Gas con ned in a nite volume tends quickly to a very
uniform state in which it occupies all the available space and all
temperature di erences are levelled out. This is the equilibrium
state. It is vastly more probable than any ordered state because
there are so many more ways in which it can be realized
microscopically. The situation is much more complicated when
gravity comes into play, since there is no well-de ned equilibrium
state for a gravitating system. Gravity is attractive, so a uniform
state is unstable and will tend to break into self-gravitating clumps.
This is the exact opposite of a gas.

Currently there is no fully satisfactory thermodynamics of

cosmology, mainly because of the way in which gravity acts. But it
does seem certain that black holes, which almost certainly exist,
have a well-de ned entropy associated with them. This was the nal
and most dramatic outcome of the intensely exciting ‘golden decade’
in the study of black holes that ended with the discovery of black-
hole evaporation by Stephen Hawking in 1974. This fascinating
story has been told with great verve by Kip Thorne in his Black
Holes and Time Warps. The entropy associated with black holes is
staggeringly large. Since there is little evidence that black holes
existed at very early times but a lot that many have since been
formed and more will be formed, the universe seems to have begun
in an extraordinarily unlikely state.

No one has done more than Roger Penrose to highlight this fact.
His The Emperor’s New Mind has an entertaining illustration of the
creating divinity seeking with a pin to nd the tiny improbable
point of the initial condition of the universe from which its utterly
unlikely history must have sprung. Penrose seeks to explain this in a
theory in which both time and actual quantum-mechanical collapse
are real, and the laws of nature are inherently asymmetric in time.
My approach is quite di erent because I think that the whole
problem of time and its arrow can – paradoxically – be formulated
more precisely and transparently in a context in which time does
not exist at all. I also believe that, far from being highly unlikely,
the kind of history and cosmos we experience are characteristic and
likely in a timeless scenario.

It all depends on how a static wave function ‘beds down’ on the

starkly asymmetric continent of Platonia. The issue of the correct
arena is all-important. The collective intuition of most physicists,
wedded to time and honed on translucent structures like absolute
space, is forced to see the observed universe as highly improbable.
But in Platonia it may appear inevitable. Wave functions have a way
of nding special structures: for example, they can create complex
molecules like proteins and DNA.

Let it be granted, not unreasonably I think, that the wave

function of the universe will be semiclassical with respect to at least
some variables in some part of Platonia. Where is it likely to be, and
how will the corresponding Hamiltonian ‘light rays’ run? Do they
emerge from Alpha? Here, one of the most famous results of
classical general relativity may be relevant. Penrose and Hawking
showed that its solutions have a remarkable propensity to evolve
into a singular state. All that is necessary is for su cient matter to
be concentrated within a certain nite region. After that, as Penrose
showed, collapse to a black hole is inevitable. Hawking showed that
there is a sense in which the Big Bang itself can be regarded as the
time-reverse of the Penrose collapse to a black hole. (Collapse here
has nothing to do with quantum-mechanical collapse of the wave
function.) Solutions that terminate at one or both ends in singular
states are characteristic of general relativity.

What happens in a quantum theory cannot be totally unrelated

to the corresponding classical theory. It therefore seems likely that
in quantum gravity there will be a semiclassical region near the
central ray in Figure 55. The Hamiltonian ‘light rays’ in it may well,
re ecting the structure of Platonia, appear to emanate from Alpha,
and rise up in a kind of jet which then spreads out and falls back, as
in a fountain, returning to small volumes but in a much more
irregular state. Alternatively, they may go on for ever, receding ever
farther from Alpha. I have described these trajectories as if they
were traced in time, but they are only paths.

Moreover, the paths are still only ‘seeds’. The nding of the full
rich structures which tell us so insistently that time exists and ows
must result from entanglement with the host of the remaining
quantum variables that constitute the expanses of Platonia. When
discussing alpha-particle tracks, I emphasized that Mott employed a
special device to concentrate the wave function on time capsules.
Considered purely in terms of the stationary Schrödinger equation,
this was arti cial. This is what created the static alpha-particle
tracks and such a strong sense of time and history out of the ‘seed’
of a spherical wave pattern.

If my proposal is along the right lines, there must be some

natural and plausible mechanism within static quantum cosmology
which performs this task. Platonia at large must force it to happen.
As I have already said, I see the cause in the rooted asymmetry of
contingent things. Platonia is necessarily skew. It is easy to imagine
that the cone of Figure 55 ‘funnels entanglement outwards’, much as
a trumpeter blows air from a bugle. I deliberately chose this last
simile. The bugle does create a nice image of what I have in mind,
but it also creates hot air. There are no hard mathematical proofs to
support my idea, but I hope you are now persuaded that at least the
arguments for a timeless universe are strong. If it nevertheless
appears intensely temporal, there must somewhere be a massive
reason for the fact. I think it is the asymmetry of being. Being can
be more or less. Sitting in the midst of things, we feel ourselves
carried forward on the mighty arrow of time. But it is an arrow that
does not move. It is simply an arrow that points from the simple to
the complex, from less to more, most fundamentally of all from
nothing to something. If we could look over our shoulder in
Platonia, we should see where this trek began: at the edge of
nothing.
 Figure 55 Explosion out of the Big Bang and recontraction to the
Big Crunch according to present standard cosmology.

A WELL-ORDERED COSMOS?

Let me end the main part of the book with a few comments on
structure, and what strikes a theoretical physicist as improbable. If
we think that dynamical histories in space and time are the
fundamental things in nature, then all statistical re ections on the
world lead to great di culties. Most histories are unutterably boring
over all but a minuscule fraction of their length. We can never
understand the miracle of the structured world. Things are
completely changed if quantum cosmology is really about some
well-behaved distribution of a static wave function over Platonia.
The con gurations at which ψ collects strongly must be special – in
some sense they must resonate with all the other con gurations that
are competing for wave function. Quantum cosmology becomes a
kind of beauty contest in eternity. The winners – those that get a
high probability density – must be exceptional, like the DNA
molecule. This is just the opposite of what classical physics leads us
to expect. There, the winners are boring.

When I wrote my book on the history of dynamics, I was exposed

to the beautiful poetic notion of the well-ordered cosmos.
Intermittent reading of Leibniz had already made me deeply
interested in structure, and this was greatly strengthened by my
collaboration with Lee Smolin on the Leibnizian idea that the actual
universe is more varied than any other conceivable universe. That
still remains a mere idea (though Paul Davies was su ciently
intrigued to include a brief account of the idea in his The Mind of
God), but I became persuaded that a scienti c theory of the universe
in which structure is created as a rst principle is possible. We may
need to get back to the wonder of childhood to comprehend what
the world really is. Yeats once wrote of Bishop Berkeley that he ‘has
brought back to us the world that only exists because it shines and
sounds. A child, smothering its laughter because the elders are
standing around, has opened once more the great box of toys.’

The single most striking thing about the universe we see around
us is its rich structure, which is so di cult to understand on a priori
statistical grounds. Until the modern scienti c age, all thinkers saw
the rst task of science as being the direct description and
explanation of this structure. This natural impulse is re ected in the
Pythagorean notion of the well-ordered cosmos. It was still very
strong in both Kepler and Galileo. However, when Newton
demonstrated the supreme importance of accelerations in dynamics,
the perspective of science changed, for the world at the present
instant became the mere consequence of its initial conditions.
Instead of asking directly how structure is fashioned, science turned
to asking how it is refashioned.

It seems to me that the abolition of time in quantum gravity

must bring us back to a more Pythagorean perspective, though with
a quantum slant, for now we must simply ask what structures are
probable. It seems to me that the rst decisive step in this direction
was the discovery by Schrödinger of his time-independent wave
equation, with its all-important condition that its solutions must be
well behaved, and Born’s probability interpretation of quantum
mechanics. For we know that these two basic elements of quantum
mechanics work together to bring forth exquisite structures in great
profusion, doing so moreover without any boundary or initial
conditions and with total disregard for what might seem statistically
likely. That is the story of atomic, molecular and solid-state physics.
I think it may even be the story of the universe.
EPILOGUE
Life Without Time

Pied Beauty

Glory be to God for dappled things –

For skies of couple-colour as a brinded cow;
For rose-moles all in stipple upon trout that swim;
Fresh- recoal chestnut-falls; nches’ wings;
Landscape plotted and pieced – fold, fallow, and plough;
And áll trádes, their gear and tackle and trim.

All things counter, original, spare, strange;

Whatever is ckle, freckled (who knows how?)
With swift, slow; sweet, sour; adazzle, dim;
He fathers-forth whose beauty is past change:
Praise him

It is a pity that Gerard Manley Hopkins’s nest line in this poem

implies that creation is a male prerogative and is so inappropriate
for the dawning millennium. But what beauty past change the wave
function does manage to nd in the nooks and crannies of Platonia!
What are we to think of life if time and motion are nothing but very
well-founded illusions? I have selected a few topics, trying to
anticipate some of the questions that the reader, as a human being
rather than a scientist, might ask. I also give some hints of how I
think the divide between impersonal science and the world of the
arts, emotions and religious aspirations might be bridged
somewhere in Platonia. I love and respect the disciplines of both.
Can they be shown to ow from a common view of the world?

Can We Really Believe in Many Worlds?

The evidence for them is strong. The history of science shows that
physicists have tended to be wrong when they have not believed
counterintuitive results of good theories. However, despite strong
intellectual acceptance of many worlds, I live my life as if it were
unique. You might call me a somewhat apologetic ‘many-worlder’!
There are occasions when the real existence of other worlds, other
outcomes, seems very hard to accept. Soon after I started writing
this book, Princess Diana was killed, and Britain – like much of the
world – was gripped by a most extraordinary mood. Watching the
funeral service live, I did wonder how seriouslyone can take a
theory which suggests that she survived the crash in other worlds.
Death appears so nal.

Such doubts may arise from the extraordinary creative power –

whatever it is – that lies behind the world. What we experience in
any instant always appears to be embedded in a rich and coherent
story. That is what makes it seem unique. I would be reassured if
the blue mist did indeed seek out only such stories. Shakespeare
wrote many plays, nearly all masterpieces. But we do not even have
a unique Hamlet: producers are always cutting di erent lines, and
producing the play in novel ways. Variety is no bad thing: I have
enjoyed many outstanding Hamlets. In the timeless many-instants
interpretation, they were all other worlds, and that is what makes
timeless quantum cosmology fascinating. Our past is just another
world. This is the message that quantum mechanics and the deep
timeless structure of general relativity seem to be telling us. If you
accept that you experienced this morning, that commits you to other
worlds. All the instants we have experienced are other worlds, for
they are not the one we are in now. Can we then deny the existence
of worlds on which ψ collects just as strongly as on our remembered
experiences?

Does Free Will Exist?

Anyone committed to science has di culty with free will. In The

Sel sh Gene (2nd edition, pp.270-71), Dawkins asks, ‘What on earth
do you think you are, if not a robot, albeit a very complicated one?’
From personal introspection, I do not believe that my conscious self
exercises free will. Certainly I ponder di cult decisions at length,
but the decision itself invariably comes into consciousness from a
di erent, unconscious realm. Brain research con rms that what we
think are spontaneous decisions, acts of free will, are prepared in
the unconscious mind before we become aware of them.

However, the many-instants interpretation puts an intriguingly

di erent slant on causality, suggesting that it operates in nothing
like the way we normally believe it to. In both classical physics and
Everett’s original scheme, what happens now is the consequence of
the past. But with many instants, each Now ‘competes’ with all
other Nows in a timeless beauty contest to win the highest
probability. The ability of each Now to ‘resonate’ with the other
Nows is what counts. Its chance to exist is determined by what it is
in itself. The structure of things is the determining power in a
timeless world.

The same applies to us, for our conscious instants are embedded
in the Nows. The probability of us experiencing ourselves doing
something is just the sum of the probabilities for all the di erent
Nows in which that experience is embedded. Everything we
experience is brought into existence by being what it is. Our very
nature determines whether we shall or shall not be. I nd that
consoling. We are because of what we are. Our existence is
determined by the way we relate to (or resonate with) everything
else that can be. Although Darwinism is a marvellous theory, and I
greatly admire and respect Richard Dawkins’s writings, one day the
theory of evolution will be subsumed in a greater scheme, just as
Newtonian mechanics was subsumed in relativity without in any
way ceasing to be great and valid science. For this reason, and for
the remarks just made, I do not think that we are robots or that
anything happens by chance. That view arises because we do not
have a large enough perspective on things. We are the answers to
the question of what can be maximally sensitive to the totality of
what is possible. That is quite Darwinian. Species, ultimately genes,
exist only if they t in an environment. Platonia is the ultimate
environment.

In Box 3, I said that Platonia is a ‘heavenly vault’ in which the

music of the spheres is played. This formulation grew out of
numerous discussions with the Celtic composer, musicologist and
poet John Purser (brother of the mathematician and cryptographer
Michael, who made the comment about my parents with which I
ended the Preface, and brought to my attention the Shakespeare
quotation). With the inimitable assurance of which only he is
capable, John is adamant that the only theory of the universe that
ever made sense was (is) the music of the spheres. My guts tell me
that he and the artists quite generally are right. But harmony rests
on mathematics, of course. Rather appropriately, given my extensive
use of meteorological metaphor, John and his wife Bar live in the
misty Isle of Skye, where at least one of the said discussions took
place while the better part of a bottle of whisky was consumed,
mostly by John.

You will naturally ask why we do not hear this music of the
spheres. Keats provides a rst answer: ‘Heard melodies are sweet,
but those unheard Are sweeter’. But Leibniz may have given the true
answer. In his monadology, he teaches that the quintessential you,
everything you experience in consciousness and the unconscious, is
precisely this music. You are the music of the spheres heard from
the particular vantage point that is you. This is taking a little liberty
with the letter but certainly not the spirit of his great philosophical
scheme. On the subject of liberties, I have taken fewer with Leibniz
than Michael did with Shakespeare. Hal does not actually ask
Falsta (‘fat-witted with drinking of old sack’) why he should be so
super uous ‘to inquire the nature of time’ but to ‘demand the time
of day’. But, were it not for the blessed Sun and its diurnal rotation
(our fortunate circumstances), the one question would be as
profound as the other.

Now for something like the original Gretchenfrage.

Is There a Role for a Creator in Quantum Cosmology?

Perhaps, but it is a somewhat strange one. It seems to me that

science can never do more than guess – theorize about – the
structure of things and then test to see whether its conjectures are
con rmed. This is an open-ended venture (with tremendous
successes behind it) and always presupposes that there is some
structure already out there waiting to be found. In the scheme I
have advanced, much is presupposed: Platonia, its detailed structure
(immensely important) and a wave function that ‘samples’
possibilities. It is the nature of theory to presuppose something, so
that always leaves a potential role for a Creator. But does invoking
something to explain what we cannot explain get us any further?

What does intrigue me is the power of structures in a timeless

scheme. They determine where the wave function collects. If one
wanted to see ψ as spirit pondering what shall be brought into
existence, it has no power in the matter. Leibniz always said that
not even God could escape the dictates of reason. He must always
act rationally. Perhaps that is more reassuring than a capricious
deity is. However, a rational universe is quite alarming too. If you
are about to perish in a concentration camp, is it any consolation to
know that what must be will be?

In The Life of the Cosmos, my friend Lee Smolin espouses a self-

creating universe, likening its growth to the often largely unplanned
development of cities. I nd his epilogue especially eloquent. In fact,
timeless quantum cosmology does give almost god-like power to
structures, ourselves included, to bring themselves into being. We
shall be if that ts the great scheme of things. The ideas of both Lee
and myself tend to pantheism. The whole universe – Platonia and
the wave function – is the closest we can get to a God.

Where Is Heaven?

I have long thought that, if only we had the wit to see it, we are
already in heaven. It is Platonia. I say this with some trepidation,
though I believe it is true. If so, Platonia must be hell and purgatory
as well. What I mean by this is really quite simple: some places in
Platonia are very admirable, pleasant and beautiful, many are
boring in the extreme, and others are horrendously nasty. The same
contrasts exist within the individual Nows. What we do not know is
where the wave function collects.

I certainly nd it di cult to believe that there is a material

world in which we currently nd ourselves, and some other, quite
di erent, immaterial world we enter after death. Apart from
anything else, modern physics suggests very strongly that so-called
gross matter – the clay from which we are made – is anything but
that. It is almost positively immaterial. Platonic forms have exact
mathematical properties, and those are all that physicists need to
model the world and to attribute to matter.

I also feel strongly that this created world is something to be

marvelled at and cherished, not dismissed as some second-best
version of what is yet to come. Disrespect for this world is disrespect
for whatever creates it. I shall not attempt to argue about these
things in detail here, but the total elimination of time, if accepted
and supported by mathematics and observation, must force
theologians to reconsider their notions. If there is a happier and
more perfect world, in which the lion lies down with the lamb and
the sword is made into a ploughshare, I think it will simply be
somewhere else in Platonia. I am sure that there are locations where
experience is much deeper and richer than here. Such experience
may be perfectly timeless – consciousness just sees what is. Perhaps
we are somehow included in that awareness. Perhaps too the world
is redeemed, and its inner con icts resolved and understood
somewhere in Platonia’s distant reaches, farther from Alpha than we
are.

It is not for nothing that I emphasized in the early part of the

book that Platonia has an Alpha but no Omega. The idea of a Point
Omega was introduced by the Jesuit biologist Teilhard de Chardin,
who conceived of it as some kind of consummation of evolution in
the ultimate future, ‘on the boundary of all future time’. I have
quoted these last words from Frank Tipler’s book The Physics of
Immortality, in which he argues that Point Omega is where our
material universe recontracts to the Big Crunch. By then, he argues,
intelligences will have become so adept and computerized that we
shall all be recreated as virtual computer programs, run so fast that
we have an e ective eternity of existence in which we are
resurrected before the universe ends in the Crunch.

I can only say that is not how I see things. I search in vain for
Omega in Platonia and nd only Alpha. But Platonia is a vast land.
Let us cherish everything around us wherever we happen to nd
ourselves in the Platonic palace.

Of one thing I feel very sure. Many poets and theologians give a
misleading image of heaven and eternity. Consider the opening lines
of Vaughan’s famous poem The World’:
 I saw Eternity the other night
Like a great Ring of pure and endless light,

All calm as it was bright;

And round beneath it, Time, in hours, days, years,

Driven by the spheres,

Like a vast shadow moved, in which the world

And all her train were hurled.

This is magni cent poetry as a description of the mystical state in

which mind comprehends all structure as uni ed. But the real
wonder of unity is when it knits together rich variety. Bliss is
comprehension of many things in harmony at once. I do not think
that eternity is pure and endless light. Such light merely illuminates,
for example, the millions of leaves of the forest of the American fall
when we see them all at once.

Is Time Travel Possible?

Time travel of a sort is possible within general relativity as a

classical theory, but is subject to strict limitations. You cannot travel
back into the past and kill your parents before your conception. In
quantum cosmology, you can travel back to a parallel universe, and
there kill your parents before they conceive you. However, we have
to be careful about the use of ‘you’. The person who ‘travels’ to
these other worlds is not exactly you now. As the discussion of the
haemoglobin molecule showed, the change within our bodies from
one instant to another is stupendous. The fact that we have such an
enduring sense of deep continuity of our personal identity is very
remarkable. I see it as another manifestation of the creative power
that brings everything into existence. Stephen Hawking long
suspected that even if time travel is logically possible, it will have a
very low probability in quantum cosmology. That is my feeling too.
Platonia certainly contains Nows in which there are beings whose
memories tell them they have travelled backwards in time.
However, I think such Nows have a very low probability.

To tell the truth, I nd the idea of time travel boring compared

with the reality of our normal existence. Each time capsule that
represents an experienced Now re ects innumerable other Nows all
over Platonia, some of them vividly. In a very real sense, our
memories make us present in what we call the past, and our
anticipations give us a foretaste of what we call the future. Why do
we need time machines if our very existence is a kind of being
present everywhere in what can be? This is very Leibnizian. We are
all part of one another, and we are each just the totality of things
seen from our own viewpoint.

Doesn’t the Denial of Motion Take All Joy and Verve out of Life?

I do feel this issue keenly. The king sher parable should make that
clear. In principle, there is no reason why we should not attempt to
put our very direct sense of change directly into the foundations of
physics. There is a long tradition, going back at least to Hamilton,
that seeks to make process the most basic thing in the world.
Roughly, the idea is that physics should be built up using verbs, not
nouns. In 1929 the English philosopher Alfred North Whitehead
published an unreadable – in my experience – book called Process
and Reality in which he advocated process. It all sounds very
exciting, but I just do not think it can be done, despite a valiant
attempt by Abner Shimony. Having translated seventy million words
of Russian into English, I can say with some feeling that sentences
do have a subject and generally an object. I could have written this
book using the one verb ‘to be’, which hardly counts as a verb. For
this reason it seldom appears in Russian; when it does, it is most
often as a surrogate: ‘to appear’. But a book without nouns is
nothing. Not even James Joyce could write it. For some reason,
disembodied verbs exert a fascination not unlike the grin of the
Cheshire cat. But when Owen Glendower claimed to be able to ‘call
up spirits from the vasty deep’, Hotspur answered: ‘Aye, and so can I
and any man, but will they come when you call them?’ I should like
to see it done.

Less provocatively, I wonder if, at root, there is that much

di erence between the Heraclitan and Parmenidean schools,
representing ‘verbs’ and ‘nouns’ respectively. If my de nition of an
instant of time is accepted, it becomes hard to say in what respect
those two great Pre-Socratics might di er. The two best-known
sayings attributed to Heraclitus are ‘Everything ows’ (Panta rei)
and the very sentence which, entirely unconsciously, I used to clinch
the argument that the cat Lucy who leapt to catch the swift was not
the cat who landed with her prey: ‘One cannot step into the same
river twice.’ There is always change from one instant to another –
no two are alike. But that is just what I have tried to capture with
the notion of Platonia as the collection of all distinct instants.
Heraclitus argued that the appearance of permanence, of enduring
substance, is an illusion created by the laws that govern change.
Obviously, he and Parmenides could not be expected to have
anticipated quantum mechanics, wave functions and the Wheeler-
DeWitt equation, But I see Bell’s account of alpha-particle track
formation as remarkable support for the Heraclitan standpoint that
appearances are the outcome of the laws that bring them forth.
Would it not be a wonderful reconciliation of opposites if the static
wave function were to settle spontaneously on time capsules that
are redolent of both ux (evidence of history) and stasis (evidence
that things endured through it)?

But the loss of motion is still poignant, a premonition of

mortality and a view of our life from outside it. There is a scene in
George Eliot’s Middlemarch in which Ladislaw and a painter friend
chance to see the heroine Dorothea in a particularly striking pose in
Rome. The painter is keen to capture it on canvas, but Ladislaw
taunts him:

Painting stares at you with an insistent imperfection. I feel

that especially about representations of women. As if a
 woman were a mere coloured super cies! You must wait for
movement and tone. There is a di erence in their very
breathing: they change from moment to moment.

Keats too, for all the beauty of his Grecian urn, addresses it with the
words

Thou, silent form! Dost tease us out of thought

As doth eternity. Cold Pastoral!

When Keats wrote these lines, he must have known that all too soon
his home would be a grave. Is Platonia a graveyard? Of a kind it
undoubtedly is, but it is a heavenly vault. For it is more like a
miraculous store of paintings by artists representing the entire range
of abilities. The best pictures are those that somehow re ect one
another. These are the paintings we nd there in profusion. There
are very few of the mediocre, dull ones. Despite what Ladislaw says,
the best paintings have a tremendous vibrancy. Turner does almost
bind you to the mast of the Ariel. Indeed, Ladislaw’s own words
immediately before the passage quoted above are: ‘After all, the true
seeing is within.’ Frozen it may be, but Platonia is the demesne
where ‘Beauty is truth, truth beauty’ and the boughs cannot shed
their leaves ‘nor ever bid the Spring adieu’. With that perfect ode,
Keats did achieve the immortality for which he so desperately
longed.
 In a ne essay entitled ‘The timeless world of a play’, Tennessee
Williams praises great sculpture because it

often follows the lines of the human body, yet the repose of
great sculpture suddenly transmutes those human lines to
something that has an absoluteness, a purity, a beauty,
which would not be possible in a living mobile form.

He argues that a play can achieve the same e ect, and so help us to
escape the ravages of time. ‘Whether or not we admit it to ourselves,
we are all haunted by a truly awful sense of impermanence.’ Again,
this is very beautiful writing, but has Williams failed to see the truth
all around us – the Platonic eternity we inhabit in each instant? Is it
blindness that drives him to seek eternity? Some people can pass a
cathedral without noticing it.

The desire for an afterlife is very understandable, but we may be

looking for immortality in the wrong place. I mentioned
Schrödinger’s curious failure to recognize in his own physics the
philosophy of ancient India (especially the Upanishads) he so
admired. There is a beautiful passage at the end of his epilogue to
What is Life? that nevertheless strikes me as wishful thinking. He
poses the question, ‘What is this “I”?’ Here is part of his answer:

If you analyse it closely you will, I think, nd that it is just a

little bit more than a collection of single data (experiences
and memories), namely the canvas upon which they are
 collected. And you will, on close introspection, nd that
what you really mean by ‘I’ is that ground-stu upon which
they are collected. You may come to a distant country, lose
sight of all your friends, may all but forget them; you
acquire new friends, you share life with them as intensely as
you ever did with your old ones … Yet there has been no
intermediate break, no death.

Thus, he argues that our personal ‘ground-stu is imperishable,

holding us together through all the changes of life. He ends with
this a rmation of faith: ‘In no case is there a loss of personal
existence to deplore. Nor will there ever be.’ But earlier he had
praised the great Upanishads for their recognition that ATHMAN =
BRAHMAN (the personal self equals the omnipresent, all-
comprehending eternal self). He seems to want to have his cake and
eat it, to be dissolved in the all-comprehending eternal self yet still
retain a personal identity. He is not nding a canvas, he is clutching
for a straw.

Tennessee Williams faced up to things more squarely: ‘About

their lives people ought to remember that when they are nished,
everything in them will be contained in a marvelous state of repose
which is the same as that which they unconsciously admired in
drama.’ Exactly: this is the truth and beauty of Keat’s Grecian urn.
Williams goes on: ‘Snatching the eternal out of the desperately
eeting is the great magic trick of human existence.’ Yes, though it’s
not a magic trick but simply the opening of our eyes.
 Some years ago, I heard Dame Janet Baker interviewed on radio.
She was asked if she ever listened to her recordings and, if so, what
were her favourites. She said she almost never listened to them. For
her, every Now was so exciting and new, it was a great mistake to
try to repeat one. In her singing she made no attempt at all to
recreate earlier performances and do the high points the same way
as the night before. Again and again she spoke with the deepest
reverence of the Now and how it should be new and happen
spontaneously. ‘The Now is what is real’, she said.

I thought it was the perfect artistic expression of how I see

timeless quantum cosmology. By de nition, every Now in Platonia
is new, for all the Nows are di erent. But some are vastly more
interesting and exciting than others. Miraculously, these are the
Nows that the wave function of the universe seems to nd with
unerring skill.

What a gift too is the specious present. Appreciation of poetry

and music would be impossible without it. I can live without motion
if I can sense it as the line that runs through a story all bound up in
one Now. Janet Baker is right. Watching motion, listening to
Beethoven, looking at a painting by Turner – all are given to us in
the Now, which we experience as the specious present. Einstein
seems to have regretted that modern science – and his own
relativity in particular – had taken the Now, the vibrant present, out
of the world. On the contrary, I think the Now may well constitute
the very essence of the physical world, the rst quantum concept (as
David Deutsch refers to time). The artists always knew it was there,
and worshipped at its altar. It was Dirac’s rediscovery of the Now at
the heart of general relativity that started my quest.

I do also feel that novelty is a genuine element of quantum

mechanics, especially in the many-worlds form, not present in
classical mechanics. In the main text I spoke of lying down to sleep
and knowing ‘not what we should wake to’. I see no fundamental
line of time and causal evolution along which we march as robots;
each experienced Now is new and distinct. I think that the many-
worlds hypothesis is the scienti c counterpart of the thrill of artistic
creation that Janet Baker feels so strongly. It is something
essentially new for which there is no adequate explanation in any
supposed past from which we have tumbled via a computer
algorithm. There is no explanation of any one triangle in terms of
any others, and the same is true of all Nows.

The Italian painter Claudio Olivieri, a friend of Bruno Bertotti,

creates paintings in which he evokes a sense of timelessness. He
expresses his aim through this poem:

È con la pittura che le apparenze si mutano in apparizioni:

Ciò che è mostrato non è la verosimiglianza ma la nascita.
È così che ci viene restituto il nostro presente,
L’assolutamente unico ma imprevedibile presente,
Somma di tutti tempi, raduno degli attimi che ci fanno
Viventi, atto sempre inaugurale dell’esistere.
A free translation is as follows:

The painting trans gures semblance in sudden apparition,

showing not likeness but birth.
That is how we are given back our present,
The absolutely unique but unforeseeable present,
sum of all times, gathering of the moments that make us
alive,
the ever inaugural act of existence.

This does express the main ideas I have tried to get across in the
nal part of the book. Each experienced instant is a separate
creation (birth), the ever inaugural act of existence, brought to life
by the gathering of all times. The thrill that Janet Baker experiences
in each Now is the assolutamente unico ma imprevedibile presente, that
nding of ourselves in one of the instants that quantum mechanics
makes resonate especially strongly with other instants.

As I began, so I end. Turner has taught us the way to look at the

world, and even how to come to terms with many worlds. Once any
painting of his had reached a certain stage of completion, all
additions to it became simply variations on an existing masterpiece.
All the stages through which his paintings then passed were perfect,
and each was – is – a separate world. Nature is an even more
consummate artist than Turner. For he too is part of Nature. Turner
is also right in the way he places us humans in the great arena. In
nearly all his pictures, human beings, though tiny on the cosmic
scale, are integral parts of some huge picture, Keats’s urn painted
large. We are simultaneously spectators and participants, subtly
changing and constantly working on an inherited landscape. We are
there in one place but bound up into something much larger.
Gretchen Kubasiak gave me, besides the Tennessee Williams essay,
some Aborigine philosophy that, but for the idea that we are
visitors, chimes with this thought:

We are all visitors to this time, this place. We are just

passing through. Our purpose here is to observe, to learn, to
grow, to love... And then we return home.

No, this is home. Mach once commented that ‘In wishing to preserve
our personal memories beyond death, we are behaving like the
astute Eskimo, who refused with thanks the gift of immortality
without his seals and walruses.’ I am not going without them, either.
I cannot even if I wanted to: they are part of me. Like you, I am
nothing and yet everything. I am nothing because there is no
personal canvas on which I am painted. I am everything because I
am the universe seen from the point, unforeseeable because it is
unique, that is me now. C’est moi. I am bound to stay. We all watch
—and participate in—the great spectacle. Immortality is here. Our
task is to recognize it. Some Nows are thrilling and beautiful beyond
description. Being in them is the supreme gift.
NOTES

PREFACE

(1) (p. 2) The article about Dirac appeared in the Süddeutsche

Zeitung for Friday, 18 October 1963, and was based on an article by
Dirac that appeared in Scienti c American in May 1963.

(2) (p. 4) On hearing about my plans for this book, Michael Purser
brought to my attention the following rebuke from Prince Hal to
Falsta :

Unless hours were cups of sack, and minutes capons, and

clocks the tongues of bawds, and dials the signs of leaping-
houses, and the blessed sun himself a fair hot wench in ame-
colour’d ta eta, I see no reason why thou shouldst be so
super uous to inquire the nature of time.

Henry IV, Part I (1. ii)

(I comment on this in the Epilogue.)

CHAPTER 1: THE MAIN PUZZLES

The Next Revolution in Physics (p. 14) The possible non-existence
of time has just begun to be discussed in authoritative books for the
general public. Both Paul Davies, in his About Time, and Kip Thorne,
in his Black Holes and Time Warps, devote a few pages to the topic.
In apocalyptic vein, Thorne likens the fate of space-time near a
black hole singularity to

a piece of wood impregnated with water . . . the wood

represents space, the water represents time, and the two
(wood and water, space and time) are tightly interwoven,
uni ed. The singularity and the laws of quantum gravity
that rule it are like a re into which the water-impregnated
wood is thrown. The re boils the water out of the wood,
leaving the wood alone and vulnerable; in the singularity
the laws of quantum gravity destroy time . . . (p. 477)

However, Thorne’s magni cent book is devoted to other topics, and

nothing prepares the reader for this dramatic and singular end of
time. Moreover, the evidence, as I read it, is that timelessness
permeates the whole universe, not just the vicinity of singularities.
Paul Davies, for his part, repeatedly expresses a deep mysti cation
about time. His book is almost a compendium of conundrums, and
he candidly consoles the reader with ‘you may well be even more
confused about time after reading this book than you were before.
That’s all right; I was more confused myself after writing it’ (p. 10).
In fact, I think Paul’s subtitle, Einstein’s Un nished Revolution, is the
key to a lot of the puzzles. As we shall see in Part 3, there are
aspects of physical time which Einstein did not address.

Among the popular books that I know, the two that undoubtedly
give most prominence to the problem of time in quantum gravity
are Lee Smolin’s The Life of the Cosmos, which contains some
discussion of my own ideas, and David Deutsch’s The Fabric of
Reality. There is considerable overlap between my book and
Deutsch’s chapter ‘Time: the rst quantum concept’. One technical
book, now going into a third edition, that from the start has taken
timelessness very seriously is Dieter Zeh’s The Physical Basis of the
Direction of Time.

It may be that the reason why a book like this one, devoted
exclusively to the idea that time does not exist, has not hitherto
been published by a physicist has a sociological explanation. For
professionals working in institutes and dependent on the opinions of
peers for research funding, such a book might damage their
reputation and put further research in jeopardy. After all, at rst it
does seem outrageous to suggest that time does not exist. It may not
be accidental that I, as an independent not reliant on conventional
funding, have been prepared to ‘come out’.

In this connection, my experience at a big international

conference in Spain in 1991 devoted to the arrow of time was very
interesting. The following is quoted from my paper in the
conference proceedings (available in paperback as Halliwell et al.,
1994):
 During the Workshop, I conducted a very informal straw-poll,
putting the following question to each of the 42 participants:

Do you believe time is a truly basic concept that must appear in the
foundations of any theory of the world, or is it an e ective concept
that can be derived from more primitive notions in the same way
that a notion of temperature can be recovered in statistical
mechanics?

The results were as follows: 20 said there was no time at a

fundamental level, 12 declared themselves to be undecided
or wished to abstain, and 10 believed time did exist at the
most basic level. However, among the 12 in the
undecided/abstain column, 5 were sympathetic to or
inclined to the belief that time should not appear at the
most basic level of theory.

Thus, a clear majority doubted the existence of time. When I

took my straw-poll, I said that I intended to publish the names with
their opinions, which was why two people abstained, to remain
anonymous. As it happens the conference generated immense media
interest in Spain, not least because of the presence of Stephen
Hawking and Nobel Laureate Murray Gell-Mann, and the reporter
from El Pais got hold of a copy of my results. One of the participants
(neither of the above), nding his own opinion quoted in a big
article the day after the conference, was none too pleased and
greeted me when we met six months later at a conference in
Cincinnati with ‘You and your damned straw-poll!’ I then realized
why the editors had meanwhile asked me to withhold the names in
my paper, which I happily did.

It was at the later conference that I learned a bon mot of Mark

Twain that somehow seems appropriate here: ‘If the end of the
world is nigh, it is time to be in Cincinnati. Everything comes to
Cincinnati twenty years late.’

The Ultimate Things (p. 15) I mentioned in the Preface the

di culty of writing without using temporal notions. The curious
state of modern physics as outlined in Box 2 compounds the
problem. Because quantum theories are obtained from classical
theories by so-called quantization, and classical concepts are much
closer to everyday experience, the language used by most physicists,
myself included, often seems to imply that the classical theories are
somehow deeper than the quantum theories obtained from them.
But that is certainly only a re ection of our way to the truth. What
is needed is a clear language in which to describe the quantum truth
directly and an explanation, based on it, of why the world appears
classical to us. I am proposing the notion of a Now as the basic
quantum notion.

Getting to Grips with Elusive Time (p. 17) The idea that instants
of time are distinct entities that should not be thought of as joined
up in a linear sequence is a powerful intuitive experience for at least
one non-scientist. A few days after the Sunday Times published its
article ‘Time’s assassin’ about my ideas in October 1998, I received
by email a ‘Question for Julian Barbour’ from Gretchen Mills
Kubasiak, who had read the article about me. She introduced herself
with: I am merely a girl who lives in Chicago, works for a
construction company and nds herself thoroughly captivated by
your ideas. In fact, I have been unable to think of little else this past
week.’ She asked if she could put a question to me. Well, who could
resist that request? I said yes, asking if by any chance, with her rst
name, she had German ancestry, and commented: ‘I guess you know
the German expression Gretchenfrage and its origin in Goethe’s Faust,
when Gretchen asks Faust about his attitude to religion and if he
believed in God. It was especially nice to get your Gretchenfrage.’
Subsequent correspondence persuades me that ‘merely a girl’ might
not be the most accurate description of her, since she is a voracious
reader and traveller (among much else). Some of her thoughts about
time are worth passing on:

Several weeks before I read the London Times article which

brought your ideas to my attention, I started having a debate
with a friend of mine on traveling. He stated that when a
person travels between two places, it is the time spent on the
journey which makes the person able to appreciate and
comprehend the nal destination. Only by making a linear
tour of the world and having a passage of time connect the two
locations are we able to understand our nal destination.
 I disagreed. I have always believed that our lives are made
up of individual moments that layer and co-exist with other
moments, not a linear sequence of events. I did not accept his
notion that time spent on a journey is relative to one’s
experience at their nal destination. The passage of time, that
for my friend constituted the journey, did not exist for me.
That is not to say that what he viewed as his journey did not
consist of moments but I could not accept that they were
relative to the moment of the nal destination simply because
they preceded it.

Despite the fact that I had these beliefs in my head, I found

that I lacked the vocabulary to make a satisfactory argument
on paper. It is one thing to state your beliefs and quite another
to be able to back up your argument. I had developed a few
descriptive examples of moments in my life that I believed
began to illustrate this idea but I knew of nothing that would
support them.

One of my ideas addressed my moments with Buckingham

Palace. As a small child I had listened to my mother recite the
poem about Christopher Robin’s visit to the changing of the
guard and I stood silently alongside him and Alice. As a young
girl I watched on television the newly married Prince and
Princess of Wales venture forth onto the balcony to greet their
public and I stood among the crowds. In both instances I was
not ‘there’ and yet I was. When I actually stood in front of the
palace as a teenager, the physical journey associated with that
moment mattered not. What mattered were these other
moments. When I stood in front of the palace, I was living not
just that moment but co-existing with the other moments as
well.

Then I came across the London Times article outlining your

notion of the illusion of time and a spark of recognition within
me was lit. Something I had always felt, but had never been
able to express, was suddenly being put into words.

If, as you say, all moments are simultaneous and there is no

linear sequence of events, does this not imply that the ‘length’
of a journey is completely irrelevant? If we exist in isolated
moments, then the notion that time spent on a journey makes
the experience cannot be true because time does not exist. If
time is merely an illusion, the time spent on a journey is also
an illusion.

My memories never fade. Memories from my supposed past

shine as clearly as my present. I remember climbing out of my
crib after a nap at 1 ½ years old as clearly as I remember
getting out of bed this morning. Aren’t memories supposed to
become less clear with time? These moments remain in my
head as individual events. I rarely think of them in conjunction
with moments that preceded or followed them. The memories
in my head feel somewhat like a piece of sedimentary rock—as
if these moments have all been compressed together and the
connector pieces—the time that I thought held them together
—has been blown away with the wind. These thoughts all exist
simultaneously in my mind yet they reveal themselves to me
one by one.

I think most important was my prevailing feeling of a

stronger connection between moments perceived as being
separated by time than between moments believed to be
connected by time. What I am unclear about, however, is what
causes this feeling of connection. Can there be a relationship
between these moments? Not in the sense of a linear
connection, but rather a feeling of empathy between them. To
a certain extent, I think there is a subconscious awareness that
there are these other moments occurring simultaneously and
that there can be an acknowledgement between moments that
are connected by subject matter.

If all moments are simultaneous, I am concurrently hearing

the Christopher Robin poem being read, watching the Prince
and Princess of Wales on the balcony, and standing in front of
the Palace myself. My conscious mind feeds them to me in a
linear sequence strung out with a bunch of other moments in
an illusion of a continuous ow of action. While I am being
read to, however, my subconscious is aware that I really am in
front of Buckingham Palace and so a sense of really being
there is brought to the Christopher Robin reading or to the
Royal Wedding viewing.
 This awareness that this other moment is occurring out
there right now has struck me at many times. Sometimes it’s
when I’m reading a book, other times I’m walking down the
street listening to music. Always, however, there is the feeling
that I am somewhat connected to that other moment and I can
almost feel there is the chance of stepping out of this moment
and into another. It is the knowledge that there is another
possibility to this moment.

To a certain extent, I often feel as if we are moving towards

a timeless existence. The increasing usage of the computer by
people on an everyday basis is one factor heading us in this
direction. At any moment, without any thought to time, we
can shop on our computers, chat, read newspapers, research,
do our banking, etc. Also, more and more we are creating
environments in which timelessness is the objectivity.
Nowhere is this more obvious than in the twentieth-century
environments of the department store, the amusement park
and the casino. The goal is one dream-like moment, where
there is no beginning and no end—no time.

Reading these comments again three months after they came, they
strike me as often very close to my position. Incidentally, I address
the original Gretchen’s questions (Glaubst du an Gott? Wie halt’s du es
mit der Religion?) in the Epilogue.
 Note for physicists (p. 18): Space plays two roles in Newtonian
physics: it binds its contents together to form the plurality within
the unity mentioned in this section (the separations between N
objects in Euclidean space are constrained by both inequalities and
algebraic relations, which give expression to this unity) and if
de nes positions at non-coincident times. In the type of physics I am
advocating, only the rst property is used, as will become clear in
Part 3.

In relativity theory, the construction of ‘three-dimensional’

snapshots from two-dimensional photographs is greatly (but not
insuperably) complicated by the fact that light travels at nite
speed, so that objects are no longer where they seem to be. Readers
familiar with relativity theory and concerned that my concept of a
Now seems very non-relativistic are asked to defer judgment until
Part 3. Einstein did not abolish Nows, he simply made them relative.

Laws and Initial Conditions (p. 22) Although Newton’s and

Einstein’s laws work equally well in both time directions, there is
one known phenomenon in quantum physics that seems to
determine a direction of time at a truly fundamental level. It is
observed in the decay of particles called kaons. Paul Davies
discusses this phenomenon in some detail in his About Time. Most
authors are agreed that this phenomenon does not seem capable of
explaining the pronounced directionality of temporal processes,
which is one of my main concerns in this book, but it is probably
very important in other respects and may provide evidence that
time really does exist as an autonomous governing factor in the
universe. However, the evidence that it de nes a direction in time is
indirect, being based on something called the TCP theorem.
Although this is most important in modern physics, what form if any
it will take in the as yet non-existent theory of quantum gravity is
not at all clear.

CHAPTER 2: TIME CAPSULES

The Physical World and Consciousness (1) (p. 26) There is a clear
and detailed account of Boltzmann’s ideas in Huw Price’s book listed
in Further Reading.

(2) (p. 27) It is worth quoting here two passages from Boltzmann
himself. In 1895 he published (in perfect English—I wonder if he
had assistance) a paper in Nature with the title ‘On certain questions
of the theory of gases’. It ends with a truly remarkable and concise
statement of what much later became known as the anthropic
principle. This expression was coined in 1970 by the English
relativist Brandon Carter (who had earlier made important
discoveries about the physics of black holes in the period leading up
to Hawking’s discovery that they can evaporate). The anthropic
principle, which gained widespread attention initially through the
book The Anthropic Cosmological Principle by John Barrow and Frank
Tipler, expresses the idea that any universe in which intelligent life
exists must have special and unexpected (from a purely statistical
viewpoint) properties, since otherwise the intelligent life that
observes these properties could not exist. Therefore we should not
be surprised to nd ourselves in a universe that does have special
and remarkable properties.

In the following passage, the summits of the H curve to which

Boltzmann refers correspond to states with very low entropy and
high order. Note that Boltzmann credits his assistant with the idea.

1 will conclude this paper with an idea of my old assistant, Dr.

Schuetz.

We assume that the whole universe is, and rests for ever, in
thermal equilibrium. The probability that one (only one) part
of the universe is in a certain state, is the smaller the further
this state is from thermal equilibrium; but this probability is
greater, the greater the universe itself. If we assume the
universe great enough we can make the probability of one
relatively small part being in any given state (however far
from the state of thermal equilibrium), as great as we please.
We can also make the probability great that, though the whole
universe is in thermal equilibrium, our world is in its present
state. It may be sayd [sic] that the world is so far from thermal
equilibrium that we cannot imagine the improbability of such
a state. But can we imagine, on the other side, how small a
part of the whole universe this world is? Assuming the
 universe great enough, the probability that such a small part of
it as our world should be in its present state, is no longer
small.

If this assumption were correct, our world would return

more and more to thermal equilibrium; but because the whole
universe is so great, it might be probable that at some future
time some other world might deviate as far from thermal
equilibrium as our world does at present. Then the
aforementioned H curve would form a representation of what
takes place in the universe. The summits of the curve would
represent the worlds where visible motion and life exist.

Boltzmann returned to this theme a year later, this time writing

in German. The following is my translation:

One has a choice between two pictures. One can suppose

that the complete universe is currently in a most unlikely state.
However, one can also suppose that the eons during which
improbable states occur are relatively short compared with all
time, and the distance to Sirius is small compared with the
scale of the universe. Then in the universe, which otherwise is
everywhere in thermal equilibrium, i.e. is dead, one can nd,
here and there, relatively small regions on the scale of our
stellar region (let us call them isolated worlds) that during the
relatively short eons are far from equilibrium. What is more,
there will be as many of these in which the probability of the
 state is increasing as decreasing. Thus, for the universe the two
directions of time are indistinguishable, just as in space there
is no up or down. But just as we, at a certain point on the
surface of the Earth, regard the direction to the centre of the
Earth as down, a living creature that at a certain time is
present in one of these isolated worlds will regard the
direction of time towards the more improbable state as
di erent from the opposite direction (calling the former the
past, or beginning, and the latter the future, or end).
Therefore, in these small regions that become isolated from the
universe the ‘beginning’ will always be in an improbable state.

Time Without Time (p. 29) In connection with my suggestion that

the brain may be deceiving us when we see motion, it is interesting
to note that, as Steven Pinker points out in his How the Mind Works,
people with speci c types of brain damage see no motion when
normal people do see motion. In his words, they ‘can see objects
change their positions but cannot see them move—a syndrome that
a philosopher once tried to convince me was logically impossible!
The stream from a teapot does not ow but looks like an icicle; the
cup does not gradually ll with tea but is empty and then suddenly
full’.

If the mind can do these things, it may be creating the

impression of motion in undamaged brains.

CHAPTER 3: A TIMELESS WORLD

First Outline (p. 36) The philosopher best known for questioning
the existence of time and its ow was John McTaggart, who is often
quoted for his espousal of the ‘unreality’ of time and the denial of
transience. The following argument of his is very characteristic of
professional philosophers:

Past, present, and future are incompatible determinations.

Every event must be one or the other, but no event can be
more than one. If I say that any event is past, that implies that
it is neither present nor future, and so with the others. And
this exclusiveness is essential to change, and therefore to time.
For the only change we can get is from future to present, and
from present to past.

The characteristics, therefore, are incompatible. But every

event has them all. if [an event] is past, it has been present
and future. If it is future, it will be present and past. If it is
present, it has been future and will be past. Thus all the three
characteristics belong to each event. How is this consistent
with their being incompatible? (McTaggart 1927, Vol. 2, p. 20)

Some thoughts here certainly match my own thinking, especially

that ‘exclusiveness is essential to change’, but McTaggart’s
arguments are purely logical and make no appeal to physics. Abner
Shimony (1997)—to whom I am indebted for several discussions—
compares McTaggart’s position with mine, but I think he has not
quite understood my notion of time capsules, so I do not feel that
his arguments force me to accept transience.

A typical example of theological thought about time is this

extract from Conversations with God—An Uncommon Dialogue by
Neale Donald Walsch (kindly sent me by Ann Gill):

Think of [time] as a spindle, representing the Eternal

Moment of Now.

Now picture leafs [sic] of paper on the spindle, one atop

the other. These are the elements of time. Each element
separate and distinct, yet each existing simultaneously with the
other. All the paper on the spindle at once! As much as there
will ever be—as much as there ever was . . .

There is only One Moment—this moment—the Eternal

Moment of Now (p-29).

Again, there is some overlap with my position. Walsch’s ‘leafs’, his

elements of time, are my Nows. But the spindle of time, the Eternal
Moment, is not at all part of my picture. My Nows are all
constructed according to the same rule. There is no Eternal Moment,
only the common rule of construction. I think Walsch is trying to
grasp eternal substance where there is none, though I think he is
right to say that the ‘leafs’ are all there at once and that this is a
consoling thought. But we should not ask for more than we can get.
Also, the image of time as a spindle is beautiful but misleading. In
my view, the ‘leafs’ of time most de nitely cannot be arranged along
a single line, as the striking spindle image implies.

The Ultimate Arena (1) (p. 39) In this section I say that all
structures that represent possible instants of time are three-
dimensional. This is because the space we actually observe has three
dimensions. However, in some modern theories (super-string
theories) it is assumed that space actually has ten or even more
dimensions. All but three of the dimensions are ‘rolled up’ so tightly
that we cannot see them. In principle, my instants of time could t
into this picture. They would then have ten (or more) dimensions.

(2) This note is for experts. Platonia is a special type of

con guration space known as a strati ed manifold. The sheets, ribs
and singular point that form the frontiers of Triangle Land are called
strata. I believe that the strati ed structure of Platonia is highly
signi cant. Mathematicians and physicists really interested in this
can consult DeWitt (1970) and Fischer (1970). The strata are
generally regarded as something of a nuisance, since at them normal
well-behaved mathematics breaks down. They are like grit in the
works. But in the world’s oyster they may be the grit from which
grows ‘a peal richer than all his tribe’: not Desdemona, but time
(Chapter 22). (After Othello had strangled Desdemona and then
realized his dreadful mistake, he said before stabbing himself that
he was ‘one whose hand, Like the base Indian, threw away a pearl
richer than all his tribe’.)
CHAPTER 4: ALTERNATIVE FRAMEWORKS

(1) (p. 61) I have written at considerable length about the early
history of astronomy and mechanics and the absolute versus relative
debate in my Absolute or Relative Motion? This has recently been
reprinted as a paperback with the new title The Discovery of
Dynamics (OUP, 2001). I still hope to complete a further volume
bringing the story up to the present, and much has already been
written, but my plans are in ux because of the developments
mentioned at the end of the Preface and at various places in these
notes. Readers wanting a full academic (and mathematical)
treatment of the topics presented in Parts 2 and 3 of this book are
asked to consult the above and the papers (Barbour 1994a, 1999,
2000, 2001), which cite earlier papers. For references to recent
developments see p. 358 and my website (www.julianbarbour.com).

(2) (p. 64) In the main body of the text, I mention the importance of
the fortunate circumstances of the world in enabling physicists to
avoid worrying about foundations. Another very important factor is
the clarity of the notion of empty space, developed so early by the
Greek mathematicians, which deeply impressed Newton. He felt that
he really could see space in his mind’s eye, and regarded it as being
rather like some in nite translucent block of glass. He and many
other mathematicians pictured its points as being like tiny identical
grains of sand that, close-packed, make up the block. But this is all
rather ghostly and mysterious. Unlike glass and tiny grains of sand,
which are just visible, space and its points are utterly invisible. This
is a suspect, unreal world.

We are not bound to hang onto old notions. We can open our
eyes to something new. Let me try to persuade you that points of
space are not what mathematicians would sometimes have us
believe. Imagine yourself in a magni cent mountain range, and that
someone asked, ‘Where are you?’ Would you kneel down with a
magnifying glass and look for that invisible ‘point’ at which you
happen to be in the ‘space’ that the mountain range occupies? You
would look in vain. Indeed, you would never do such a silly thing.
You would just look around you at the mountains. They tell you
where you are. The point you occupy in the world is de ned by what
the world looks like as seen by you: it is a snapshot of the world as
seen by you. Real points of space are not tiny grains of sand, they
are actual pictures. To see the point where you are in the world, you
must look not inward but outward.

The plaque near the grave of Christopher Wren in St. Paul’s

Cathedral says simply: ‘If you seek a monument, look around you.’
The point where you are is a monument too, and you see it by
looking around you. It is this sort of change of mindset that I think
we need if we are to understand the universe and time.

To conclude this note, a word about what is perhaps the most

serious problem in my approach. It is how to deal with in nity. As
so far de ned, each place in Platonia corresponds to a con guration
of a nite number of objects. Such a universe is like an island of
 nite extent. One could allow the con gurations to have in nite
extent and contain in nitely many objects. That is not an
insuperable problem. The di culty arises with the operations that
one needs to perform. As presented in this book, the operations
work only if the points in Platonia, the instants of time, are in some
sense nite. There may be ways around this problem—Einstein’s
theory can deal beautifully with either nite or in nite universes—
but in nity is always rather di cult. There is something ‘beyond
the horizon’, and we can never close the circle of cause and e ect.
In short, we cannot build a model of a completely rational world.
Precisely for this reason Einstein’s rst and most famous
cosmological model was spatially nite, closed up on itself. The
constructions of this book are to be seen as a similar attempt to
create a rational model of the universe in which the elusive circle
does close.

In fact, if the work with Niall O Murchadha mentioned at the

end of the Preface, which suggests that absolute distance can be
eliminated as a basic concept (see Box 3), can be transformed into a
complete theory, the problem of in nity may well be solved in the
process. If size has no meaning, the distinction between a spatially
nite or in nite universe becomes meaningless.

CHAPTER 5: NEWTON’S EVIDENCE

The Aims of Machian Mechanics (1) (p. 71) In creating the

beautiful diagrams that form such an important part of this section,
Dierck Liebscher was able to draw on initial data devised by
Douglas Heggie (University of Edinburgh), using software written by
Piet Hut (Institute for Advanced Study), Steve McMillan (Drexel
University) and Jun Makino (University of Tokyo). Dierck has
written a very interesting book (alas, as yet published only in
German) on the connection between di erent possible geometries
and Einstein’s relativity theory (Liebscher 1999). It contains many
striking computer-generated diagrams.

(2) Poincaré’s discussion is contained in his Science and Hypothesis,

which, along with the writings of his contemporary, Mach, became
a popular-science best-seller. In fact, in this book I am actually
revisiting many of the themes discussed by Poincaré and Mach, but
with the advantage of hindsight. How are the great issues they
raised changed by the discovery of general relativity and quantum
mechanics? I have adapted Poincaré’s discussion somewhat to match
the requirements of a timeless theory (he considered only the
possibility of eliminating absolute space).

(3) Since writing Box 3, which draws attention to the present

unsatisfactory use of absolute dislance in physics, I have discovered
a way to create dynamical theories in which distance is not
absolute. This is achieved by a very natural extension of the best-
matching idea described later in the book. The new insights that I
mention in the Preface are in part connected with this development.
One of the most exciting is that, if such theories do indeed describe
the world, gravitation and the other forces of nature are precisely
the mechanism by means of which absolute distance is made
irrelevant. Since this work is still in progress, I shall make no
attempt to describe it in detail, but I shall keep my website
(www.julianbarbour.com) up to date with any progress (see also
p.358).

CHAPTER 6: THE TWO GREAT CLOCKS IN THE SKY

The Inertial Clock (p. 99) Tait’s work, which I feel is very
important, passed almost completely unnoticed. This is probably
because two years later the young German Ludwig Lange introduced
an alternative construction for nding inertial frames of reference,
coining the expression ‘inertial system’. Lange deserves great credit
for bringing to the fore the issue of the determination of such
systems from purely relative data, but Tait’s construction is far more
illuminating. Lange’s work is discussed in detail in Barbour (1989)
and Tait’s in Barbour (forthcoming).

The Second Great Clock (p. 107) A very nice account of the history
of the introduction of ephemeris time was given by the American
astronomer Gerald Clemence (1957).

CHAPTER 7: PATHS IN PLATONIA

Nature and Exploration (p. 109) For physicists and
mathematicians who do not know the book, a wonderful account of
the variational principles of mechanics, together with much
historical material, is given by Lanczos (1986).

Developing Machian Ideas (p. 115) Translations of the papers by

Hofrnann, Reissner and Schrödinger, along with other historical and
technical papers on Mach’s principle, can be found in Barbour and
P ster (1995).

Exploring Platonia (p. 115) The special properties of Newtonian

motions with vanishing angular momentum were discovered
independently of the work of Bertotti and myself by A. Guichardet
in the theory of molecular motions and by A. Shapere and E Wilczek
in the theory of how micro-organisms swim in viscous uids! A rich
mathematical theory has meanwhile developed, and is excellently
reviewed in the article by Littlejohn and Reinsch (1997), which
contains references to the original work mentioned above. All
mathematical details, as well as references to the earlier work by
Bertotti and myself, can be found in Barbour (1994a).

CHAPTER 8: THE BOLT FROM THE BLUE

Historical accidents (p. 123) Poincaré’s paper can be found in his

The Value of Science, Chapter 2. Pais’s book is in the Bibliography.
Background to the Crisis (p. 124) The best (moderately technical)
historical background to the relativity revolution that I know of is
the book by Max Born. It is available in paperback.

The Forgotten Aspects of Time (p. 135) My claims about the

topics that somehow escaped Einstein’s attention are spelled out in
detail in Barbour (1999, forthcoming). I have tried to make good the
gap in the literature on the theory of clocks and duration in Barbour
(1994a).

CHAPTER 10: THE DISCOVERY OF GENERAL

RELATIVITY

Einstein’s Way to General Relativity (p. 151) Einstein’s papers

and correspondence are currently being published (with translations
into English) by Princeton University Press. The letter to his wife
mentioned in this section can be found in the rst volume of
correspondence (Stachel et al. 1987).

CHAPTER 11: GENERAL RELATIVITY: THE TIMELESS

PICTURE

Platonia for Relativity (p. 167) This is a technical note about the
de nition of superspace. The equations of general relativity lead to
a great variety of di erent kinds of solution, including ones in
which there are so-called closed time-like loops. These are solutions
in which a kind of time travel seems to be possible. The question
then arises of whether a given solution of general relativity—that is,
a space-time that satis es Einstein’s equations—can be represented
as a path in superspace, in technical terms, as a unique succession of
Riemannian three-geometries. If this is always so, then superspace
does indeed seem a natural and appropriate concept. Unfortunately,
it is de nitely not so. There are two ways in which we can attempt
to get round this di culty. We could say that classical general
relativity is not the fundamental theory of the universe, since it is
not a quantum theory. This allows us to argue that superspace is the
appropriate quantum concept and that it will allow only certain
‘well-behaved’ solutions of general relativity to emerge as
approximate classical histories. For these, superspace will be an
appropriate concept. Alternatively, we could extend the de nition of
super-space to include not only proper Riemannian 3-geometries (in
which the geometry in small regions is always Euclidean), but also
pscudo-Riemannian 3-geometries (in which the local geometry has a
Minkowski type signature), and also geometries in which the
signature changes within the space. For the reasons given in the
long note starting on p. 348 below, I prefer the second option.

The above note was written before my new insights mentioned at the
end of the Preface. I now believe that there is a potentially much more
attractive resolution of the di culty: the true arena of the world is not
superspace but conformal superspace, which I describe on p. 350.
Catching Up with Einstein (1) (p. 175) Figure 30 is modelled
directly on well-known diagrams in Wheeler (1964) and Misner et
al. (1973).

(2) Technical note: Einstein’s eld equations relate a four-

dimensional tensor formed from geometrical quantities to the four-
dimensional energy-momentum tensor, which is formed from the
variables that describe the matter. Machian geometrodynamics
shows how these four-dimensional tensors are built up from three-
dimensional quantities. The two principles by which this is done are
best matching, and Minkowski’s rule that the space and time
directions must be treated in exactly the same way (see the
following note). As far as I know, the mathematics of how this is
done when matter is present was rst spelled out in a recent paper
by Domenico Giulini (1999), to whom I am indebted for numerous
discussions on this and many other topics covered in this book.

A Summary and the Dilemma (1) (p. 177) This is another

technical note. My image of space-time as a tapestry of interwoven
lovers rests on the following property of Einstein’s eld equations.
If, in any given space-time that is a solution of the eld equations,
we lay out an arbitrary four-dimensional grid in any small region of
the space-time, we can then, in principle, attempt to take the data
on one three-dimensional hypersurface and use Einstein’s equations
to evolve these data and recover the space-time in the complete
region. Normally, we attempt to do this in a time-like direction.
However, the form of the equation is exactly the same whichever
direction in which we choose to attempt the evolution from initial
data. This is an immediate consequence of an aspect of the relativity
principle that Minkowski gave a special emphasis: as regards the
structure of the equations, whatever holds for space holds for time
and vice versa.

What is more, however we choose the ‘direction of attempted

evolution’, Einstein’s equations always have a very characteristic
structure. There are ten equations in all. One of them does not
contain any derivative with respect to the variable in which we are
going to attempt the evolution. Three of them contain only rst
derivatives with respect to that variable. The remaining six
equations contain second derivatives with respect to it and have the
form of equations that are suitable for evolution in the chosen
direction. But we must rst solve the other four equations, which
are so-called constraints. Unless the initial data satisfy these four
equations, evolution is impossible.

There are two ways to look at a space-time that satis es

Einstein’s equations: either as a structure obtained from initial data
that have been (somehow) obtained in a form that satis es the
constraints and then built by the more or less conventional
evolution equations, or as a structure that satis es everywhere the
constraints however we choose to draw the coordinate lines. In the
second way of looking at space-time, conventional evolution does
not come into the picture at all. Much suggests that this is the more
fundamental way of looking at Einstein’s equations (see, in
particular, Kuchaf’s beautiful 1992 paper).The connection with my
timeless way of thinking about general relativity is expressed by the
fact that the three constraint equations containing only rst
derivatives of the evolution variable are precisely the expression of
the fact that a best-matching condition holds along the
corresponding ‘initial’ hypersurface, while the fourth constraint
equation, containing no derivatives of the evolution variable,
expresses the fact that proper time is determined in
geometrodynamics as a local analogue of the astronomers’
ephemeris time. It is this complete freedom to draw coordinate lines
as we wish and, at least formally, to attempt evolution in any
direction, that makes me feel that the second alternative envisaged
in the Platonia for Relativity note is appropriate. I think it is also
very signi cant that Einstein’s equations have the same form
whatever the signature of space-time. The signature is not part of
the equations, it is a condition normally imposed on the solutions.
The demonstration that Einstein’s general relativity is the unique
theory that satis es the criterion (mentioned at the end of this
section) of a higher four-dimensional symmetry was given by
Hojman et al. (1976).

1 mentioned on p. 346 at the end of the notes on Chapter 4 my

recent discovery of a way to create dynamical theories of the
universe in which absolute distance is no longer relevant. My Irish
colleague Niall Ó Murchadha, of University College Cork, and I are
currently working on the application of the new idea to theories like
general relativity, in which geometry is dynamical. There is a
possibility that this work will not only give new insight into the
structure of general relativity, in which a kind of residual absolute
distance does play a role, but also lead to a rival alternative theory
in which no distance of any kind occurs.

The key step is to extend the principle of best matching from

superspace to so-called conformal superspace. In the context of
geometrodynamics, this is analogous to the passage from Triangle
Land to Shape Space as described in Box 3. However, whereas in
Box 3 it is only the overall scale that is removed, and it is still
meaningful to talk about the ratios of lengths of sides, the transition
to conformal superspace is much more drastic and removes from
physics all trace of distance comparison at spatially separated
points.

In more technical terms, for people in the know, each point of

conformal superspace has a given conformal geometry and is
represented by the equivalence class of metrics related by position-
dependent scale transformations.

The potentially most interesting implication of this work is that

it could resolve the severe problem of the criss-cross fabric of space-
time illustrated by Figure 31. At the level of conformal superspace,
the universe passes through a unique sequence of states. For latest
developments, please consult my website (www.julianbarbour.com)
and the nal entries in these notes and the notes on p.358.
CHAPTER 12: THE DISCOVERY OF QUANTUM
MECHANICS

(p. 191) On the connection between particles and elds, let me

mention here that I assume the appropriate ‘Platonic’ representation
at the level of quantum eld theory to be in terms of the states of
elds, not particles.

CHAPTER 13: THE LESSER MYSTERIES

(p. 202) Wheeler and Zurek (1983) have published an excellent

collection of original papers on the interpretational problems of
quantum mechanics.

CHAPTER 14: THE GREATER MYSTERIES

The Many-Worlds Interpretation (p. 221) Everett’s original Ph.D.

thesis, his published paper and the papers of DeWitt (and some
other people) relating to the many-worlds idea can be found in the
book by DeWitt and Graham (1973).

CHAPTER 16: ‘THAT DAMNED EQUATION’

History and Quantum Cosmology (p. 240) More details on the

Leibnizian idea that the actual universe is more varied than any
other conceivable universe are given in Smolin (1991), Barbour and
Smolin (1992), and Barbour (1994b). The quotation from T. S. Eliot
is in Eliot (1964). My book is Barbour (1989).

‘That Damned Equation’ (p. 247) Technical note: In connection

with Chapter 11, it is interesting that the form of the Wheeler-
DeWitt equation is independent of the signature of space-time.

(1) (p. 247) For physicists I should mention that there is an

important alternative to regarding the Wheeler-DeWitt equation as
analogous to the stationary Schrödinger equation. It also bears a
resemblance to the relativistic Klein-Gordon equation, the role of
time in that equation being played, essentially, by the volume of the
universe in the case of the Wheeler-DeWitt equation.

(2) (p. 247) Kuchařs objections to my timeless interpretation of the

Wheeler-DeWitt equation can be found in the discussion sessions at
the end of Barbour and P ster (1995). Comprehensive reviews of
the problems of time in quantum gravity can be found in Kuchaf
(1992) and Isham (1993).

(3) (p. 247) In discussions with me in 1994 at an international

conference on quantum gravity held at Durham, Bryce DeWitt
expressed two main reservations about his ‘damned equation’. The
rst was that it required a division of space-time into space and
time, which he felt was running counter to the great tradition of
relativity initiated by Einstein and Minkowski. I have already
explained why I feel that this may not necessarily be an objection;
indeed, it may not be possible to give objective content to general
relativity unless such a split is made. DeWitt’s second objection was
that the ‘damned equation’ had not as yet yielded any concrete
results and was (is) plagued with mathematical di culties. This is
certainly true, and I have omitted all discussion of these di culties,
which are certainly great. However, I think it is worth noting that as
physicists’ understanding of the equations that describe nature
becomes deeper, the equations themselves become more
sophisticated and harder to solve. It is much harder to nd solutions
of Einstein’s equations than Newton’s. This tendency—deeper
understanding of principles bringing with it greater intractability of
equations—will almost certainly mean that progress in quantum
gravity is very slow. In fact, for over a decade, a group centred on
the relativist Abhay Ashtekar, including my friend Lee Smolin and
another friend Carlo Rovelli, has been working intensively on a
particular approach to canonical quantum gravity (the broad
framework in which DeWitt derived his equation) and have
certainly resolved some of the di culties. An account of this work
can be found in Lee’s The Life of the Cosmos and Three Roads to
Quantum Gravity. Kuchař too has made many important
contributions.

If the ideas described in the note on p. 350 work out as Niall Ó

Murchadha and I believe they could, the di cult issues raised in the
nal part of Chapter 16 and in the above notes will be to a very
large degree resolved. The conceptual uncertainties about the
correct way to proceed that have plagued the theory for four
decades could all be removed. Both for general relativity and the
alternative theory that might replace it, the wave function of the
universe will certainly be static and give probabilities for
con gurations as explained in the main text. The main di erence is
that only the intrinsic structure will count, so that all con gurations
that have the same structure and di er only in the local scales will
have the same probability. They will merely be di erent
representations of the same instants of time. However, for general
relativity there will be a curious residual scale that represents a
volume of the universe. It will be meaningful to say that the
universe has a volume but not how the volume is distributed
between the intrinsic structures contained within in.

CHAPTER 17: THE PHILOSOPHY OF TIMELESSNESS

(p. 255) On the subject of the aims and methods of science, I

strongly recommend David Deutsch’s The Fabric of Reality.

CHAPTER 18: STATIC DYNAMICS AND TIME CAPSULES

Dynamics Without Dynamics (p. 258) In this section I refer to

investigations by various authors. Their studies will be found in the
bibliography. Physicists really interested in the semiclassical
approach may also like to consult the review article by Vilenkin
(1989), the paper by Brout (1987), the nal part of Zeh (1992,
1999) and the introductory article by Kiefer (1997). The fullest
account of my own ideas is Barbour (1994a).

CHAPTER 19: LATENT HISTORIES AND WAVE POCKETS

Schrödinger’s Heroic Failure (p. 278) In the rst draft of this book
I included a long section on the very interesting interpretation of
quantum mechanics advanced originally by de Broglie, and revived
by Bohm, whose 1952 paper I strongly recommend to physicists
together with Peter Holland’s book (Holland 1993). With regret I
omitted it, as I felt that it made this book too long, especially since I
believe that the interpretation does not really solve the problem.
However, I particularly value the way in which it shows that all the
results of quantum mechanics can be obtained in a framework in
which positions are taken as basic. This made the theory attractive
to John Bell, as we shall see in the next chapters.

CHAPTER 20: THE CREATION OF RECORDS

The Creation of Records: First Mechanism (1) (p. 284) Bell’s

paper can be found in his collected publications Speakable and
Unspeakable in Quantum Mechanics.

(2) (p. 284) Mott’s paper is reproduced in Wheeler and Zurek

(1983). Heisenberg’s treatment is in his Physical Principles of
Quantum Theory. I am very grateful to Jim Hartle, who rst drew my
attention to Mott’s paper. At that time he was considering seriously
an interpretation of quantum cosmology that is quite close to my
own present position. He has since backed away somewhat, and
now advocates an interpretation of quantum mechanics in which
history is the fundamental concept. I should also like to express my
thanks here to Dieter Zeh. Zeh, who was in this business long before
me, also made me realize the importance of Mott’s work, and,
crucially, alerted me to Bell’s paper. There are not many physicists
who take the challenge of timelessness utterly seriously, but Dieter
Zeh and his student Claus Kiefer, from both of whom I have gained
and learned much, are two of them.

CHAPTER 21: THE MANY-INSTANTS INTERPRETATION

Bell’s ‘Many-Worlds’ Interpretation (p. 299) In his ‘cosmological

interpretation’ of quantum mechanics, Bell combined elements
derived from both Everett and the de Broglie-Bohm interpretation
(see the note to Chapter 19). In fact, Bell’s account of his mixed
interpretation is rather terse, and can be misunderstood. I am most
grateful to Fay Dowker and Harvey Brown for drawing my attention
to an error I made in reporting Bell’s idea in my rst draft of this
book. In this section I follow their interpretation of Bell, which I am
sure is what he did mean.
The Many-Instants Interpretation (p. 302) I hope I have made it
clear that probability ‘to be experienced’ or ‘to exist’ is a
problematic concept. If consciousness is determined by structure,
the consciousness is already in the Nows and must be experienced
irrespective of their probabilities. What role remains for
probabilities? It is a very di cult issue. Probability is already
puzzling in ordinary quantum mechanics, and even in classical
physics. Cold water could boil spontaneously, but we never see this
happen. Standard probability arguments suggest that what is
possible but hugely improbable will not be experienced. Much
suggests that probabilities in some form are inescapable in quantum
theory simply because it explores mutually exclusive possibilities.
Instants of time are natural candidates for the ultimate exclusive
possibilities. If certain very specially structured instants do get
hugely larger probabilities than others, and are the ones habitually
experienced, that must, I feel, count as explanation. But as an
indication of the depth of this problem, I add here in Box 16 an
edited email exchange I had with Fay Dowker of Imperial College,
London. I had especially asked her to read my rst draft, since she is
a very clear thinker but is sceptical about both many worlds and
canonical quantization, the approach to quantum gravity that I
favour.

CHAPTER 22: THE EMERGENCE OF TIME AND ITS

ARROW
Soccer in the Matterhorn (p. 307) In the second edition of The
Physical Basis of the Direction of Tune, Zeh says that the intrinsic
dynamical asymmetry of quantum gravity ‘o ers the possibility of
deriving an arrow of time (perhaps even without imposing any
special conditions)’.

Timeless Descriptions of Dynamics (1) (p. 309) For specialists: for

each stage of a perturbation expansion, Mott always chooses a
kernel in an integral representation corresponding to outgoing
waves. However, nothing in the mathematics rules out the
(occasional) choice of incoming waves. This would mess up
everything.

(2) (p. 309) I should emphasize that Mott, like Bell, never used any
expression like ‘time capsule’, and clearly did not think in such
terms about alpha-particle tracks. Neither did Mott’s work on alpha-
particle tracks seem to have prompted him to any intimations of a
many-worlds type interpretation of quantum mechanics. I learned
this from Jim Hartle. Over a decade ago, when collaborating with
Stephen Hawking in Cambridge, Jim lodged at his college, Gronville
and Caius (featured famously in Chariots of Fire), which was also
Mott’s college. Over dinner Jim asked Mott whether his paper had
not led him to anticipate some form of Everett’s idea, and was told
no. Apparently, all the ‘young Turks’ followed the Copenhagen line
without hesitation at that time. Shortly before his death about two
years ago, when he was still mentally very alert, I contact Mott and
asked if I could talk to him about his paper. Alas, he was too ill to
keep the appointment, telling his secretary he was very disappointed
‘since the man wanted to talk about work I did nearly sixty years
ago’.

A Well-Ordered Cosmos? (p. 321) This nal section follows closely

the nal section of Barbour (1994a).

BOX 16 An Email Dialogue

DOWKER. It seems to me that you provide no scheme for making

predictions, and I would further claim that no such scheme can exist
which contains the two aspects that are fundamental to your
scheme: canonical quantum gravity (CQG) and the Bell version of
the many-worlds interpretation (MWI).

BARBOUR. I think you are right, subject to what one means by

prediction. I cannot make the kinds of prediction you want, and you
correctly identify the reasons. I feel the arguments for CQG and
MWI outweigh desire for predictions of the kind you would like.

DOWKER. I freely admit that I am rather attached to the notion of

the universe (and I) having had, and being about to have, a
continuous history. But my criticism here is not the absence of
history in your approach, but, to repeat, that there is no way to
make predictions about the results of our observations. In my view
this is a de ciency that cannot be overcome. Whatever else science
tells us about the world, it must allow us to make predictions about
our observations that we can check.

BARBOUR. I am not sure we can impose such a criterion on Nature.

The Greeks had the notion of saving appearances ( nding a rational
explanation for the phenomena we observe) that is already very
valuable. You may be asking more of Nature than she is prepared to
give.

DOWKER. In backing up this criticism I shall focus on the aspect

that I am most familiar with and on which I have most con dence
in my own views, which is the aspect of the interpretation of
quantum mechanics. I am pleased that your book draws attention to
the work of Bell on the many-worlds interpretation, since it has not
had the recognition it deserves. In my view his version is the only
well-de ned many-worlds interpretation (I’ll call it BMWI) that
exists in the literature.

BARBOUR. I agree it is well de ned, but with reservations about

the role of time. The time of an observation, like any other
observable, must be extracted from present records. When you start
to ask how that is done in practice and how Nature does put time
into the records, I think things may become less well de ned. I do
believe that almost all physicists this century have blindly followed
Einstein in declining to try to understand duration at a fundamental
level. A lot of the rst part of my book is about that. I think my
position might be stronger than you realize.

DOWKER. I think that neither your version (which I’ll call JMWI)
nor BMWI allows us to make predictions about what we observe (so
I disagree with Everett’s statement ‘the theory itself predicts that our
experience will be what it in fact is’). Let me take your version.
There we have many con gurations at time t. The most serious
problem is that in a scheme like yours, in which all the possibilities
are realized, there is no role for the probabilities. The usual
probabilistic Copenhagen predictions for the results of our
observations cannot be recovered. An excellent reference which
analyses the MWI literature and the various attempts to derive the
Born interpretation from MWI is Adrian Kent [1990, International
Journal of Modern Physics, A5, 1745]. Adrian concludes that they
fail. I’ll just state again the main reason that they fail: when all the
elements in a sample space of possibilities are realized, then
probability is not involved. Your idea is that it is the sample space
itself, i.e. how many copies of each con guration are included in the
sample space, which is determined by the (squares of the)
coe cients of the terms in the wave function. That is all well and
good (if bizarre). But there’s no reason then to call those numbers
probabilities, and no way to recover the probabilistic predictions of
Copenhagen quantum mechanics. In fact the MWI proponents
themselves agree that the failure to reproduce the Copenhagen
predictions is a problem and do try to address it, but without
success.

BARBOUR. I accept that this is a strong critique. I nevertheless feel

that my scheme does in principle have predictive strength. If you
could see Platonia and Born’s probability density concentrated
incredibly strongly on a tiny proportion of its points that all turn out
to be time capsules as I de ne them and Bell describes them, would
you not nd that impressive, and something like a rational
explanation for our experiences?

DOWKER. As well as the MWI, you base your conjecture of

timelessness on the technical result that when a canonical
quantization scheme is applied to general relativity, the wave
function cannot contain the time. My understanding of the state of
a airs in canonical quantum gravity is that, because of this, no one
knows how to make the kind of predictions we’d like to make:
explanations such as ‘What happens in the nal stages of black hole
collapse?’, ‘Why is the cosmological constant so very small?’, etc.

BARBOUR. I agree with your rst example (and do not think it is

too serious—there may be questions that it is just not sensible to
ask), but in principle my scheme could predict that virtually all time
capsules will appear to have been created in nearly classical
universes with a very small cosmological constant. After all, that is
what our present records indicate. If all probable con gurations
seem to contain records that indicate a small cosmological constant,
I am okay.

DOWKER. My reaction to the situation is that formulating general

relativity in a canonical way has been shown to be the wrong thing
to do—we did what we weren’t supposed to—divided up space-time
into space and time again. Even if it wasn’t clear from the beginning
that it would be incredibly di cult to maintain general covariance
of the theory whilst trying to treat space and time di erently, I nd
the lack of any insight into how to recover predictions within the
canonical quantum gravity program convinces me that we should
look elsewhere for a quantum theory of gravity.

BARBOUR. As he was creating general relativity, Einstein was

convinced general covariance had deep physical signi cance. Two
years later, correctly in my opinion, he completely abandoned that
position. In my opinion, general covariance is an empty shell (I say
something about this at the end of Chapter 10 and in the notes to
it). I believe it is not possible to give any meaning to the objective
content of general relativity without saying how the three-
dimensional slices in space-time are related to each other. That is
the very content of the theory. That is why I think the arguments for
canonical quantum gravity are very strong indeed. The constraints
of the canonical theory are its complete content.
DOWKER. Having made my basic points, let me now just say that I
nd it incredibly hard to understand how, as a solipsist of the
moment, you must view science and the scienti c enterprise.

BARBOUR. Answered above I think. Science should explain what

we observe. We habitually observe and experience time capsules.
Even granting the real di culties with calling the square of a static
amplitude a probability, should it turn out that the Wheeler-DeWitt
equation does strongly concentrate the square of the amplitude on
time capsules, I think that would be an incredibly strong and
suggestive result.

DOWKER. Take the idea that a good scienti c theory should be

falsi able.

BARBOUR. I think my idea is falsi able in the following sense.

There may well be con gurations of the universe with records of my
idea and mathematical proofs that the Wheeler-DeWitt equation
most de nitely does not concentrate the square of the amplitude on
time capsules. If I too am in them, I would have to say my proposal
for an explanation of why we think time ows has failed.

DOWKER. That presupposes that there will be a future in which we

can try new experiments that test the theory and nd that these
experiments may be in contradiction to our predictions. The very
word ‘prediction’, which I have used so many times in this letter, is
laden with time-meaning. A prediction is a statement of expectation
of something that will happen. Prediction is the lifeblood of science.
How could we do science without it?

BARBOUR. I totally agree about the importance of prediction. But it

does not necessarily have to involve time in the way you suggest.
From observations of one side of the Moon, astronomers tried to
predict what was on the other side. They got it wrong when the
other side was seen. I do not think time comes into such predictions
at all signi cantly. Consider, as Jim Hartle once did when he was
quite close to my present position, geology. The rocks of the Earth
hardly change. Suppose the idea of continental drift had been
proposed before America had been discovered. It would have
predicted the existence of America and the geology of its east coast
(the west of Ireland exactly matches Newfoundland, I believe).
Again, time is not essentially involved in this prediction. I think Bell
puts my case very well: ‘We have no access to the past. We have
only our “memories” and “records”. But these memories and records
are in fact present phenomena.’ The italics are Bell’s. Predictions are
always veri ed in the present. That is my apologia.

Notes Added for This Printing.

As mentioned at the end of the Preface and at various places in

the Notes, there have been some promising developments of the
ideas presented in this book since it was sent to press in spring
1999. They are contained in two joint papers published
electronically and available on the web: Julian Barbour and Niall Ó
Murchadha, ‘Classical and quantum gravity on conformal
superspace’, http//xxx.lanl.gov/abs/gr-qc/9911071 and Julian
Barbour, Brendan Z. Foster, and Niall Ó Murchadha, ‘Relativity
without relativity’, http//xxx.lanl.gov/abs/gr-qc/0012089 [the xxx
is correct].

The potential signi cance of the rst paper has already been
explained on p. 349/350. At this stage, I do not wish to make any
rm statements about this new work since it is incomplete and has
not yet been exposed to scrutiny by other physicists, but I can at
least give some idea of what is at stake. The basic issue is the status
of the relativity principle. When Einstein and Minkowski created
special relativity, they deliberately made no attempt to explain the
remarkable structure that their work had brought to light: the
existence of spacetime and its associated light cone, both being
re ected in the Lorentz invariance of the laws of nature. They
adopted Lorentz invariance, which assumes the existence of length,
as the basis of physics. In small regions of spacetime, this still
remains true in general relativity.

Taken together the two papers cited above suggest that all of the
presently known facts of relativity and electromagnetism can be
derived in a new and hitherto unsuspected manner from three
assumptions: 1) an independent time plays no role in dynamics; 2)
best matching (pp. 116/7) is the essential element in the action
principle of the universe; 3) any theory satisfying these principles
must have nontrivial solutions. It is the third assumption that makes
a dramatic di erence. Hitherto, in common with other colleagues, I
had assumed (see p. 181) many di erent theories could satisfy the
rst two conditions, but, as my collaborator Niall Ó Murchadha
discovered, this is not the case. The reasons for this and its
remarkable potential consequences are spelled out in the second
cited paper. It is frustrating not to be able to say more at the present
moment, but at a time of uncertainty about the nal outcome it is
better to say less rather than too much. My website
(julianbarbour.com) will carry more detailed information.

I conclude with some additional comments that have mostly

already appeared in the UK paperback.

My website now carries an email dialogue between myself and

Don Page (whom Hawking, in A Brief History of Time, credits with
pointing out, with Raymond La amme, ‘his greatest blunder’). I
think Don has made some interesting and valuable comments,
including both valid criticism of some points but also reassurance on
other issues on which I had doubts. Don is another person who takes
the timelessness of physics utterly seriously, and, in fact, our views
are very close. He has written several papers on the problem of
consciousness and quantum mechanics (Sensible Quantum
Mechanics). Full details can be found on my website. I should also
like to mention here an idea that Dieter Zeh put forward in the
meeting at Huelva in Spain at which I conducted my straw poll. This
is that the universe must, of necessity, always be observed as
expanding. I nd this an intriguing idea and, if it is correct, it would
t beautifully with the open-ended, owerlike structure of Platonia.
I think Dieter’s idea, which I hope is correct though I have hedged
my bets in this book, in uenced me somewhat in the writing of Box
3 and parts of the Epilogue.

I regret that in the nal chapter I made no mention of the idea of

in ation, which is explained in a gripping and candid book by Alan
Guth that I have at least now included in the books recommended
for further reading. From reading Guth’s book I also learned that an
interesting proposal of a mechanism for the ‘creation of the
universe’ was made by Edward Tryon in 1973. I should also have
mentioned that in 1982 Alexander Vilenkin proposed an in uential
alternative proposal to Hawking’s no-boundary idea (1981) and that
Jim Hartle played a signi cant role in its development, which
culminated the Hartle-Hawking wave function. My apologies to
these authors (none of whom have registered any complaint).
Details can be found in Guth’s book.

It has also been pointed out to me by several email

correspondents that there is a clear anticipation of some of my ideas
about time in Fred Hoyle’s novel October the First Is Too Late, which
Paul Davies discusses in his About Time. Sir Fred’s ‘pigeonholes’ are
essentially my time capsules. American reviewers and
correspondents also noted a similarity with the philosophy of time
that underlies Kurt Vonnegut’s Slaughterhouse Five. Having now
read the book, I can con rm that this is the case. Of course, as I
make clear in the text, John Bell also formulated the idea of time
capsules (without giving them any name) quite clearly long before
me. Another correspondent, Andrew Clifton, regretted that I had not
devoted at least a few words to demolishing the idea that there
really is a ‘moving present’. I think he was right. Happily, David
Deutsch has done the job very well in his The Fabric of Reality.

I should also like to thank Damien Broderick, who has reviewed

my book for The Australian, for drawing my attention to various
misprints.

It is also now clear to me that in the body of the text I should

have said more about possible ways in which my ideas could be
refuted. A theory is no use to science unless it is capable of disproof.
In the email exchange with Fay Dowker, I did mention the
possibility of mathematical disproof of my conjecture that the
Wheeler—DeWitt equation concentrates its solutions on time
capsules. However, I think that (in normal parlance) that might take
decades. Something that might occur much sooner is a completely
convincing de nitive form of superstring theory (or some other
uni ed theory) that reintroduces an external time (string theory
does currently use background structures). That would kill my idea.
My own feeling is that in fact superstring theory will, if and when it
is found, turn out to be timeless.

Then there is one other quite di erent way in which my ideas

could be disproved. This is if experimental evidence can be found
that shows collapse of the quantum wave function to be a real
physical process. In this connection, I should like to mention
especially an experiment proposed by Roger Penrose to test this very
possibility. He is developing it in collaboration with Anton
Zeilinger, the Austrian physicist based in Vienna, who has
performed so many incredibly beautiful quantum experiments.
Penrose, very understandably, nds the many-worlds interpretation
of quantum mechanics extremely hard to accept (see my comments
about the death of Diana), and with great persistence is trying to
nd a way round it. He has certainly identi ed the greatest single
issue in modern physics. If his experiment, which could perhaps be
performed within a decade, works out in the way he hopes, it will
be a huge development and destroy my approach (because it will
show that quantum mechanics does not hold macroscopically). The
volume containing my paper (Barbour 2000) also contains Penrose’s
most important paper on the subject, and also a related paper by
Joy Christian, who was a student of Abner Shimony. Joy, following
Abner (see my Epilogue), is trying to establish transience as a real
physical thing. I think that if collapse of the wave function could be
demonstrated as a real physical phenomenon, that would be true
demonstration of something that one might call transience.

JB
January 2001
FURTHER READING

Barrow, John, 1992, Theories of Everything, Oxford University Press,

Oxford.

Barrow, John and Tipler, Frank, 1986, The Anthropic Cosmological

Anthropic Principle, Clarendon Press, Oxford.

Bondi, Hermann, 1962, Relativity and Common Sense: A New

Approach to Einstein, Dover, New York.

Coleman, James, 1954, Relativity for the Layman, William-Frederick

Press, New York (1990, Penguin, London).

Coveney, Peter and High eld, Roger, 1991, The Arrow of Time,
Flamingo, London.

Davies, Paul, 1991, The Mind of God, Simon & Schuster, New York.

Davies, Paul, 1995, About Time: Einstein’s Un nished Revolution,

Penguin Press, London.

Deutsch, David, 1997, The Fabric of Reality, Penguin Press, London.

Eddington, Arthur, 1920, Space, Time and Gravitation, Cambridge

University Press, Cambridge.

Einstein, Albert, 1960, Relativity: The Special and the General Theory.
A Popular Exposition, Routledge, London.
Greene, Brian, 1999, The Elegant Universe: Superstrings, Hidden
Dimensions, and the Quest for the Ultimate Theory, Vintage Books,
New York.

Gribbin, John, 1984, In Search of Schrödinger’s Cat: Quantum Physics

and Reality, Corgi, London.

Guth, Alan H., 1997, The In ationary Universe: The Quest for a New
Theory of Cosmic Origins, Perseus Books Group, New York.

Lippincott, Kristen, Eco, Umberto, Gombrich, E. H. et al., 1999, The

Story of Time, Merrell Holberton, London. [This is a splendid book
on the most diverse aspects of time in culture and science.]

Lockwood, Michael, 1989, Mind, Brain and the Quantum, Basil

Blackwell, Oxford.

Novikov, Igor, 1998, The River of Time, Cambridge University Press,

Cambridge.

Penrose, Roger, 1989, The Emperor’s New Mind: Concerning

Computers, Minds, and the Laws of Physics, Oxford University
Press, Oxford.

Price, Huw, 1996, Time’s Arrow and Archimedes’ Point, Oxford

University Press, New York.

Rees, Martin, 1997, Before the Beginning: Our Universe and Others,
Simon & Schuster, London.

Rees, Martin, 1999, Just Six Numbers: The Deep Forces that Shape the
Universe, Weidenfeld & Nicolson, London.
Smolin, Lee, 1997, The Life of the Cosmos, Weidenfeld & Nicolson,
London (Oxford University Press, New York).

Thorne, Kip, 1994, Black Holes and Time Warps: Einstein’s Outrageous
Legacy, Norton, New York.

Tipler, Frank, 1995, The Physics of Immortality, Doubleday, New

York.

Weinberg, Steven, 1977, The First Three Minutes, Basic Books, New
York (André Deutsch, London).

Weinberg, Steven, 1993, Dreams of a Final Theory: The Search for the
Fundamental Laws of Nature, Vintage, London.

Wheeler, John Archibald, 1990, A Journey into Gravity and

Spacetime, Scienti c American Library, New York.

Will, Cli ord, 1986, Was Einstein Right?, Basic Books, New York.
BIBLIOGRAPHY

Arnowitt, R., Deser, S. and Misner, C.W., 1962, ‘The dynamics of

general relativity’, in Gravitation: An Introduction to Current
Research, L. Witten (ed.), Wiley, New York.

Baierlein, R.F., Sharp, D.H. and Wheeler, J.A., 1962, ‘Three-

dimensional geometry as a carrier of information about time’,
Physical Review, 126, 1864.

Banks, T., 1985, ‘TCP, quantum gravity, the cosmological constant

and all that . . .’, Nuclear Physics, B249, 332.

Barbour, J.B., 1989, Absolute or Relative Motion? Vol. 1: The

Discovery of Dynamics, Cambridge University Press, Cambridge;
reprinted as the paperback The Discovery of Dynamics, Oxford
University Press, New York (2001).

Barbour, J.B., 1994a, ‘The timelessness of quantum gravity: I. The

evidence from the classical theory; II. The appearance of
dynamics in static con gurations’, Classical and Quantum Gravity,
11, 2853.

Barbour, J.B., 1994b, ‘On the origin of structure in the universe’, in

Philosophy, Mathematics and Modern Physics’, E. Rudolph and I.-O.
Stamatescu (eds), Springer, Berlin.
Barbour, J.B., 1999, ‘The development of Machian themes in the
twentieth century’, in The Arguments of Time, J. Butter eld (ed.),
published for The British Academy by Oxford University Press.

Barbour, J.B., 2001, ‘General covariance and best matching’, in

Physics Meets Philosophy at the Planck Length, C. Callender and N.
Huggett (eds), Cambridge University Press.

Barbour, J.B., forthcoming, Absolute or Relative Motion?, Vol. 2,

Oxford University Press, New York (see note to Chapter 4 on
p.344).

Barbour, J.B., Foster, Brendan Z. and Ó Murchadha, N., 2000,

‘Relativity without relativity’, http//xxx.lanl.gov/abs/gr-
qc/0012089.

Barbour, J.B. and Ó Murchadha, N., 1999, ‘Classical and quantum

gravity on con-formal superspace’, http//xxx.lanl.gov/abs/gr-
qc/9911071.

Barbour, J.B. and P ster, H. (eds), 1995, Mach’s Principle: From

Newton’s Bucket to Quantum Gravity, Birkhauser, Boston.

Bell, J., 1987, Speakable and Unspeakable in Quantum Mechanics,

Cambridge University Press, Cambridge.

Bohm, D., 1952, ‘A suggested interpretation of the quantum theory

in terms of “hidden” variables: I and II’, Physical Review, 85, 166
(reprinted in Wheeler and Zurek 1983).

Boltzmann, L., 1895, ‘On certain questions of the theory of gases’,

Nature, 51, 413.
Born, M., 1982, Einstein’s Theory of Relativity, Dover, New York.

Brout, R., 1987, ‘On the concept of time and the origin of the
cosmological temperature’, Foundations of Physics, 17, 603.

Brown, H.R., 1996, ‘Mindful of quantum possibilities’, British Journal

for the Philosophy of Science, 47, 189.

Carnap, R., 1963, ‘Autobiography’, in The Philosophy of Rudolf

Carnap, P.A. Schilpp (ed.), Library of Living Philosophers, P.A.

Clemence, G.M., 1957, ‘Astronomical time’, Reviews of Modern

Physics, 29, 2.

DeWitt, B.S., 1967, ‘Quantum theory of gravity’, Physical Review,

160, 1113.

DeWitt, B.S., 1970, ‘Spacetime as a sheaf of geodesics in

superspace’, in Relativity, M. Carmeli et al. (eds), Plenum, New
York.

DeWitt, B.S. and Graham, N. (eds), 1973, The Many-Worlds

Interpretation of Quantum Mechanics, Princeton University Press,
Princeton.

Dirac, P.A.M., 1983, ‘The evolution of the physicists’ picture of

nature’, Scienti c American, 208, 45.

Eliot, T.S., 1964, Knowledge and Experience in the Philosophy of F. II.

Bradley, Faber & Faber, London.

Fischer, A.E., 1970, ‘The theory of superspace’, in Relativity, M.

Carmeli et al. (eds), Plenum, New York.
Giulini, D., 1999, ‘The generalized thin-sandwich problem and its
local solvability’, Journal of Mathematical Physics, 40, 2470.

Halliwell, J. and Hawking, S.W., 1985, ‘Origin of structure of the

universe’, Physical Review D, 31, 1777.

Halliwell, J.J., Pérez-Mercader, J. and Zurek, W.H. (eds), 1994, The

Physical Origins of Time Asymmetry, Cambridge University Press,
Cambridge.

Heisenberg, W., 1930, The Physical Principles of the Quantum Theory,

Chicago University Press, Chicago.

Hofman, S.A., Kuchar, K. and Teitelboim, C., 1976,

‘Geometrodynamics regained’, Annals of Physics, 96, 88.

Holland, P., 1993, The Quantum Theory of Motion, Cambridge

University Press, Cambridge.

Isham, C., 1993, ‘Canonical quantum gravity and the problem of

time’, in Integrable Systems, Quantum Groups, and Quantum Field
Theories, L.A. Ibort and M.A. Rodriguez (eds), Kluwer Academic,
London.

Kiefer, C., 1997, ‘Does time exist at the most fundamental level?’, in
Time, Temporality, Now, H. Atmanspacher and E. Ruhnau (eds),
Springer, Berlin.

Kuchar, K., 1992, ‘Time and interpretations of quantum gravity’, in

Proceedings of the 4th Canadian Conference on General Relativity
and Relativistic Astrophysics, G. Kunstatter et al. (eds), World
Scienti c, Singapore.
Lanczos, C., 1986, The Variational Principles of Mechanics, Dover,
New York.

Lapchinskii, V.G. and Rubakov, V.A., 1979, ‘Canonical quantization

of gravity and quantum eld theory in curved space-time’, Acta
Physica Polonica BIO, 1041.

Liebscher, D.-E., 1999, Einsteins Relativitätstheorie und die Geometrien

der Ebene, Teubner, Stuttgart.

Littlejohn, R.G. and Reinsch, M., 1997, ‘Gauge elds in the

separation of rotations and internal motions in the n-body
problem’, Reviews of Modern Physics, 69, 213.

Mach, E., 1883, Die Mechanik in Hirer Entwickelung. Historisch-kritisch

dargestellt, J.A. Barth, Leipzig. (English translation (1960): The
Science of Mechanics: A Critical and Historical Account of its
Development, Open Court, LaSalle, IL.)

McTaggart, J.M.E., 1927, The Nature of Existence, Cambridge

University Press, Cambridge.

Misner, C.W., Thorne, K.S. and Wheeler, J.A., 1973, Gravitation,

W.H. Freeman, San Francisco.

Page, D. and Wootters, W., 1983, ‘Evolution without evolution:

Dynamics described by stationary observables’, Physical Review D,
27, 2885.

Pais, A., 1982, ‘Subtle is the Lord . . . ‘The Science and Life of Albert
Einstein, Oxford University Press, Oxford.
Pinker, S., 1997, How the Mind Works, Penguin, London/Norton,
New York.

Poincaré, H., 1902, La Science et l’hypothesis, Flammarion, Paris.

(English translation (1905): Science and Hypothesis, Walter Scott,
London.)

Poincaré, H., 1904, La Valeur de la Science, Flammarion, Paris.

(English translation (1907): The Value of Science, Science Press,
New York; reprinted in 1958 by Dover, New York.)

Schrödinger, E., 1944, What is Life?, Cambridge University Press,

Cambridge.

Shimony, A., 1997, ‘Implications of transience for spacetime

structure’, in The Geometric Universe: Science, Geometry, and the
Work of Roger Penrose, S.A. Huggett et al. (eds), Oxford University
Press, Oxford.

Silk, J., 1994, A Short History of the Universe, W.H.

Freeman/Scienti c American Library, New York, p. 87.

Smolin, L., 1991, ‘Space and time in the quantum universe’, in

Conceptual Problems of Quantum Gravity, A. Ashtekar and J.
Stachel (eds), Birkhauser, Boston.

Smolin, L., 2001, Three Roads to Quantum Gravity, Weidenfeld &

Nicolson, London (Basic Books, New York).

Stachel, J. (ed.), 1987, The Collected Papers of Albert Einstein, Vol. 1,

The Early Years, 1879–1902, Princeton University Press,
Princeton.
Tait, P.G., 1883, ‘Note on reference frames’, Proceedings of the Royal
Society’ of Edinburgh, Session 1883–4, p. 743.

Vilenkin, A., 1989, ‘Interpretation of the wave function of the

Universe’, Physical Review D, 39, 1116.

Walsch, N.D., 1997, Conversations with God—An Uncommon Dialogue,

Hodder & Stoughton, London.

Wheeler, J.A., 1964a, ‘Mach’s principle as boundary condition for

Einstein’s equations’, in Gravitation and Relativity, H.-Y. Chiu and
W.F. Hofmann (eds), Benjamin, New York.

Wheeler, J.A., 1964b, ‘Geometrodynamics and the issue of the nal

state’, in Relativity, Groups, and Topology: 1963 Les Houches
Lectures, C. DeWitt and B. DeWitt (eds), Gordon & Breach, New
York.

Wheeler, J.A. and Zurek, W. (eds), 1983, Quantum Theory and

Measurement, Princeton University Press, Princeton.

Williams, T., 1951, ‘The timeless world of a play’, in The Rose

Tattoo, New Directions Books, New York.

Zeh, H.-D., 1992, The Physical Basis of the Direction of Time, 2nd edn,
Springer, Berlin (3rd edn, 1999).
INDEX

aberration 128

absolute distance 74–5

absolute motion 62, 119

absolute rest 126

absolute space 20, 62–3, 83–5, 92, 119–20, 235–7

absolute time 20, 64, 83–5, 92, 119–20

action 111–13

ADM formalism 167, 253

aether 124, 126–8, 130–1, 153

Alpha 42, 46

alpha particle tracks 284–7, 290–6

angular momentum 86–90

anthropic principle 25, 342

antiparlicles 191

Arnowitt, Richard 167

arrow of time 19, 25, 27, 34, 308

Ashtekar, Abhay 351

Aspect, Alain 217

atomic time 107

atoms 24–5, 49

ionization 284–7, 290–6

number in a pea, 46–8
spectral lines 188–9

Augustine, St 11

Baierlein, Ralph 1 75–6

Baker, Dame Janet 332

Banks, Tom 258

Barbour, Julian

collaboration with Bertotti, 113–18, 167–72, 238

Discovery of Dynamics, The 239
website, 5, 7, 345

Barrow, John 341

Bell, John 220–1, 284, 287, 299–302

Bell’s inequalities 221

Berkeley, Bishop 63, 322

Bernoulli, John 110

Bertotti, Bruno 113–18, 167–72, 238

Besso, Michele 163

best matching 116–17, 170–2, 173, 175–6, 236–7

bicycle riding 88

bifurcation 55

Big Bang 12, 46, 251, 317, 320, 321

Big Crunch 263, 317, 321, 328

black holes 38, 166, 319, 320

block universe 143

body, human, as a time capsule 33

Bohm, David 218, 352

Bohr, Niels 13, 188–9, 219

Boltzmann, Ludwig 24–5, 27–8, 34, 342–3

Born, Max 198, 286, 310

Bose-Einstein statistics 191

bosons 191

bound states 234–5

boundary conditions 253

‘brachistochrone’ problem 110

brain
 damage 33
motion processing 28–30, 267
organization 31–2
state 26
as a time capsule 32–3

Broderick, Damien 358

Broglie, Louis de 189–90, 352

Brout, Robert 352

Brown, Harvey 353

Bruno, Giordano 75

BSW paper 176, 253

bucket experiment 62–3

calculus 63

Candide 111

canonical quantization 166

Capelin, Simon 239

Carnap, Rudolf 143

Carter, Brandon 341

causality 312–13
cells 33

centre of mass 79–81

centrifugal force 62–3

Chardin, Teilhard de 328

Christian, Joy 359

chronology of the universe 313–15

Clarke, Samuel 63

Clemence, Gerald 347

Clifton, Andrew 358

clocks 93–108

de ned 135
earth’s rotation 106
gravitational e ects 154–5
inertial 99–104
solar 97, 98
solar system 106–7
stellar 97, 98
universe as 107–8
water-clock 95

collapse of wave function 199–200, 204, 217–18, 224, 289, 295,

297
College Farm 3, 4

complementarity 204

complex numbers 197

con guration space 43, 208–9

con gurations 41, 43, 228

collinear 75

Confucius 17

consciousness 26–8, 31, 52–3

continuum spectrum 235

coordinate grid 129–32, 140–1, 206

Copenhagen interpretation 202–4

Copernicus, Nicolaus 12, 32

correlations 224, 257

cosmology see quantum cosmology

cosmos, well-ordered 322

covariance, general 160, 163–4

creation 22, 64, 110–11, 229, 232–5, 251, 254–6

Creator 326–7

curved surfaces 157–8

extrinsic/intrinsic curvature 158

shortest path on 111–12
cycling 88

Davies, Paul 322, 336, 337, 341, 358

Davisson, Clinton 190

Dawkins, Richard 48, 210, 324–5

De Broglie, Louis 189–90

De Chardin, Teilhard 328

degrees of freedom 79

two true 244

Descartes, René

on existence 49–50, 53
on motion 61–2

Deser, Stanley 167

Deutsch, David 333, 337, 352, 358

DeWitt, Bryce 38, 39, 221, 224, 246–7, 257, 351

on many-worlds 224–5

Diana, Princess of Wales, death of 324, 360

Dido, Queen 109–10

di erence

intrinsic 116–17, 172

 provisional 171–2

di raction 285

Dirac, Paul 2–3, 13, 15, 40, 167

on mental pictures 206

on self-interference 197
on space-time 2
spinor elds 191
transformation theory 206

Discovery of Dynamics, The 239, 344

discrete spectrum 234

dispersion relation 278

distance 74–5, 147, 150

absolute 74–5
in space-time 147, 150, 151

distinguished simpli er 120, 170–1

Dowker, Fay 353, 354–7, 359

duration 19, 100–3, 107, 123, 133, 136, 173

dynamics 12

Earth
 as a clock 106
rotation 32, 87
as a time capsule 33

eclipses 98–9

Eddington, Arthur 165

eigenfunctions 234

eigenstates 201, 202

eigenvalues 234–5

eikonal equation 270

Einstein, Albert 11, 134–6, 153

equivalence principle 154

on general covariance 163–4
home in Pavia 114
on Mach 65, 66
Nobel Prize for Physics 187
on Nows 143
on quantum theory 13, 186–7, 188
on real objects 220
relativity theory 12, 15, 129–32, 151–2, 155
on rotating systems 156–7
and space-time 155–6, 159, 160, 161
Einstein-Podolsky-Rosen paradox 218–20

electromagnetism 124–6

electrons 188, 189–90, 191

Eliot, George 330–1

Eliot, T.S. 240

energy 23, 90–2 see also kinetic energy; potential energy

energy levels 188

entangled states 216–17, 222–4

entropy 23–5, 27–8, 289, 318–9

ephemeris time 107

EPR paradox 218–20

equation of time 98

equivalence principle 154, 176–7

eternity 36, 40, 328

Euler, Leonhard 111

events 139, 143

Everett, Hugh 221–7

everything, theory of 192

evolution 252, 256

excited states 234

experience 51–2, 53–4

F

faces 33

falling bodies 93–6

Faraday, Michael 15, 124

Fermat, Pierre de 110

Fermat’s principle 271, 272

Fermi-Dirac statistics 191

fermions 191

Feynman, Richard 244

on time 2

elds 124, 190–1, 222

Fischer, Arthur 344

Fitzgerald, George 128

fossils 33

Foster, Brendan 358

Fourier, Joseph 201

frames of reference 81–2

free will 324–6

Fresnel, Augustin Jean 124, 269

Friedmann, Alexander 317

G

Galileo Galilei 12, 61

falling ball experiment 93–6

odd numbers rule 93–4
parabolic motion 94, 95, 96
on relativity 81–2, 151–2
water-clock of 95

Gamow, George 284

gas 318

Gauss, Carl Friedrich 157–8

Gaussian wave packets 278

general covariance 160, 163–4

geodesics 112, 116–17, 118

geometrical optics 270–1, 288

geometrodynamics 167

geometry, non-Euclidean 157–8

Germer, Lester 190

Gill, Ann 344

Giulini, Domenico 348

God 22, 64, 110–11, 326–7

gra ti describing time 44–5

grand uni ed theory 192

gravitons 244

gravity 12, 15, 91

curving space-time 12, 159, 160, 161

and time 154–5

Grossman, Marcel 159, 161

ground state 234

group velocity 278

Guichardet, A. 347

Guth, Alan 359

haemoglobin 48

Halliwell, Jonathan 258

Hamilton, William Rowan 111, 268, 269–74

Hartle-Hawking wave function 359

Hartle, Jim 353, 354, 359

Hawking, Stephen 39, 258

on Big Bang 320

on black holes 166, 319
on end of physics 13–14
 on time travel 329
no-boundary proposal 312

heaven 327–8

Heggie, Douglas 346

Heisenberg, Werner 13, 189, 284

Heisenberg cut 296

Heisenberg’s uncertainty principle 205

Heraclitus 1, 330

Hero of Alexandria 110

Hilbert space 193, 206

Hipparchus 98

history 69, 282–3, 287–92 see also records

Hofmann, Wenzel 115

Holland, Peter 352

Hopkins, Gerard Manley 75, 323

Hoyle, Fred 359

human body as a time capsule 33

Hut, Piet 346

hyperplanes 143–4

I
identity of indiscernibles 85–6

inertia 62, 65, 93

inertial clock 99–104

inertial frame of reference 82

in ation 359

initial conditions 22, 253

instants of time 16, 18, 247, 265–6

interactions 104–5

interference 124, 125, 195–6

self- 197
fringes 196

intrinsic di erence 116–17, 172

intrinsic time 245

invariable plane 88

ionization 284–7, 290–6

Isham, Christopher 351

James, William 28

Kant, Immanuel 207

Kaon decay 342

Keats, John 326, 331

Kepler, Johannes 2, 12, 98, 164, 269

Kiefer, Claus 352, 353

kinematic relativity 152

kinetic energy 90, 108, 112

Klein-Gordon equation 351

Kretschmann, Erich 164

Kubasiak, Gretchen Mills 334, 338–40

Kuchař, Karel 172–3, 238–9, 246, 349, 351

Kuhn, Thomas 115

LaFlamme, Raymond 358

Lagrange, Joseph-Louis de 111

Lanczos, Cornelius 268–9

Lange, Ludwig 346

Lapchinskii, V. 258

Laplace, Pierre 71, 76, 88

Laplacian determinism, perfect 83, 156, 177

Laue, Max von 190

least action, principle of 111–3

least-time theory 110

Leibniz, Willhelm Gottfried 16, 63–4, 74, 110, 119, 240, 322

on creation 63–4, 110–11

identity of indiscernibles 85–6
principle of su cient reason 64, 86
satirized as Dr Pangloss 111

Leibniz-Clarke Correspondence 63–4

length of time 19, 100–2, 107, 123, 133, 173

Liebscher, Dierck 6, 73, 346

light

cone 148
di raction 285
interference 124, 125
polarization 244
quanta (photons) 187, 188, 191, 243–4
rays 269–73
refraction 110
wave theory 124, 125, 269–71

‘line of time’ 27–8

linearity
 in quantum mechanics 225, 231, 275
of time 19, 27–8

Littlejohn, Robert 347

Lorentz, Hendrik Anton 128

Lorentz frames 141

Lorentz transformation 141

lottery, cosmic 52–3

Lovelock, James 4

Mach, Ernst 65, 153, 334

on inertia 65–6
on time 67

Machian distinguished simpli er 120, 170–1

Mach’s principle 65–6, 113–5

magnetism 124, 126

Makino, Jun 346

many-instants 302–5

many-worlds 53, 221–7, 229–302, 323–4

mass

centre of 79–81
 of particles 243–4

matrix mechanics 13, 189, 280

Maupertuis, Pierre 111

Maxwell, James Clerk 15, 124–5

McMillan, Steve 346

McTaggart, John 343

measurement problem 200, 224

measurements 200, 222–4

mechanics 12

Megamolecule Land 43

memory 19, 28, 32–3, 53, 55–6, 300

mental pictures 206–7

Mercury 161

Michelson-Morley experiment 128

Middlemarch 330–1

Midsummer Night’s Dream, A 276

Minkowski, Hermann 12, 138–42, 144, 147, 150–1

Minkowski space-time 141–2, 143, 145, 146, 174

Misner, Charles 167

mist(s) 51, 197

molecules 210, 235

momentum 200

angular 86–90
eigenstates 201

Monadology 111, 240

Moon 98–9, 205, 298

motion 61–70

absolute 62, 119

brain processing 28–30, 267
Descartes on 61–2
as an illusion 39, 46–9, 266–7, 329–35
Newtonian concepts 12, 20–1, 22, 62–3
‘philosophical’ 61
relative 61–2, 63, 119

Mott, Nevill 284, 307, 310, 354

movement see motion

music of the spheres 75, 325–6

Neumann, Carl 99–100

Newton, Isaac 11, 12, 20–2, 119

on absolute motion 62
on absolute space 20, 62–3
 on absolute time 20
on creation 22
on Descartes 62
laws of 12, 20–1, 22, 64
on making his discoveries 74
on motion 12, 20–1, 22, 62–3

Newtonian space-time 139–40, 146, 174

Newtonian time 107

no-boundary proposal of Hawking 312

Nobel Prize for Physics 187, 235

Nows 16, 18, 34, 40–1, 43, 44–5, 55–6, 174, 177, 333

Einstein on 143
and experiences 51–2, 53–4
and memories 55–6
in relativity 142–6
self-awareness 52–3

observables 204

odd numbers rule 93–4

Olivieri, Claudio 333

Omego 328
Ó Murchadha, Niall 35, 349, 352, 358

ontology 138, 301

optics 269–71

optimization 109–13

ovens, heating 185

Page, Don 257, 358

Pais, Abraham 123

Pangloss, Dr 111

parabolic motion 94, 95, 96

paradigms 115

Parmenides 1

particles 185–92

mass of 243–4

‘past’ 299, 300 see also history

paths 43–4, 69, 111–3, 118

Pavia 114

pea, number of atoms in 46–8

Penrose, Roger 7, 31, 166, 319, 320, 360

perihelion, advance of 161

persistence of objects 252

phase locking 259

‘philosophical’ motion 61

photons 187, 188, 191, 243–4

physics 38

end of 13–14
quantum versus classical 13, 15–16

Pied Beauty 75, 323

Pinker, Steven 342–3

Planck, Max 13, 185–6

Planck length 311

Planck mass 311

Planck time 311

Planck’s constant 186

planets see solar system

Plato 44, 45

Platonia 44, 45–6, 47, 344

asymmetry of 55, 307–9

for Minkowski space-time 145
for general relativity 167–9
geodesics in 116–17, 118
 ‘mist’ over 51–6, 197–8

Platonic forms 44, 251

Podolsky, Boris 218

Poincaré, Henri 71, 76, 128, 135–6, 152–3, 156, 180, 346

on duration 123

Point Omega 328

polarized light 244

pond argument 127

position eigenstates 202

potential energy 90–1, 108, 112, 117

preferred-basis problem 226–7, 301

Price, Huw 341

principal function 272

principle theories 134

probability 24, 198–9, 203–4, 356

proper time 150–1, 175

psychophysical parallelism 226, 301–2

Ptolemy, Claudius 99

pulsars 166

Purser, John 325–6

Purser, Michael 5, 325, 336

Pythagoras’ theorem 150

Q space 209

quanta 13 see also photons

quantization 15, 166

quantum cosmology 16, 197, 247, 251, 284, 311–21

quantum gravity 15–16, 240, 242–8

quantum inseparability 216–17

quantum mechanics 183–245

discovery of 12–13, 185–92

good measurement in 223
measurement problem 200, 224
measurements in 200–4, 212–24

quantum of action 185–6

radioactive decay 284

realism 252

real objects 220, 227, 228

records 30–1, 32–3, 284–7, 292–6, 300

refraction 110
Reinsch, M 347

Reissner, Hans 115

relative motion 61–2, 63, 119

relativity 12, 121–81

Einstein on 12, 15, 129–32, 151–2, 155

Galilean 81–2, 151–2
general 12, 155
kinematic 152
special 12, 152

rest 126–7

Ricci tensor 161

Riemann, Bernhard 158–9

Riemannian spaces 159, 169

Rosen, Nathan 218

rotation 86–90, 157

Rovelli, Carlo 351

Rubakov, V. 258

Rutherford, Ernest 188

Saturn 89, 90
scalar elds 191

Schrödinger, Erwin 13, 115, 189–90, 207, 208–9

on ‘I’ 331–2
Nobel Prize for Physics 235
stationary equation 230, 231, 232–5, 253
time-dependent Schrödinger equation 230, 231, 258
wave-packets 275–80

self-awareness 52–3

self-interference 197

semiclassical approach 258, 259–64, 273

semiclassical solution 258, 288, 296

Shape Space 73, 74–7

Shapere, A. 348

Sharp, David 175–6

Shimony, Abner 329, 343, 360

shortest paths 111–13, 117–18

shortest-time theory 110

sidereal time 97, 98

simultaneity 12, 123, 129–32

simultaneity hyperplanes 143–4

Smolin, Lee 239–41, 322, 327, 337, 351

snapshots 18–19, 20

solar system 12, 88–9

as a clock 106–7

solar time 97, 98

solipsism 50, 252, 300

space 138

absolute 20, 62–3, 83–5, 92, 119–20, 235–7

space-like 149

space-time 12, 138–42, 145–6, 155–6

curvature 12, 159, 160, 161

distance in 147, 150, 151
Minkowski 141–2, 143, 145, 146, 174
Newtonian 139–40, 146, 174
past, present and future 147–50

spatiotemporal framework 136

specious present 28

spectral lines 188–9

speed 96, 97

group 278

spin 86–90, 157, 218

spinor elds 191

spiral galaxies 88

St Augustine 11

stars as clocks 97, 98

‘state’ 26

stationary points 172

stationary Schrödinger equation 230, 231, 232–5, 253

stationary states 230, 231

statistical behaviour 191

stellar clocks 97, 98

strata 345

strati ed manifold 344

su cient reason, principle of 64, 86, 156

Sun 12, 316–17

as a clock 97, 98

superposition 201–2, 210–12, 225

superspace 169, 348

superstring theory 39, 166, 192, 359

supersymmetry 192

syzgies 78

T
Tait, Peter 100–1

Tail’s problem 100–4, 135, 139, 144, 347

tensors 161, 191

Tetrahedron Land 43, 168

theory of everything 192

thermodynamics 23

Thomson, George 190

Thomson, James 123, 135

Thorne, Kip 166, 319, 336–7

3-spaces 171

time 17–19

absolute 20, 64, 83–5, 92, 119–20

arrow of 19, 25, 27, 34, 308
atomic 107
directionality of 19
ephemeris 107
equation of 98
experiencing 19–20
gra ti describing 44–5
and gravity 154–5
instants of 16, 18, 247, 265–6
 intrinsic 245
length of 19, 100–2, 107, 123, 133, 173
‘line of 27–8
linearity of 19, 27–8
Newtonian 107
non-existence of 14–15, 35–8, 39, 247, 251, 309–11
proper 150, 175
sidereal 97, 98
solar 97, 98

time capsules 30–4, 251, 261, 264–7, 299, 300, 308

de ned 31

time-dependent Schrödinger equation 230, 231, 258

time-independent (stationary)

Schrödinger equation 230, 231, 232–5, 253

time-like 148

time travel 328–9

timelessness 14–15, 35–8, 39, 247, 251, 309–11

Tipler, Frank 328, 341

transformation theory 206

trial pairing 171

Triangle Land 40, 41, 42, 45–6, 116–17, 344

Twain, Mark 338

two-and-a-bit puzzle 109

two-slit experiment 194–6

two-snapshots problem 92, 103

unbound states 235

unity of unities 75

universal equation 254

universe 22

chronology of 313–15
as a clock 107–8
expansion 262–4
modelling 40–6
order of 23
wave function of 242

Vaughan, Henry 328

vector elds 191

velocity 96, 97

group 278
Vilenkin, Alexander 352, 359

Voltaire 111

von Laue, Max 190

Vonnegut, Kurt 359

Walsch, Neale Donald 344

water-clock 95

wave function 194–205, 241–2

collapse 199–200, 204, 217–18, 224, 297

of the universe 242

wave mechanics 13, 189–90, 194–205

wave optics 270

wave packets 201, 205, 262, 275–81

wave theory of light 124, 125, 269–71

website 5, 7, 344

Weinberg, Steven 244

well behaved, conditions of being 254

Wheeler, John Archibald 38, 44, 136, 167, 169, 175–6, 246–7

Wheeler-DeWitt equation 38–9, 242, 247, 256, 351, 359

Whitehead, Alfred North 329

Wilczek, Frank 347

Williams, Tennessee 331, 332

Wootters, William 257

World, The 328

world line 140

X-rays 190

Yeats, W.B. 322

Young, Thomas 124, 125

Zeh, Dieter 264, 307, 338, 354

Zeilinger, Anton 360

Zeno’s paradox 49