Questioning Basics of Physics

T H E
F R O N T I E R S
C O L L E C T I O N
Anthony Aguirre
Brendan Foster
Zeeya Merali (Eds.)
QUESTIONING
THE
FOUNDATIONS
OF PHYSICS
Which of Our Fundamental Assumptions
Are Wrong?
THE FRONTIERS COLLECTION

Series editors
Avshalom C. Elitzur
Unit of Interdisciplinary Studies, Bar-Ilan University, 52900 Ramat-Gan, Israel
e-mail: [email protected]
Laura Mersini-Houghton
Department of Physics, University of North Carolina, Chapel Hill,
NC 27599-3255, USA
T. Padmanabhan
Inter University Centre for Astronomy and Astrophysics (IUCAA) Pune, India
Maximilian Schlosshauer
Department of Physics, University of Portland, Portland, OR 97203, USA
Mark P. Silverman
Department of Physics, Trinity College, Hartford, CT 06106, USA
Jack A. Tuszynski
Department of Physics, University of Alberta, Edmonton, AB T6G 1Z2, Canada
Rdiger Vaas
Center for Philosophy and Foundations of Science, University of Giessen,
35394 Giessen, Germany
Series Editors
A.C. Elitzur L. Mersini-Houghton T. Padmanabhan
M.P. Silverman J.A. Tuszynski R. Vaas
M. Schlosshauer
The books in this collection are devoted to challenging and open problems at the
forefront of modern science, including related philosophical debates. In contrast to
typical research monographs, however, they strive to present their topics in a
manner accessible also to scientically literate non-specialists wishing to gain
insight into the deeper implications and fascinating questions involved. Taken as a
whole, the series reflects the need for a fundamental and interdisciplinary approach
to modern science. Furthermore, it is intended to encourage active scientists in all
areas to ponder over important and perhaps controversial issues beyond their own
speciality. Extending from quantum physics and relativity to entropy, consciousness and complex systemsthe Frontiers Collection will inspire readers to push
back the frontiers of their own knowledge.
More information about this series at http://www.springer.com/series/5342
For a full list of published titles, please see back of book or springer.com/series/5342
Anthony Aguirre Brendan Foster

Zeeya Merali
Editors
QUESTIONING
THE FOUNDATIONS
OF PHYSICS
Which of Our Fundamental Assumptions
Are Wrong?
123
Editors
Anthony Aguirre
Department of Physics
University of California
Santa Cruz, CA
USA
Zeeya Merali
Foundational Questions Institute
New York, NY
USA
Brendan Foster
Foundational Questions Institute
New York, NY
USA
ISSN 1612-3018
ISSN 2197-6619 (electronic)
ISBN 978-3-319-13044-6
ISBN 978-3-319-13045-3 (eBook)
DOI 10.1007/978-3-319-13045-3
Library of Congress Control Number: 2014957159
Springer Cham Heidelberg New York Dordrecht London
Springer International Publishing Switzerland 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.
Printed on acid-free paper
Springer International Publishing AG Switzerland is part of Springer Science+Business Media
(www.springer.com)
Preface
This book is a collaborative project between Springer and The Foundational

Questions Institute (FQXi). In keeping with both the tradition of Springers
Frontiers Collection and the mission of FQXi, it provides stimulating insights into a
frontier area of science, while remaining accessible enough to benet a nonspecialist audience.
FQXi is an independent, nonprot organization that was founded in 2006. It
aims to catalyze, support, and disseminate research on questions at the foundations
of physics and cosmology.
The central aim of FQXi is to fund and inspire research and innovation that is
integral to a deep understanding of reality, but which may not be readily supported
by conventional funding sources. Historically, physics and cosmology have offered
a scientic framework for comprehending the core of reality. Many giants of
modern sciencesuch as Einstein, Bohr, Schrdinger, and Heisenbergwere also
passionately concerned with, and inspired by, deep philosophical nuances of the
novel notions of reality they were exploring. Yet, such questions are often overlooked by traditional funding agencies.
Often, grant-making and research organizations institutionalize a pragmatic
approach, primarily funding incremental investigations that use known methods and
familiar conceptual frameworks, rather than the uncertain and often interdisciplinary methods required to develop and comprehend prospective revolutions in
physics and cosmology. As a result, even eminent scientists can struggle to secure
funding for some of the questions they nd most engaging, while younger thinkers
nd little support, freedom, or career possibilities unless they hew to such strictures.
FQXi views foundational questions not as pointless speculation or misguided
effort, but as critical and essential inquiry of relevance to us all. The Institute is
dedicated to redressing these shortcomings by creating a vibrant, worldwide
community of scientists, top thinkers, and outreach specialists who tackle deep
questions in physics, cosmology, and related elds. FQXi is also committed to
engaging with the public and communicating the implications of this foundational
research for the growth of human understanding.
vi
Preface
As part of this endeavor, FQXi organizes an annual essay contest, which is open
to everyone, from professional researchers to members of the public. These contests
are designed to focus minds and efforts on deep questions that could have a profound impact across multiple disciplines. The contest is judged by an expert panel
and up to 20 prizes are awarded. Each year, the contest features well over a hundred
entries, stimulating ongoing online discussion for many months after the close
of the contest.
We are delighted to share this collection, inspired by the 2012 contest, Questioning the Foundations: Which of Our Basic Physical Assumptions Are Wrong?
In line with our desire to bring foundational questions to the widest possible
audience, the entries, in their original form, were written in a style that was suitable
for the general public. In this book, which is aimed at an interdisciplinary scientic
audience, the authors have been invited to expand upon their original essays and
include technical details and discussion that may enhance their essays for a more
professional readership, while remaining accessible to non-specialists in their eld.
FQXi would like to thank our contest partners: The Gruber Foundation, SubMeta, and Scientic American. The editors are indebted to FQXis scientic
director, Max Tegmark, and managing director, Kavita Rajanna, who were
instrumental in the development of the contest. We are also grateful to Angela
Lahee at Springer for her guidance and support in driving this project forward.
2014
Anthony Aguirre
Brendan Foster
Zeeya Merali
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Anthony Aguirre, Brendan Foster and Zeeya Merali
The Paradigm of Kinematics and Dynamics Must Yield

to Causal Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Robert W. Spekkens
Recognising Top-Down Causation . . . . . . . . . . . . . . . . . . . . . . . .

George Ellis
17
On the Foundational Assumptions of Modern Physics . . . . . . . . .

Benjamin F. Dribus
45
The Preferred System of Reference Reloaded . . . . . . . . . . . . . . . .

Israel Perez
61
Right About Time? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Sean Gryb and Flavio Mercati
87
A Critical Look at the Standard Cosmological Picture . . . . . . . . .

Daryl Janzen
103
Not on but of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Olaf Dreyer
131
Patterns in the Fabric of Nature . . . . . . . . . . . . . . . . . . . . . . . . .

Steven Weinstein
139
10
Is Quantum Linear Superposition an Exact Principle

of Nature? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Angelo Bassi, Tejinder Singh and Hendrik Ulbricht
151
vii
viii
Contents
11
Quantum-Informational Principles for Physics . . . . . . . . . . . . . .

Giacomo Mauro DAriano
165
12
The Universe Is Not a Computer . . . . . . . . . . . . . . . . . . . . . . . . .

Ken Wharton
177
13
Against Spacetime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Giovanni Amelino-Camelia
191
14
A Chicken-and-Egg Problem: Which Came First,

the Quantum State or Spacetime? . . . . . . . . . . . . . . . . . . . . . . . .
Torsten Asselmeyer-Maluga
205
15
Gravity Can Be Neither Classical Nor Quantized . . . . . . . . . . . . .

Sabine Hossenfelder
219
16
Weaving Commutators: Beyond Fock Space . . . . . . . . . . . . . . . .

Michele Arzano
225
17
Reductionist Doubts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Julian Barbour
235
18
Rethinking the Scientic Enterprise: In Defense

of Reductionism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ian T. Durham
251
Is Life Fundamental? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sara Imari Walker
259
Titles in this Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
269
Appendix: List of Winners. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
273
19
Chapter 1
Introduction
Anthony Aguirre, Brendan Foster and Zeeya Merali
Our conceptions of Physical Reality can never be definitive; we

must always be ready to alter them, to alter, that is, the
axiomatic basis of physics, in order to take account of the facts
of perception with the greatest possible logical completeness.
(Einstein, A: Maxwells influence on the evolution of the idea of
physical reality. In: Thomson, J. J., ed.: James Clerk Maxwell: a
commemoration volume, pp. 6673. Cambridge University
Press (1931).)
Albert Einstein (1931)
Scientific development depends in part on a process of
non-incremental or revolutionary change. Some revolutions are
large, like those associated with the names of Copernicus,
Newton, or Darwin, but most are much smaller, like the
discovery of oxygen or the planet Uranus. The usual prelude to
changes of this sort is, I believe, the awareness of anomaly, of an
occurrence or set of occurrences that does not fit existing ways of
ordering phenomena. The changes that result therefore require
putting on a different kind of thinking-cap, one that renders the
anomalous lawlike but that, in the process, also transforms the
order exhibited by some other phenomena, previously
unproblematic. (Kuhn, T.S.: The Essential Tension (1977).)
Thomas S. Kuhn (1977)
Over the course of history, we can identify a number of instances where thinkers
have sacrificed some of their most cherished assumptions, ultimately leading to
scientific revolutions. We once believed that the Earth was the centre of the universe;
now, we know that we live in a cosmos littered with solar systems and extra-solar
A. Aguirre (B)
Department of Physics, University of California, Santa Cruz, CA, USA
B. Foster Z. Merali
Foundational Questions Institute, New York, NY, USA
Z. Merali
A. Aguirre et al. (eds.), Questioning the Foundations of Physics,
The Frontiers Collection, DOI 10.1007/978-3-319-13045-3_1
A. Aguirre et al.
planets. Cosmologists today are even questioning whether our universe is itself
unique or one of many parallel cosmoses.
Such paradigm shifts can be forced by experiment, an internal inconsistency in
accepted physics, or simply a particular philosophical intuition. Based, in part, on
the theoretical insight that the speed of light in a vacuum should be a constant, in
the early twentieth century, Einstein developed his special theory of relativity, which
threw out the common-sense belief that time and space are absolute. With his general
theory of relativity, Einstein went on to claim that space and time are stitched together
creating a four-dimensional fabric pervading the universe and that gravity manifests
as this fabric warps and bends around massive cosmic objects. Around the same time,
at the other extremity of scale, physicists realised that in order to explain perplexing
experimental results they must formulate a new set of rules for the behaviour of
subatomic entitiesquantum physicsthat muddies the boundaries between what
we define to be particles and what we traditionally think of as waves. Inherently
probabilistic, quantum theory also forces us to relinquish some of our deepest-held
intuitions and to accept that, at its core, reality may be indeterministic.
But those revolutions in our understanding raised as many questions as they
answered. Almost a century on, the time appears ripe for reassessing our current
assumptions. Relativity and quantum theory together form the cornerstones of modern physics but they have brought us to an impasse. Both theories have been corroborated by experiments; yet physicists have failed to bring the two descriptions
together into one overarching framework of quantum gravity, suggesting that one
or other, or even both, must be modified.
Astronomical observations also mock our understanding of the contents of the
universe. By monitoring galaxies, astronomers have surmised that most of the mass of
the universe resides in some unknown form, dubbed dark matter, that is detectable
only through its gravitational pull on visible matter. Furthermore, at the end of the
twentieth century, cosmologists were blind-sided by the discovery that the universe
is expanding at an accelerated rate, without any clear cause. This propulsive push is
now attributed to dark energy, but the origin and nature of this entity remains a
mystery.
The worlds biggest experiment at the Large Hadron Collider, at the CERN
laboratory, has recently helped to verify the standard model of particle physics with
unprecedented precision. Yet, this success has left physics with many unanswered
questions. The standard model cannot explain the nature of dark matter, or why
certain known particles have their observed masses and properties. In fact, if the
standard model is correct, it is difficult to understand how we even came to exist,
since it predicts that equal amounts of matter and antimatter should have been produced during the big bang, and that this matter and antimatter should subsequently
have annihilated leaving nothing behind to form stars, galaxies, or people.
It seems clear that we are lacking some fundamental insight. In order to understand
the origin of the universe, its contents and its workingsand our own existenceit
is likely that we will once again need to give up one or more of the notions that lie
at the base of our physical theories and which we currently hold sacred.
1 Introduction
So which of our current underlying preconceptionstacit or explicitneed

rethinking? That is the question that we posed in the 2012 FQXi contest: Questioning the Foundations: Which of Our Basic Physical Assumptions Are Wrong?
This was one of our broadest and most ambitious essay topics and it drew over 270
entries from Africa, Asia, Australasia, Europe, and North and South America. It
also generated record levels of discussion on our online forums. This volume brings
together the top 18 prize-winning entries.
Our first prize winner, Robert Spekkens, questions the distinction between a
theorys kinematicsthat is, the specification of the space of physical states it
allowsand its dynamicswhich encompasses the description of how these states
may evolve. Though this conceptual separation has traditionally been central to the
way that physicists build theories, whether classical or quantum, in Chap. 2, Spekkens
argues that it is a convention that should be abandoned. In its stead, he champions
underpinning new theories with a causal structure that explicitly relates variables
in terms of how they have been influenced by, or could in turn affect, other variables.
Chapters 3 and 4 also deal with causation. George Ellis scrutinizes the implicit
assumption that causation flows from the bottom upthat is, from micro to macro
scalesinstead positing that complexity in biology, and even the arrow of time,
emerge from a top-down causal flow from macroscopic scales downwards. Benjamin
Dribus meanwhile rejects the traditional spacetime manifold invoked by relativity in
favour of a new central principle based on considering causal order.
The tenets upon which relativity are built are examined in more detail in Chaps. 5
and 6. In particular, Israel Perez questions Einsteins assumption that there are no
preferred reference frames in the universe. In their essay, Sean Gryb and Flavio
Mercati propose unstitching time from space in Einsteins fabric and argue that the
fundamental description of reality must be based on shape.
Daryl Janzen also tackles physicists accepted conceptions of time. In Chap. 7, he
argues that by rethinking time in cosmological contexts, we may get a better handle
on cosmic expansion and the origin of dark energy. Chapter 8 also deals with current
mysteries in cosmology. Olaf Dreyer derives observable consequences that relate to
both dark energy and dark matter by reformulating these problems in a framework
in which particles are described as emergent excitations of the background, rather
than as existing on a background.
Connecting cosmology and quantum mechanics in Chap. 9, Steven Weinstein
challenges the orthodox view that physical facts at one point in space must be held
independent from those at another point. In so doing, he argues, we may better
understand the surprising homogeneity of the universe on cosmic scales and also the
origin of quantum entanglementthe spooky property that appears to link distant
quantum particles so that measurements of one influence the properties of its partners.
Chapters 1012 deal specifically with aspects at the foundations of quantum
theory. Angelo Bassi, Tejinder Singh and Hendrik Ulbricht question the principle
of quantum linear superposition (that is, the consensus notion that the actual state
of a quantum particle is the sum of its possible states). Although this has been
experimentally confirmed for relatively small particles and molecules, they note that
superposition breaks down for macroscopic objects; tables are never seen in two
A. Aguirre et al.
places at once, for instance. The team proposes experiments to test whether quantum
theory is an approximation to a stochastic non-linear theory. In his essay, Giacomo
DAriano searches for new quantum-information principles at the foundations of
physics based on epistemological and operational rules. In Chap. 12, Ken Wharton
argues that aspects of quantum physics would feel less paradoxical and may be open
to explanation if we let go of the intuitive implicit belief that the universe is effectively a computer, processing itself in the same time-evolved manner that we use
when performing calculations.
The challenge of devising a theory of quantum gravity that will unite quantum
theory with Einsteins general theory of relativity occupies the authors of Chaps. 13
16. Debates over the best approach for developing such a unified theory often focus
on whether quantum theory or our general-relativistic view of spacetime is more
fundamental. Giovanni AmelinoCamelia argues that when quantum mechanical
effects dominate, the assumption that spacetime exists becomes a hindrance and
should be thrown out. By contrast, Torsten AsslemeyerMaluga reviews both options
in Chap. 14that either spacetime must be quantized or that spacetime emerges from
something deeperand then presents an alternative view in which spacetime defines
the quantum state. Sabine Hossenfelder also makes the case for a third way, arguing
that the final theory need not be either classical or quantized. In Chap. 16, Michele
Arzano opens a new avenue for approaching a potential theory of quantum gravity
by scrutinizing the founding principles of quantum field theory that determine the
structure of the quantum fields.
To close the volume, we include award-winning entries that looked at the
philosophical stance of reductionism. In Chap. 17, Julian Barbour argues that while
reductionism has been a successful approach in science, in order to understand quantum mechanics and other mysteries such as the arrow of time, we may require a more
holistic approach. Ian Durham defends reductionism in Chap. 18, but questions the
paradigm that modern science simply consists of posing questions and then testing them. Finally, in Chap. 19, Sara Walker examines the merits of reductionism
for tackling perhaps the biggest unanswered question of allthe origin of lifeby
challenging the edict that all life is just chemistry.
In summary, the volume brings together an eclectic mix of approaches for addressing current mysteries that range from the peculiarities of the subatomic quantum scale
to those that span cosmic distances, examining our beliefs about time, causation, and
even the source of the spark of life, along the way. The winners include experts in
physics, mathematics, astronomy, astrobiology, condensed-matter physics, aerospace
engineering, and cosmology and each provides ample food for thought for the basis
of our next scientific revolution.
Chapter 2
The Paradigm of Kinematics and Dynamics

Must Yield to Causal Structure
Robert W. Spekkens
Abstract The distinction between a theorys kinematics and its dynamics, that is,
between the space of physical states it posits and its law of evolution, is central
to the conceptual framework of many physicists. A change to the kinematics of a
theory, however, can be compensated by a change to its dynamics without empirical
consequence, which strongly suggests that these features of the theory, considered
separately, cannot have physical significance. It must therefore be concluded (with
apologies to Minkowski) that henceforth kinematics by itself, and dynamics 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 notion of causal structure seems to provide
a good characterization of this union.
Proposals for physical theories generally have two components: the first is a
specification of the space of physical states that are possible according to the theory,
generally called the kinematics of the theory, while the second describes the possibilities for the evolution of the physical state, called the dynamics. This distinction
is ubiquitous. Not only do we recognize it as a feature of the empirically successful
theories of the past, such as Newtonian mechanics and Maxwells theory of electromagnetism, it persists in relativistic and quantum theories as well and is even
conspicuous in proposals for novel physical theories. Consider, for instance, some
recent proposals for how to unify quantum theory and gravity. Fay Dowker describes
the idea of causal histories as follows [1]:
The hypothesis that the deep structure of spacetime is a discrete poset characterises causal
set theory at the kinematical level; that is, it is a proposition about what substance is the
subject of the theory. However, kinematics needs to be completed by dynamics, or rules
about how the substance behaves, if one is to have a complete theory.
She then proceeds to describe the dynamics. As another example, Carlo Rovelli
describes the basics of loop quantum gravity in the following terms [2]:
The kinematics of the theory is well understood both physically (quanta of area and volume,
discrete geometry) and from the mathematical point of view. The part of the theory that is not
yet fully under control is the dynamics, which is determined by the Hamiltonian constraint.
R.W. Spekkens (B)
Perimeter Institute for Theoretical Physics, Waterloo, Ontario N2L 2Y5, Canada
R.W. Spekkens
In the field of quantum foundations, there is a particularly strong insistence that any
well-formed proposal for a physical theory must specify both kinematics and dynamics. For instance, Sheldon Goldstein describes the de Broglie-Bohm interpretation [3]
by specifying its kinematics and its dynamics [4]:
In Bohmian mechanics a system of particles is described in part by its wave function, evolving, as usual, according to Schrdingers equation. However, the wave function provides
only a partial description of the system. This description is completed by the specification of
the actual positions of the particles. The latter evolve according to the guiding equation,
which expresses the velocities of the particles in terms of the wave function.
John Bell provides a similar description of his proposal for a pilot-wave theory for
fermions in his characteristically whimsical style [5]:
In the beginning God chose 3-space and 1-time, a Hamiltonian H, and a state vector |0.
Then She chose a fermion configuration n (0). This She chose at random from an ensemble
of possibilities with distribution D (0) related to the already chosen state vector |0. Then
She left the world alone to evolve according to [the Schrdinger equation] and [a stochastic
jump equation for the fermion configuration].
The distinction persists in the Everett interpretation [6], where the set of possible
physical states is just the set of pure quantum states, and the dynamics is simply given
by Schrdingers equation (the appearance of collapses is taken to be a subjective
illusion). It is also present in dynamical collapse theories [7, 8], where the kinematics
is often taken to be the same as in Everetts approachnothing but wavefunction
while the dynamics is given by a stochastic equation that is designed to yield a good
approximation to Schrdinger dynamics for microscopic systems and to the von
Neumann projection postulate for macroscopic systems.
While proponents of different interpretations of quantum theory and proponents of
different approaches to quantizing gravity may disagree about the correct kinematics
and dynamics, they typically agree that any proposal must be described in these
terms.
In this essay, I will argue that the distinction is, in fact, conventional: kinematics
and dynamics only have physical significance when considered jointly, not separately.
In essence, I adopt the following methodological principle: any difference between
two physical models that does not yield a difference at the level of empirical phenomena does not correspond to a physical difference and should be eliminated. Such a
principle was arguably endorsed by Einstein when, from the empirical indistinguishability of inertial motion in free space on the one hand and free-fall in a gravitational
field on the other, he inferred that one must reject any model which posits a physical
difference between these two scenarios (the strong equivalence principle).
Such a principle does not force us to operationalism, the view that one should
only seek to make claims about the outcomes of experiments. For instance, if one
didnt already know that the choice of gauge in classical electrodynamics made no
difference to its empirical predictions, then discovery of this fact would, by the lights
of the principle, lead one to renounce real status for the vector potential in favour of
only the electric and magnetic field strengths. It would not, however, justify a blanket
rejection of any form of microscopic reality.
2 The Paradigm of Kinematics and Dynamics
As another example, consider the prisoners in Platos cave who live out their lives
learning about objects only through the shadows that they cast. Suppose one of the
prisoners strikes upon the idea that there is a third dimension, that objects have a threedimensional shape, and that the patterns they see are just two-dimensional projections
of this shape. She has constructed a hidden variable model for the phenomena. Suppose a second prisoner suggests a different hidden variable model, where in addition
to the shape, each object has a property called colour which is completely irrelevant
to the shadow that it casts. The methodological principle dictates that because the
colour property can be varied without empirical consequence, it must be rejected
as unphysical. The shape, on the other hand, has explanatory power and the principle finds no fault with it. Operationalism, of course, would not even entertain the
possibility of such hidden variables.
The principle tells us to constrain our model-building in such a way that every
aspect of the posited reality has some explanatory function. If one takes the view that
part of achieving an adequate explanation of a phenomenon is being able to make
predictions about the outcomes of interventions and the truths of counterfactuals,
then what one is seeking is a causal account of the phenomenon. This suggests
that the framework that should replace kinematics and dynamics is one that focuses
on causal structure. I will, in fact, conclude with some arguments in favour of this
approach.
Different Formulations of Classical Mechanics

Already in classical physics there is ambiguity about how to make the separation
between kinematics and dynamics. In what one might call the Newtonian formulation of classical mechanics, the kinematics is given by configuration space, while in
the Hamiltonian formulation, it is given by phase space, which considers the canonical momentum for every independent coordinate to be on an equal footing with the
coordinate. For instance, for a single particle, the kinematics of the Newtonian formulation is the space of possible positions while that of the Hamiltonian formulation
is the space of possible pairs of values of position and momentum. The two formulations are still able to make the same empirical predictions because they posit different
dynamics. In the Newtonian approach, motion is governed by the Euler-Lagrange
equations which are second-order in time, while in the Hamiltonian approach, it is
governed by Hamiltons equations which are first order in time.
So we can change the kinematics from configuration space to phase space and
maintain the same empirical predictions by adjusting the dynamics accordingly. Its
not possible to determine which kinematics, Newtonian or Hamiltonian, is the correct
kinematics. Nor can we determine the correct dynamics in isolation. The kinematics
and dynamics of a theory can only ever be subjected to experimental trial as a pair.
R.W. Spekkens
On the Possibility of Violating Unitarity

in Quantum Dynamics
Many researchers have suggested that the correct theory of nature might be one
that shares the kinematics of standard quantum theory, but which posits a different
dynamics, one that is not represented by a unitary operator. There have been many
different motivations for considering this possibility. Dynamical collapse theorists,
for instance, seek to relieve the tension between a systems free evolution and its
evolution due to a measurement. Others have been motivated to resolve the black
hole information loss paradox. Still others have proposed such theories simply as
foils against which the predictions of quantum theory can be tested [9].
Most of these proposals posit a dynamics which is linear in the quantum state
(more precisely, in the density operator representing the state). For instance, this is
true of the prominent examples of dynamical collapse models, such as the proposal
of Ghirardi et al. [7] and the continuous spontaneous localization model [8]. This
linearity is not an incidental feature of these models. Most theories which posit
dynamics that are nonlinear also allow superluminal signalling, in contradiction with
relativity theory [10]. Such nonlinearity can also lead to trouble with the second law
of thermodynamics [11].
There is an important theorem about linear dynamics that is critical for our
analysis: such dynamics can always be understood to arise by adjoining to the system
of interest an auxiliary system prepared in some fixed quantum state, implementing
a unitary evolution on the composite, and finally throwing away or ignoring the auxiliary system. This is called the Stinespring dilation theorem [12] and is well-known
to quantum information theorists.1
All proposals for nonunitary but linear modifications of quantum theory presume
that it is in fact possible to distinguish the predictions of these theories from those
of standard quantum mechanics. For instance, the experimental evidence that is
championed as the smoking gun which would rule in favour of such a modification
is anomalous decoherencean increase in the entropy of the system that cannot be
accounted for by an interaction with the systems environment. Everyone admits that
such a signature is extremely difficult to detect if it exists. But the point Id like to
make here is that even if such anomalous decoherence were detected, it would not
vindicate the conclusion that the dynamics is nonunitary. Because of the Stinespring
dilation theorem, such decoherence is also consistent with the assumption that there
are some hitherto-unrecognized degrees of freedom and that the quantum system
under investigation is coupled unitarily to these.2
1 It is analogous to the fact that one can simulate indeterministic dynamics on a system by deterministic dynamics which couples the system to an additional degree of freedom that is probabilistically
distributed.
2 A collapse theorist will no doubt reject this explanation on the grounds that one cannot solve
the quantum measurement problem while maintaining unitarity. Nonetheless, our argument shows
that someone who does not share their views on the quantum measurement problem need not be
persuaded of a failure of unitarity.
So, while it is typically assumed that such an anomaly would reveal that quantum
theory was mistaken in its dynamics, we could just as well take it to reveal that quantum theory was correct in its dynamics but mistaken in its kinematics. The experimental evidence alone cannot decide the issue. By the lights of our methodological
principle, it follows that the distinction must be purely conventional.
Freedom in the Choice of Kinematics

for Pilot-Wave Theories
The pilot-wave theory of de Broglie and Bohm supplements the wavefunction with
additional variables, but it turns out that there is a great deal of freedom in how to
choose these variables. A simple example of this arises for the case of spin. Bohm,
Schiller, and Tiomno have proposed that particles with spin should be modeled as
extended rigid objects and that the spinor wavefunction should be supplemented not
only with the positions of the particles (as is standardly done for particles without
spin), but with their orientation in space as well [13]. In addition to the equation
which governs the evolution of the spinor wavefunction (the Pauli equation), they
propose a guidance equation that specifies how the positions and orientations evolve
over time.
But there is another, more minimalist, proposal for how to deal with spin, due
to Bell [14]. The only variables that supplement the wavefunction in his approach
are the particle positions. The particles follow trajectories that are different from the
ones they would follow if they did not have spin because the equations of motion for
the particle positions depend on the spinor wavefunction.
The Bohm, Schiller and Tiomno approach and the Bell approach make exactly
the same experimental predictions. This is possible because our experience of quantum phenomena consists of observations of macroscopic variables such as pointer
positions rather than direct observation of the properties of the particle.
The non-uniqueness of the choice of kinematics for pilot-wave theories is not
isolated to spin. It is generic. The case of quantum electrodynamics (QED) illustrates this well. Not only is there a pilot-wave theory for QED, there are multiple
viable proposals, all of which produce the same empirical predictions. You could
follow Bohms treatment of the electromagnetic field, where the quantum state is
supplemented by the configuration of the electric field [15]. Alternatively, you could
make the supplementary variable the magnetic field, or any other linear combination of the two. For the charges, you could use Bells discrete model of fermions
on a lattice (mentioned in the introduction), where the supplementary variables are
the fermion numbers at every lattice point [5]. Or, if you preferred, you could use
Colins continuum version of this model [16]. If you fancy something a bit more
exotic, you might prefer to adopt Struyve and Westmans minimalist pilot-wave theory for QED, which treats charges in a manner akin to how Bell treats spin [17]. Here,
the variables that are taken to supplement the quantum states are just the electric field
strengths. No variables for the charges are introduced. By virtue of Gausss law, the
10
R.W. Spekkens
field nonetheless carries an image of all the charges and hence it carries an image of
the pointer positions. This image is what we infer when our eyes detect the fields.
But the charges are an illusion. And, of course, according to this model the stuff of
which we are made is not charges either: we are fields observing fields.
The existence of many empirically adequate versions of Bohmian mechanics has
led many commentators to appeal to principles of simplicity or elegance to try to
decide among them. An alternative response is suggested by our methodological
principle: any feature of the theory that varies among the different versions is not
physical.
Kinematical Locality and Dynamical Locality

I consider one final example, the one that first set me down the path of doubting
the significance of the distinction between kinematics and dynamics. It concerns
different notions of locality within realist models of quantum theory. Unlike a purely
operational interpretation of quantum theory, a realist model seeks to provide a causal
explanation of the experimental statistics, specifically, of the correlations that are
observed between control variables in the preparation procedure and outcomes of the
measurement procedure. It is presumed that it is the properties of the system which
passes between the devices that mediates the causal influence of the preparation
variable on the measurement outcome [18]. We refer to a full specification of these
properties as the systems ontic state.
It is natural to say that a realist model has kinematical locality if, for any two
systems A and B, every ontic state AB of the composite is simply a specification of
the ontic state of each component,
AB = ( A , B ) .
In such a theory, once you have specified all the properties of A and of B, you
have specified all of the properties of the composite AB. In other words, kinematical
locality says that there are no holistic properties.3
It is also natural to define a dynamical notion of locality for relativistic theories:
a change to the ontic state A of a localized system A cannot be a result of a change
to the ontic state B of a localized system B if B is outside the backward light-cone
of A. In other words, against the backdrop of a relativistic space-time, this notion of
locality asserts that all causal influences propagate at speeds that are no faster than
the speed of light.
Note that this definition of dynamical locality has made reference to the ontic
state A of a localized system A. If A is part of a composite system AB with holistic
properties, then the ontic state of this composite, AB , need not factorize into A
and B therefore we cannot necessarily even define A . In this sense, the dynamical
notion of locality already presumes the kinematical one.
3
The assumption has also been called separability [19].
11
It is possible to derive Bell inequalities starting from the assumption of kinematical

and dynamical locality together with a few other assumptions, such as the fact that the
measurement settings can be chosen freely and the absence of retrocausal influences.
Famously, quantum theory predicts a violation of the Bell inequalities. In the face of
this violation, one must give up one or more of the assumptions. Locality is a prime
candidate to consider and if we do so, then the following question naturally arises:
is it possible to accommodate violations of Bell inequalities by admitting a failure
of the dynamical notion of locality while salvaging the kinematical notion?
It turns out that for any realist interpretations of quantum theory wherein the ontic
state encodes the quantum state, termed -ontic models4 in Ref. [19], there is a
failure of both sorts of locality. In such models, kinematical locality fails simply by
virtue of the existence of entangled states. This is the case for all of the interpretations
enumerated in the introduction: Everett, collapse theories, de Broglie-Bohm. Might
there nonetheless be some alternative to these interpretations that does manage to
salvage kinematical locality?
Ive told the story in such a way that this seems to be a perfectly meaningful
question. But I would like to argue that, in fact, it is not.
To see this, it suffices to realize that it is trivial to build a model of quantum
theory that salvages kinematical locality. For example, we can do so by a slight
modification of the de Broglie-Bohm model. Because the particle positions can be
specified locally, the only obstacle to satisfying kinematical locality is that the other
part of the ontology, the universal wavefunction, does not factorize across systems
and thus must describe a holistic property of the universe. This conclusion, however,
relied on a particular way of associating the wavefunction with space-time. Can we
imagine a different association that would make the model kinematically local? Sure.
Just put a copy of the universal wavefunction at every point in space. It can then pilot
the motion of every particle by a local causal influence. Alternatively, you could put
it at the location of the center of mass of the universe and have it achieve its piloting
by a superluminal causal influenceremember, we are allowing arbitrary violations
of dynamical locality, so this is allowed. Or, put it under the corner of my doormat
and let it choreograph the universe from there.
The point is that the failure of dynamical locality yields so much leeway in the
dynamics that one can easily accommodate any sort of kinematics, including a local
kinematics. Of course, these models are not credible and no one would seriously
propose them,5 but what this suggests to me is not that we should look for nicer
models, but rather that the question of whether one can salvage kinematical locality
was not an interesting one after all. The mistake, I believe, was to take seriously the
distinction between kinematics and dynamics.
4 Upon learning this terminology, a former student, Chris Granade, proposed that the defining
feature of these types of modelthat the ontic state encodes the quantum stateshould be called
-ontology. I and other critics of -ontic approaches have since taken every opportunity to score
cheap rhetorical points against the -ontologists.
5 Norsen has proposed a slightly more credible model but only as a proof of principle that kinematical
locality can indeed be achieved [20].
12
R.W. Spekkens
Summary of the Argument

A clear pattern has emerged. In all of the examples considered, we seem to be able to
accommodate wildly different choices of kinematics in our models without changing
their empirical predictions simply by modifying the dynamics, and vice-versa. This
strikes me as strong evidence for the view that the distinction between kinematics and
dynamicsa distinction that is often central to the way that physicists characterize
their best theories and to the way they constrain their theory-buildingis purely
conventional and should be abandoned.
From Kinematics and Dynamics to Causal Structure

Although it is not entirely clear at this stage what survives the elimination of the
distinction between kinematics and dynamics, I would like to suggest a promising
candidate: the concept of causal structure.
In recent years, there has been significant progress in providing a rigorous mathematical formalism for expressing causal relations and for making inferences from
these about the consequences of interventions and the truths of counterfactuals. The
work has been done primarily by researchers in the fields of statistics, machine learning, and philosophy and is well summarized in the texts of Spirtes et al. [21] and
Pearl [22]. According to this approach, the causal influences that hold among a set of
classical variables can be modeled by the arrows in a directed acyclic graph, of the
sort depicted in Figs. 2.1 and 2.2, together with some causal-statistical parameters
describing the strengths of the influences.
The causal-statistical parameters are conditional probabilities P (X |Pa (X )) for
every X , where Pa (X ) denotes the causal parents of X , that is, the set of variables
that have arrows into X . If a variable X has no parents within the model, then one
simply specifies P(X ). The graph and the parameters together constitute the causal
model.
It remains only to see why this framework has some hope of dispensing with the
kinematics-dynamics distinction in the various examples I have presented.
The strongest argument in favour of this framework is that it provides a way to
move beyond kinematical and dynamical notions of locality. John Bell was someone
who clearly endorsed the kinematical-dynamical paradigm of model-building, as
the quote in the introduction illustrates, and who recognized the distinction among
notions of locality, referring to models satisfying kinematical locality as theories
of local beables [23]. In his most precise formulation of the notion of locality,
howeverwhich, significantly, he called local causalityhe appears to have transcended the paradigm of kinematics and dynamics and made an early foray into the
new paradigm of causal structure.
Consider a Bell-type experiment. A pair of systems, labeled A and B, are prepared
together and then taken to distant locations. The variable that specifies the choice of
13
Fig. 2.1 The causal graph

associated with Bells notion
of local causality
measurement on A (respectively B) is denoted S (respectively T ) and the variable

specifying the measurements outcome is denoted X (respectively Y ). Bell interprets
the question of whether a set of correlations P (X Y |ST ) admits of a locally causal
explanation as the question of whether the correlations between X and Y can be
entirely explained by a common cause , that is, whether they can be explained by a
causal graph of the form illustrated in Fig. 2.1. From the causal dependences in this
graph, we derive that the sorts of correlations that can be achieved in such a causal
model are those of the form

P (X |S) P (Y |T ) P().
P (X Y |ST ) =
Such correlations can be shown to satisfy certain inequalities, called the Bell inequalities, which can be tested by experiments and are found to be violated in a quantum
world.
If we think of the variable as the ontic state of the composite AB, then we see
that we have not needed to specify whether or not factorizes as ( A , B ). Bell
recognized this fact and emphasized it in his later writing: It is notable that in this
argument nothing is said about the locality, or even localizability, of the variable
[24]. Indeed, whether is localized in the common past of the two measurement
events and effects them by means of intermediary influences that propagate subluminally, or whether it is localized under my doormat and influences them superluminally, or whether it is not even localized at all, is completely irrelevant. All that is
needed to prove that P(X Y |ST ) must satisfy the Bell inequalities is that whatever the
separate kinematics and dynamics might be, together they define the effective causal
structure that is depicted in Fig. 2.1. By our methodological principle, therefore, only
the effective causal structure should be considered physically relevant.6
We see that Bells argument manages to derive empirical claims about a class of
realist models without needing to make any assumptions about the separate nature
of their kinematics and dynamics. This is a remarkable achievement. I propose that
it be considered as a template for future physics.
This analysis also suggests that the concepts of space and time, which are primitive within the
paradigm of kinematics and dynamics, ought to be considered as secondary concepts that are
ultimately defined in terms of cause-effect relations. Whereas in the old paradigm, one would
consider it to be part of the definition of a cause-effect relation that the cause should be temporally
prior to the effect, in the new paradigm, what it means for one event to be temporally prior to another
is that the first could be a cause of the second.
14
R.W. Spekkens
Fig. 2.2 Causal graphs for

Hamiltonian (left) and
Newtonian (right)
formulations of classical
mechanics
It is not as clear how the paradigm of causal structure overcomes the conventionality of the kinematics-dynamics distinction in the other examples Ive presented,
but there are nonetheless some good reasons to think that it is up to the task.
Consider the example of Hamiltonian and Newtonian formulations of classical
mechanics. If we let Q i denote a coordinate at time ti and Pi its canonically conjugate
momentum, then the causal models associated respectively with the two formulations
are depicted in Fig. 2.2. The fact that Hamiltonian dynamics is first-order in time
implies that the Q and P variables at a given time are causally influenced directly
only by the Q and P variables at the previous time. Meanwhile, the second-order
nature of Newtonian dynamics is captured by the fact that Q at a given time is
causally influenced directly by the Qs at two previous times. In both models, we have
a causal influence from Q 1 to Q 3 , but in the Newtonian case it is direct, while in the
Hamiltonian case it is mediated by P2 . Nonetheless, the kinds of correlations that can
be made to hold between Q 1 and Q 3 are the same regardless of whether the causal
influence is direct or mediated by P2 .7 The consequences for Q 3 of interventions
upon the value of Q 1 also are insensitive to this difference. So from the perspective
of the paradigm of causal structure, the Hamiltonian and Newtonian formulations
appear less distinct than they do if one focusses on kinematics and dynamics.
Empirical predictions of statistical theories are typically expressed in terms of
statistical dependences among variables that are observed or controlled. My guiding
methodological principle suggests that we should confine our attention to those causal
features that are relevant for such dependences. In other words, although we can
convert a particular claim about kinematics and dynamics into a causal graph, not all
features of this graph will have relevance for statistical dependences. Recent work
that seeks to infer causal structure from observed correlations has naturally gravitated
towards the notion of equivalence classes of causal graphs, where the equivalence
relation is the ability to produce the same set of correlations. One could also try to
7
There is a subtlety here: it follows from the form of the causal graph in the Newtonian model that
Q 1 and Q 4 are conditionally independent given Q 2 and Q 3 , but in the Hamiltonian case, this fact
must be inferred from the causal-statistical parameters.
15
characterize equivalence classes of causal models while allowing for restrictions on

the forms of the conditional probabilities and while allowing not only observations of
variables but interventions upon them as well. Such equivalence classes, or something
like them, seem to be the best candidates for the mathematical objects in terms of
which our classical models of physics should be described.
Finally, by replacing conditional probabilities with quantum operations, one
can define a quantum generalization of causal modelsquantum causal models
[25, 26]which appear promising for providing a realist interpretation of quantum
theory. It is equivalence classes of causal structures here that are likely to provide
the best framework for future physics.
The paradigm of kinematics and dynamics has served us well. So well, in fact,
that it is woven deeply into the fabric of our thinking about physical theories and
will not be easily supplanted. I have nonetheless argued that we must abandon it.
Meanwhile, the paradigm of causal structure is nascent, unfamiliar and incomplete,
but it seems ideally suited to capturing the nonconventional distillate of the union of
kinematics and dynamics and it can already claim an impressive achievement in the
form of Bells notion of local causality.
Rest in peace kinematics and dynamics. Long live causal structure!
Acknowledgments My thanks to Howard Wiseman and Travis Norsen for valuable discussions,
especially those on the subject of kinematical locality. Research at Perimeter Institute is supported
by the Government of Canada through Industry Canada and by the Province of Ontario through the
Ministry of Research and Innovation.
References
1. F. Dowker. Causal sets and the deep structure of spacetime, ed. by A. Ashtekar. 100 Years of
Relativity, Space-Time Structure: Einstein and Beyond, pp. 445464 (2005).
2. C. Rovelli. Loop quantum gravity. Living Rev. Relativ. 11 (2008).
3. D. Bohm, A suggested interpretation of the quantum theory in terms of hidden variables I.
Phys. Rev. 85(2), 166 (1952).
4. S. Goldstein. Bohmian mechanics. The Stanford Encyclopedia of Philosophy (Fall 2012 edn.).
5. J.S. Bell, Beables for quantum field theory. Phys. Rep. 137, 4954 (1986).
6. H. Everett, III., Relative state formulation of quantum mechanics. Rev. Mod. Phys. 29(3),
454 (1957).
7. G.C. Ghirardi, A. Rimini, T. Weber, Unified dynamics for microscopic and macroscopic systems. Phys. Rev. D 34(2), 470 (1986).
8. G.C. Ghirardi, P. Pearle, A. Rimini, Markov processes in Hilbert space and continuous spontaneous localization of systems of identical particles. Phys. Rev. A 42(1), 78 (1990).
9. S. Weinberg, Testing quantum mechanics. Ann. Phys. 194(2), 336386 (1989).
10. N. Gisin, Weinbergs non-linear quantum mechanics and supraluminal communications. Phys.
Lett. A 143(1), 12 (1990).
11. A. Peres, Nonlinear variants of Schrdingers equation violate the second law of thermodynamics. Phys. Rev. Lett. 63(10), 1114 (1989).
12. W.F. Stinespring, Positive functions on C*-algebras. Proceedings of the American Mathematical Society 6, 211216 (1955).
13. D. Bohm, R. Schiller, J. Tiomno, A causal interpretation of the Pauli equation (A). Il Nuovo
Cimento 1, 4866 (1955).
16
R.W. Spekkens
14. J.S. Bell, On the impossible pilot wave. Found. Phys. 12(10), 989999 (1982).
15. D. Bohm, A suggested interpretation of the quantum theory in terms of hidden variables II.
Phys. Rev. 85(2), 180 (1952).
16. S. Colin, A deterministic Bell model. Phys. Lett. A 317(5), 349358 (2003).
17. W. Struyve, H. Westman, A minimalist pilot-wave model for quantum electrodynamics. Proc.
R. Soc. A: Math. Phys. Eng. Sci. 463, 31153129 (2007).
18. R.W. Spekkens, Contextuality for preparations, transformations, and unsharp measurements.
Phys. Rev. A 71(5), 052108 (2005).
19. N. Harrigan, R.W. Spekkens, Einstein, incompleteness, and the epistemic view of quantum
states. Found. Phys. 40(2), 125157 (2010).
20. T. Norsen, The theory of (exclusively) local beables. Found. Phys. 40(12), 18581884 (2010).
21. P. Spirtes, C.N. Glymour, R. Scheines, Causation, Prediction, and Search, 2nd edn. (The MIT
Press, 2001).
22. J. Pearl, Causality: Models, Reasoning, and Inference, 2nd edn. (Cambridge University Press,
2009).
23. J.S. Bell, The theory of local beables. Epistemo. Lett. 9, 1124 (1976).
24. J.S. Bell, Bertlmanns socks and the nature of reality. Le J. de Phys. Colloq. 42(C2), 4161
(1981).
25. M.S. Leifer, Quantum dynamics as an analog of conditional probability. Phys. Rev. A 74(4),
(2006).
26. M.S. Leifer, R.W. Spekkens, Towards a formulation of quantum theory as a causally neutral
theory of Bayesian inference. Phys. Rev. A 88(5), 052130 (2013).
Chapter 3
Recognising Top-Down Causation

George Ellis
The Theme
A key assumption underlying most present day physical thought is that causation
in the hierarchy of complexity is bottom up all the way: particle physics underlies
nuclear physics, nuclear physics underlies atomic physics, atomic physics underlies
chemistry, and so on, and this is all that is taking place. Thus all the higher level
subjects are in principle reducible to particle physics, which is therefore the only
fundamental science; all the rest are derivative, even if we do not have the computing
power to demonstrate this in detail. As famously claimed by Dirac, chemistry is just
an application of quantum physics (see [60]). It is implied (or sometimes explicitly
stated) that the same is true for biology and psychology.
Interlevel Causation
However there are many topics that one cannot understand by assuming this one-way
flow of causation. Inter alia,
The flourishing subject of social neuroscience makes clear how social influences
act down on individual brain structure [13], giving the mechanism by which our
minds get adapted to the society in which they are imbedded [5];
the study of sensory systems shows conclusively that our senses do not work in
a bottom up way with physical input from the environment uniquely determining
what we experience; rather our expectations of what we should see play a key
role [32];
studies in physiology demonstrate that downward causation is key in physiological
systems. For example it is needed to understand the functioning of the heart, where
this form of causation can be represented as the influence of initial and boundary
G. Ellis(B)
University of Cape Town, Cape Town, South Africa
17
18
G. Ellis
conditions on the solutions of the differential equations used to represent the lower
level processes [51];
epigenetic and developmental studies demonstrate that biological development
is crucially influenced by the environment in which development takes place
[34, 51];
evolutionary theory makes clear that living beings are adapted to environmental
niches, which means that environmental influences shape animal structure, function, and behavior [11].
In each case, the larger environment acts down to affect what happens to the components at lower levels of the hierarchy of structure. This does not occur by violating
physical laws; on the contrary, it occurs through directing the effects of the laws of
physics by setting constraints on lower level interactions.
Now many believe that insofar as this view is not just a trivial restatement of
the obvious, it is wrong, because it implies denying the validity of the physics that
underlies our material existence. This is however not the case; at no point do I deny
the power and validity of the underlying physical causation. The crucial point is that
even though physical laws always completely characterize the interactions at their
own level of interaction (that between the physical components out of which complex
entities arise), they do not by themselves determine unique outcomes either at the
lower or higher levels. The specific outcomes that in fact occur are determined by
the context in which those physical interactions take place; for example whether the
electrons and protons considered are imbedded in a digital computer, a rock, a dog,
a river, an elephant, an aircraft, or a trombone. Context has many dimensions: complexity arises out of suitable modular hierarchical structures, each layer influencing
both those above and those below. Indeed that is why there are so many different
subjects (chemistry, biochemistry, molecular biology, physiology, ecology, environmental science, evolutionary biology, ecology, psychology, anthropology, sociology,
economics, politics, and so on) that characterize our complex existence. Only in the
case of physical chemistry is there some chance of reducing the higher level emergent
behaviour to nothing but the outcome of lower level causal interactions; and even
there it does not actually work, inter alia because of the issue of the arrow of time
(see sect. The Arrow of Time).
What about physics itself? In this essay I make the case that top-down causation
is also prevalent in physics [23], even though this is not often recognized as such.
Thus my theme is [26],
Interlevel causation: The assumption that all causation is bottom up is wrong, even in the
case of physics. Both bottom up and top down effects occur in the hierarchy of complexity,
and together enable higher level emergent behaviour involving true complexity to arise
through the existence of inter-level feedback loops.
Some writers on this topic prefer to refer to contextual effects or whole-part constraints. These are perfectly acceptable terms, but I will make the case that the stronger
term top-down causation is appropriate in many cases. As stated above, this is not
an exercise in mysticism, or denial of physical causation; it is an important assertion
about how causality, based essentially in physics at the bottom levels, works out in
the real world.
3 Recognising Top-Down Causation
19
Two Basic Issues

To set the scene, I give some definitions on which what follows is based.
Causation
The nature of causation is highly contested territory, although less so than before
[53, 54]. I will take a pragmatic view:
Definition 1 (Causal Effect) If making a change in a quantity X results in a reliable
demonstrable change in a quantity Y in a given context, then X has a causal effect
on Y.
Example I press the key labelled A on my computer keyboard; the letter A appears
on my computer screen.
Note: the effect may occur through intermediaries, e.g. X may cause C which in
turn causes Y. It still remain true that (through this chain) X causes Y. What is at issue
here is what mechanism enables X to influence Y. The overall outcome is unaffected
by this issue.
Now there are of course a myriad of causal influences on any specific event: a
network of causation is always in action. What we usually do is to have some specific
context in mind where we keep almost all parameters and variables fixed, allowing
just one or two remaining ones to vary; if they reliably cause some change in the
dependent variable in that context, we then label them as the cause. For example in
the case of a digital computer, we have
(Physics, computer structure, specific software, data) (output)
(3.1)
Now in a particular run on a specific computer, the laws of physics do not change
and the high level software loaded (e.g. Microsoft Word) will be fixed, so the above
reduces to
(3.2)
(data) ===
(output)
MS-Word
If however we load new high level software (e.g. now we run Photoshop) we will
end up with a different relation than (3.2):
(data) ===
(output)
(3.3)
Photoshop
Hence both the data and the software are causes of the output, in the relevant context.
The laws of physics are also a cause, as is the existence of the Earth and the designer
of the computer, but we usually dont bother to mention this.
20
G. Ellis
Existence
Given this understanding of causation, it implies a view on ontology (existence) as
follows: I assume that physical matter (comprised of electrons, protons, etc.) exists.
Then the following criterion for existence makes sense
Definition 2 (Existence) If Y is a physical entity made up of ordinary matter, and X
is some kind of entity that has a demonstrable causal effect on Y as per Definition 1,
then we must acknowledge that X also exists (even if it is not made up of such matter).
This is clearly a sensible and testable criterion; in the example above, it leads to the
conclusion that both the data and the relevant software exist. If we do not adopt this
definition, we will have instances of uncaused changes in the world; I presume we
wish to avoid that situation.
Hierarchy of Scales and Causation

The basic hierarchy of physical matter is indicated in Table 3.1,1 indicating physical
scales. It is also an indication of levels of bottom-up causation, with each lower level
underlying processes at the one next above in functional terms; hence one can call
it a hierarchy of causation. The computer hierarchy and life sciences hierarchy are
rather different: those hierarchies are based on causation rather than scale (for the
computer case, see section Complex Structures: Digital Computers).
Now a core aspect of this hierarchy of emergent properties is that one needs different vocabularies and language at the different levels of description. The concepts that
are useful at one level are simply inapplicable at other levels. The effective equations
at the various levels are valid at a specific level, describing same-level (emergent)
behaviour at that level. They are written in terms of the relevant variables at those
levels. These emergent variables can sometimes be obtained by coarse graining of
lower level states, but not always so (examples will be given below). One of the
characteristics of truly complex systems is that higher level variables are not always
just coarse grainings of lower level variables.
The key feature then is that the higher level dynamics is effectively decoupled
from lower level laws and details of the lower level variables [60]: except for some
highly structured systems, you dont have to know those details in order to understand
higher level behaviour. Thus you dont have to know particle physics in order to be
A fuller description is given here: http://www.mth.uct.ac.za/~ellis/cos0.html.
21
Table 3.1 The hierarchy of physical matter

Level Domain
Scale (m) Mass (kg) Example
L17
L16
L15
L14
L13
L12
L11
L10
L9
L8
L7
L6
L5
L4
L3
L2
L1
Cosmology
Large Scale Structures
Galaxy Clusters
Galaxies
Star clusters
Stellar systems
Stars
Planets
Continents
Land forms
Macro objects
Materials
Molecules/ chemistry
Atomic physics
Nuclear physics
Particle physics
Quantum gravity
1026
1023
1022
1021
1020
1012
1010
109
107
104
1
102
109
1010
1014
1015
1035
1053
1047
1045
1042
1035
1030
1030
1024
1017
108
10
101
1025
1026
1027
1033
Observable Universe
Great Attractor, Sloan Great wall
Virgo cluster, Coma cluster
M31, NGC 1300, M87
Messier 92, Messier 69
Binaries, Solar system
Sun, Proxima Centauri, Eta Carinae
Earth, Mars, Jupiter
Africa, Australia
Atlas mountains, Andes
Rocks, chairs, computers
Conductors, Insulators, semi-conductors
H2 0, SiO2 , C6 H12 O6 , C9 H13 N5 O12 P3
Hydrogen atom, Carbon atom
Neutron, Proton, Carbon nucleus
Quarks, electrons, gluons
Superstrings
either a motor mechanic or a zoologist. However you may need some knowledge of
chemistry if you are a doctor.
A sensible view is that the entities at each classical level of the hierarchy (Table 3.1)
are real [19, 51]. A chair is a real chair even though it is made of atoms, which in
turn are real atoms even though they are made of a nucleus and electrons, and so on;
and you too are real (else you could not read this paper), as is the computer on which
you are reading it. Issues of ontology may be unclear at the quantum level [28, 35],
but they are clear at the macro level.
In highly ordered structures, sometimes changes in some single micro state can have
major deterministic outcomes at the macro level (which is of course the environment for the micro level); this cannot occur in systems without complex structure.
Examples are,
1. A single error in microprogramming in a digital computer can bring the whole
thing to a grinding halt;
2. A single swap of bases in a gene can lead to a change in DNA that results in
predictable disease;
3. A single small poison pill can debilitate or kill an animal, as can damage to some
very specific micro areas in the brain.
22
G. Ellis
This important relation between micro structure and macro function is in contrast to
statistical systems, where isolated micro changes have no effect at the macro level,
and chaotic systems, where a micro change can indeed lead to a macro change, but
its unpredictable. Perhaps this dependable reliance on some specific lower level
details is a characteristic of genuine complexity.
The sequel: I will now look in turn at digital computers (sect. Complex
Structures: Digital Computers); life and the brain (sect. Complex Structures: Life
and the Brain); astronomy and cosmology (sect. Astronomy and Cosmology);
and physics (sect. Contextual Effects in Physics). The latter study will open the
way to considering how there can be the needed causal slack for top-down effects to
take place: how is there room at the bottom for all these effects, without overriding
the lower level physics? (sect. Room at the Bottom). The essay ends with some
comments on what this all implies for our understanding of causality (sect. The Big
Picture: The Nature of Causation).
Complex Structures: Digital Computers

Structured systems such as a computer constrain lower level interactions, and thereby
paradoxically create new possibilities of complex behaviour. For example, the specific connections between p-type and n-type transistors can be used to create NOR,
NAND, and NOT gates [45]; these can then be used to build up adders, decoders, flipflops, and so on. It is the specific connections between them that channels causation
and so enable the lower level entities to embody such logic; the physical structure
constrains the movement of electrons so as to form a structured interaction network.
The key physical element is that the structure breaks symmetry (cf: [2]), thereby
enabling far more complex behaviour than is possible in isotropic structures such as
a plasma, where electrons can go equally in any direction. This hardware structure
takes the form of a modular physical hierarchy (networks, computers, logic boards,
chips, gates, and so on [62]).
The Software Hierarchy

However hardware is only causally effective because of the software which animates
it: by itself hardware can do nothing. Software is also modular and hierarchically
structured [62], with the higher level logic of the program driving the lower level
events. The software hierarchy for digital computers is shown in Table 3.2. All but
the lowest level are virtual machines.
Entering data by key strokes is a macro activity, altering macro variables. This
acts down (effect T1) to set in motion a great number of electrons at the micro
physics level, which (effect D1) travel through transistors arranged as logic gates
at the materials level; finally (effect B1) these cause specific patterns of light on

Table 3.2 The software
hierarchy in a digital
computer system
23
Levels
Software hierarchy
Level 7
Level 6
Level 5
Level 4
Level 3
Level 2
Level 1
Level 0
Applications programs
Problem oriented language level
Assembly language level
Operating system machine level
Instruction set architecture level
Microarchitecture level
Digital logic level
Device level
a computer screen at the macro level, readable as text. Thus we have a chain of
top-down action T1 followed by lower level dynamical processes D1, followed by
bottom up action B1, these actions composed together resulting in a same level
effective macro action D2:
D2 = B1 o D1 o T 1
(3.4)
This is how effective same-level dynamics D2 at the higher level emerges from
the underlying lower level dynamics D1. This dynamics is particularly clear in the
case of computer networking [40], where the sender and receiver are far apart. At the
sender, causation flows downwards from the Application layer through the Transport,
Network and Link layers to the Physical layer; that level transports binary code to
the other computer through cable or wireless links; and then causation flows up the
same set of layers at the receiver end, resulting in effective same-level transport of
the desired message from source to destination. The same level effective action is not
fictitious: it is a reliable determinable dynamic effect relating variables at the macro
level of description at the sender and receiver. If this were not the case you would not
be able to read this article, which you obtained by such a process over the internet.
There was nothing fictitious about that action: it really happened. Emergent layers
of causation are real [2, 19, 51, 60].
The result is that the user sees a same-level interaction take place, and is unaware
of all the lower levels that made this possible. This is information hiding, which is
crucial to all modular hierarchical systems [8]. The specific flow of electrons through
physical gates at the physical level is determined by whether the high level software
is a music playing program, word processor, spreadsheet, or whatever: a classic case
of top-down causation (the lower level interactions would be different if different
software were loaded, cf. (3.2) and (3.3)). Hence what in fact shapes the flow of
electrons at the gate level is the logic of the algorithms implemented by the top level
computer program [43, 45].
24
G. Ellis
Key Issues
Four crucial points emerge.
A: Causal Efficacy of Non Physical entities: Both the program and the data are
non-physical entities, indeed so is all software. A program is not a physical thing
you can point to, but by Definition 2 it certainly exists. You can point to a CD or
flashdrive where it is stored, but that is not the thing in itself: it is a medium in
which it is stored. The program itself is an abstract entity, shaped by abstract logic.
Is the software nothing but its realisation through a specific set of stored electronic
states in the computer memory banks? No it is not because it is the precise pattern
in those states that matters: a higher level relation that is not apparent at the scale of
the electrons themselves. Its a relational thing (and if you get the relations between
the symbols wrong, so you have a syntax error, it will all come to a grinding halt).
This abstract nature of software is realised in the concept of virtual machines, which
occur at every level in the computer hierarchy except the bottom one [62]. But this
tower of virtual machines causes physical effects in the real world, for example when
a computer controls a robot in an assembly line to create physical artefacts.
B: Logical relations rule at the higher levels: The dynamics at all levels is
driven by the logic of the algorithms employed in the high level programs [41]. They
decide what computations take place, and they have the power to change the world
[43]. This abstract logic cannot be deduced from the laws of physics: they operate in a
completely different realm. Furthermore the relevant higher level variables in those
algorithms cannot be obtained by coarse graining any lower level physical states.
They are not coarse-grained or emergent variables: they are assigned variables, with
specific abstract properties that then mediate their behaviour.
C: Underlying physics allows arbitrary programs and data: Digital computers
are universal computers. The underlying physics does not constrain the logic or type
of computation possible, which Turing has shown is universal [14]. Physics does
not constrain the data used, nor what can be computed (although it does constrain
the speed at which this can be done). It enables the higher level actions rather than
constraining them. The program logic dictates the course of things.
D: Multiple realisability at lower levels. The same high level logic can be implemented in many different ways: electronic (transistors), electrical (relays), hydraulic
(valves), biological (interacting molecules) for example. The logic of the program
can be realised by any of these underlying physical entities, which clearly demonstrates that it is not the lower level physics that is driving the causation. This multiple
realisability is a key feature characterising top-down action [4]: when some high
level logic is driving causation at lower levels, it does not matter how that logic is
physically instantiated: it can be realised in many different ways. Thus the top-down
map T1 in (3.5) is not unique: it can be realised both in different physical systems,
and in different micro states of the same system.
25
Equivalence Classes
The last point means that we can consider as the essential variables in the hierarchy,
the equivalence classes of lower level states that all that correspond to the same high
level state [4]. When you control a top level variable, it may be implemented by any
of the lower level states that correspond to the chosen high level state; which one
occurs is immaterial, the high level dynamics is the same. You even can replace the
lower level elements by others with the same functionality, the higher entity remains
the same (a classic example: the cells in your body are now all different than they
were 7 years ago; you are made up of different matter, but you are still the same
person).
In digital computers, there are such equivalences all over the place:
at the circuit level: one can use Boolean algebra to find equivalent circuits;
at the implementation level: one can compile or interpret (giving completely different lower level functioning for same higher level outcome);
at the hardware level, one can run the same high level software on different microprocessors;
even more striking is the equivalence of hardware and software in much computing
(there is a completely different nature of lower level entities for the same higher
level outcomes).
In each case this indicates top-down effects are in action: the higher level function
drives the lower level interactions, and does not care how it is implemented (information hiding is taking place).
As to the use of the computer for represent text, the keyboard letters are never
exactly identical: yet the abstract letter A represented by them is still the letter A
despite many possible variations. It is also the letter A if
you change font (Times New Roman to Helvetica)

you change to bold or italic
you change size of the font
you change colour of the font
you change the medium from light on a computer screen to ink on paper
Such multiple realisability occurs at all levels in a text. One of the key problems in
generating intelligent understanding is to assign all these different representations
to the same abstract entity that they all represent. This way varied lower level representations of a higher level entity occur is characteristic of top-down causation
[4]: what matters is the equivalence class of all these representations, which is the
characteristic of the higher level entity, not which particular representation has been
chosen (see the Appendix).
26
G. Ellis
Implications
Hence although they are the ultimate in algorithmic causation as characterized so
precisely by Turing [14], digital computers embody and demonstrate the causal
efficacy of non-physical entities. The physics allows this; it does not control what
takes place, rather it enables the higher level logic to be physically implemented.
Computers exemplify the emergence of new kinds of causation out of the underlying
physics, not implied by physics but rather by the logic of higher level possibilities
as encoded in data structures and algorithms [8, 41, 43]. This leads to a different
phenomenology at each of the levels of Table 3.2, described by effective laws for
that level, and an appropriate language. A combination of bottom up causation and
contextual affects (top-down influences) enables their complex functioning.
Complex Structures: Life and the Brain

Living systems are modular hierarchical system, for the same reasons as in the
case of digital computers: this structuring enables complex behaviour inter alia it
because it allows class structures with inheritance, information hiding, and abstraction [8]. The lower level interactions are constrained by recurrent network structures,
thereby creating possibilities of truly complex behaviour. But the core dynamic that
allows true complexity to emerge is adaptive selection. Biological systems are finely
adapted to their physical, ecological, and social environments: and that cannot take
place without topdown information flows from the environment to the structure and
behaviour of organisms. This takes place on evolutionary, developmental, and functional timescales.
Microbiology
The rise of epigenetics has shown that the functioning of molecular biology is controlled by many epigenetic mechanisms that are sensitive to environmental effects,
such as DNA methylation [34]. Consequently the view that biology is controlled
bottom up by the actions of genes alone is fatally flawed.2 Contextual effects are
crucial in determining physical outcomes.
For a comprehensive discussion, see the many links on Denis Nobles webpage at http://
musicoflife.co.uk/.
27
Physiology
The molecular basis behind the physiology of an animal obeys the laws of physics
and chemistry. But by themselves they do not create entities such as a sensory or
circulatory system, nor determine their mode of functioning and resulting physical
outcomes. When you study the physiology of the heart you find it cannot be understood except in terms of the interplay of bottom up and top down causation, which
determines which specific molecular interactions take place where and at what time
[50, 51]. Bottom up physics alone cannot explain how a heart comes into being, nor
what its design is, nor its regular functioning.
The Brain
Top-down causation is prevalent at all levels in the brain: for example it is crucial
to vision [32, 39] as well as the relation of the individual brain to society [13] and
social decision making [63]. The hardware (the brain) can do nothing without the
excitations that animate it: indeed this is the difference between life and death. The
mind is not a physical entity, but it certainly is causally effective: proof is the existence
of the computer on which you are reading this text. It could not exist if it had not
been designed and manufactured according to someones plans, thereby proving the
causal efficacy of thoughts, which like computer programs and data are not physical
entities.
This is made possible firstly by the hierarchical structuring in the brain described
in [31]. His forward connections are what I call bottom up, and his backward
connections are what I call top-down (the difference in nomenclature is obviously
immaterial). He makes quite clear that a mix of bottom up and top down causation
is key as to how the brain works; backward connections mediate contextual effects
and coordinate processing channels. For example, the visual hierarchy includes 10
levels of cortical processing; 14 levels if one includes the retina and lateral geniculate
nucleus at the bottom as well as the entorhinal cortex and hippocampus at the top
[27]. Secondly, it depends on context-dependent computation by recurrent dynamics
in prefrontal cortex [46]. And thirdly, it happens by top-down reorganization of
activity in the brain after learning tasks on developmental timescales [33, 61] and by
environmental influence (of the effect of stress) on childrens physiological state by
influencing telomere length in chromosomes [48]. On evolutionary timescales, basic
social responses have been built into the brain through evolutionary processes even
in animals as simple as C. elegans, where a hub-and-spoke circuit drives pheromone
attraction and social behaviour [44].
These structural features and related cortical functioning are reflected in the way
the brain functions at the higher functional levels. I will give two illustrations of
to-down processes in psychology.
28
G. Ellis
Example 1: Reading
How does reading work? Heres a remarkable thing.
You can read this even through words are misspelt,
and this through letters are wrong,
And this through words missing.
How can it be we can make sense of garbled text in this way? One might think
the brain would come to a grinding halt when confronted with such incomplete or
grammatically incorrect text. But the brain does not work in a mechanistic way, first
reading the letters, then assembling them into words, then assembling sentences.
Instead our brains search for meaning all the time, predicting what should be seen
and interpreting what we see based on our expectations in the current context.
Actually words by themselves may not make sense without their context. Consider:
The horses ran across the plane,
The plane landed rather fast,
I used the plane to smooth the wood.
what plane means differs in each case, and is understood from the context. Even
the nature of a word (noun or verb) can depend on context:
Her wound hurt as she wound the clock
This example shows you cant reliably tell from spelling how to pronounce words
in English, because not only the meaning, but even pronunciation depends on context.
The underlying key point is that we are all driven by a search for meaning: this
is one of the most fundamental aspects of human nature, as profoundly recorded by
Viktor Frankl in his book Mans Search for Meaning [30]. Understanding this helps
us appreciate that reading is an ongoing holistic process: the brain predicts what
should be seen, fills in what is missing, and interprets what is seen on the basis of
what is already known and understood. And this is what happens when we learn to
read, inspired by the search for understanding. One learns the rules of grammar and
punctuation and spelling too of course; but such technical learning takes place as the
process of meaning making unfolds. It is driven top down by our predictions on the
basis of our understandings, based in meaning.
Example 2: Vision
Vision works in essentially the same way, as demonstrated by Dale Purves in his
book Brains: How They Seem to Work [59]. The core of his argument is as follows
[58]
The evolution of biological systems that generate behaviorally useful visual percepts has
inevitably been guided by many demands. Among these are: 1) the limited resolution of
photoreceptor mosaics (thus the input signal is inherently noisy); 2) the limited number of
29
neurons available at higher processing levels (thus the information in retinal images must
be abstracted in some way); and 3) the demands of metabolic efficiency (thus both wiring
and signaling strategies are sharply constrained). The overarching obstacle in the evolution
of vision, however, was recognized several centuries ago by George Berkeley, who pointed
out that the information in images cannot be mapped unambiguously back onto real-world
sources (Berkeley, 1975). In contemporary terms, information about the size, distance and
orientation of objects in space are inevitably conflated in the retinal image. In consequence,
the patterns of light in retinal stimuli cannot be related to their generative sources in the
world by any logical operation on images as such. Nonetheless, to be successful, visually
guided behavior must deal appropriately with the physical sources of light stimuli, a quandary
referred to as the inverse optics problem.
The resolution is top-down shaping of vision by the cortex, based in prediction of

what we ought to see. Visual illusions are evidence that this is the way the visual
system solves this problem [31, 59].
Adaptive Selection
As was mentioned above, adaptive selection is one of the most important types of
top-down causation. It is top-down because the selection criteria are at a different
conceptual level than the objects being selected: in causal terms, they represent a
higher level of causation [22]. Darwinian selection is the special case when one has
repeated adaptive selection with heredity and variation [56]. It is top-down because
the result is crucially shaped by the environment, as demonstrated by numerous
examples: for example a polar bear is white because the polar environment is white.
Section Complex Structures: Life and the Brain of [47] emphasizes how downward
causation is key to adaptive selection, and hence to evolution. An important aspect is
that multilevel selection must occur in complex animals: environmental changes have
no causal handle directly to genes but rather the chain of causation is via properties
of the group, the individual, or physiological systems at the individual level, down
to the effects of genes; that is, it is inherently a multilevel process [24].
The key point about adaptive selection (once off or repeated) is that it lets us locally
go against the flow of entropy, and lets us build up and store useful information
through the process of deleting what is irrelevant. Paul Davies and Sara Walker
explain that this implies that evolutionary transitions are probably closely related to
top-down causation [64]:
Although it has been notoriously difficult to pin down precisely what it is that makes life
so distinctive and remarkable, there is general agreement that its informational aspect is
one key property, perhaps the key property. The unique informational narrative of living
systems suggests that life may be characterized by context-dependent causal influences,
and in particular, that top-down (or downward) causation where higher-levels influence
and constrain the dynamics of lower-levels in organizational hierarchies may be a major
contributor to the hierarchal structure of living systems. Here we propose that the origin of
life may correspond to a physical transition associated with a fundamental shift in causal
structure. The origin of life may therefore be characterized by a transition from bottomup to top-down causation, mediated by a reversal in the dominant direction of the flow of
30
G. Ellis
information from lower to higher levels of organization (bottom-up), to that from higher to
lower levels of organization (top-down). Such a transition may be akin to a thermodynamic
phase transition, with the crucial distinction that determining which phase (nonlife or life)
a given system is in requires dynamical information and therefore can only be inferred by
identifying causal relationships. We discuss one potential measure of such a transition, which
is amenable to laboratory study, and how the proposed mechanism corresponds to the onset
of the unique mode of (algorithmic) information processing characteristic of living systems.
However adaptive selection occurs far more widely than that; e.g. it occurs in state
vector preparation [22], as I indicate below.
Astronomy and Cosmology

I now turn to the physical sciences: first astronomy and cosmology, and then physics
itself.
Astronomy
In the context of astronomy/astrophysics, there is a growing literature on contextual
effects such as suppression of star formation by powerful active galactic nuclei [52].
This is a top-down effect from galactic scale to stellar scale and thence to the scale
of nuclear reactions. Such effects are often characterised as feedback effects.
Cosmology
In cosmology, there are three venerable applications of the idea of top-down effects:
Olbers paradox, Machs Principle, and the Arrow of Time [7, 20] In each case it has
been strongly suggested that boundary conditions on the Universe as a whole are the
basic cause of crucial local effects (the dark night sky, the origin of inertia, and the
local direction of time that characterizes increasing entropy). Machs Principle is not
now much discussed, and I will not consider it further. More recent examples are
nucleosynthesis and structure formation in the expanding universe, though they are
not usually expressed in this way.
Nucleosynthesis
In the case of element formation in the early universe, macro-level variables (average
cosmic densities) determine the expansion rate of the cosmos, which determines the
31
temperature-time relation T (t). The Einstein equations for a radiation dominated

cosmology lead to a hot state evolving as

S(t) t 2/3 , t =
c2
15.5 GaT 4
1/2

2
t
T
=
1010 K
1.92 sec
(3.5)
This is the context within which nucleosynthesis takes place, and in turn determines
the rates of nuclear reactions (micro-level processes) and hence the outcome of nucleosynthesis, leading to macro variables such as the overall mass fraction of Helium
and Lithium in the early Universe. Occurrence of this top-down effect is why we can
use element abundance observations to constrain cosmological parameters [18].
Structure Formation
Another example is structure formation in the expanding universe, studied by examining the dynamics of perturbed cosmological models. Again macro-level variables
occur as coefficients in the relevant equations, determining the growth of perturbations and hence leading to macro variables such as the power spectrum of structure
in the universe. Occurrence of this top-down effect is the reason we can use observations of large scale structural features, such as power spectra and identification of
baryon acoustic oscillations, to constrain cosmological parameters [18, 49].
Olbers Paradox
The calculations leading to understanding of the CBR spectrum [18, 49] are basically
the present day version of the resolution of Olbers paradox (why the is the night sky
not as bright as the surface of the Sun): one of the oldest calculations of global to
local effects [7]. The essence of the resolution is that the universe is not infinite and
static: it has expanded from a state with a low entropy to baryon ratio.
The Arrow of Time

Boltzmanns H-theorem wonderfully proves, in a purely bottom up way by coarsegraining the microphysics states, that entropy S always increases with time:
dS/dt > 0. However the proof also works if one reverses the direction of time:
define T := t, exactly the same proof holds again, step by step, for this time T too,
which proves that also dS/dT > 0 [55]. The same holds for Weinbergs proof of
entropy increase ([66, p.150]), which is formulated in the language of quantum field
theory and avoids approximations such as the Born approximation. His H-Theorem
(3.6.20) will hold for both directions of time (just reverse the direction of time and
relabel to : the derivation goes through as before) because the derivation depends
32
G. Ellis
only on unitarity. Unitary transformations however are time reversible; there is therefore nothing in the dynamics that can choose one time direction as against the other as
far as any dynamical development is concerned, just as there is no intrinsic difference
between the particles and .3
This is Loschmidts paradox: because the microphysics considered in these two
derivations is time reversible, whichever direction of time you choose, exactly the
same proof shows that S cannot decrease in the opposite direction of time too. So
you cannot determine the arrow of time from time reversible microphysics alone; it
must come from somewhere else and the best candidate is special initial conditions
at the start of the universe on a macro scale [1, 12, 55]. In fact this is obvious: if the
micro physics is time symmetric, you cannot derive the macro arrow of time from it
alone, because of the symmetry under t t; it must derive from the cosmological
context. A clear presentation of why some kind of non-local condition is necessary
to resolve the arrow of time issue is given by [10]: bottom up effects alone are not
able to resolve it.
Thus the time asymmetry can come not from the dynamic equations of the theory
but from the boundary conditions. Consequently, one needs some global boundary
condition to determine a consistent arrow of time in local physics:a top-down effect
from the cosmological scale to everyday scales. This global coordination is plausibly
provided by a macro-scale low entropy condition of the early universe [12, 55]. Some
will claim this is nothing but the combined effect of all the particles and fields in the
very early universe. Well yes and no: the effect is indeed due to all those particles, but it
depends crucially on the specific relationship between them. Change that relationship
from smooth to very clumpy: exactly the same particles will be there, but will have a
very different effect. Its the spatial distribution that matters: the relationship between
them is the specific cause of the local physical effect of existence of a unique arrow
of time. You cannot describe that higher level relationship in terms of the variables
and scales relevant at the lower levels in Table 3.1. And the outcome is crucially
important: life would not be possible without a well-established local arrow of time.
Contextual Effects in Physics

Top-down causation happens wherever in addition to initial conditions, boundary conditions materially determine physical results; then environmental variables
(a macro scale concept) act down to determine the values of physical fields locally.
Such effects occur in both classical and quantum physics.
3 This derivation does not refer to the time irreversibility of the weak interaction, which has no
direct effect on everyday life: it cannot be the source of the arrow of time in physical chemistry and
biology.
33
Classical Physics
Examples of top-down effects in classical physics are standing waves; hearing the
shape of a drum [38]; outcomes of the reaction diffusion equation, for example in
pattern formation in embryos. In each case the outcome is largely determined by
the shape of the system boundary. Physicists often try to minimise such boundary
effects through the idealisation of an isolated system; but no real system is isolated
either in space or in time. Furthermore, that is in fact just another example of specific
boundary conditions: asymptotic flatness or outgoing radiation conditions are just as
much contextual constraints as periodic boundary conditions are.
There is nothing new in all this: its just that we dont usually talk about this as
a top-down effect. It may be helpful to do so; so I give two other rather different
examples.
The Ideal Gas Law

An example of top down action is given by the ideal gas law for gas constrained in
a cylinder by a piston. The relation
PV = nRT
(3.6)
between the macroscopic variables pressure P, volume V , amount of gas n, and

temperature T does not depend on detailed microscopic variables such as velocities,
positions and masses of any specific molecules; indeed we dont know those values.
The variables T and V at the macro level are the only handle we have on the state
of a given body of gas n: we cant (except in unusual circumstances) manipulate the
micro level variables directly. But we can change them by altering macro variables
(e.g. by compressing the gas, so changing V ), which then changes lower level states,
speeding up the molecules so as to increase the pressure. This is the top down effect
of the higher level variables on lower level states. The universal constant R is the link
between the micro and macro states, because it relates the energy of micro states to
the values observed at the bulk level.
Electromagnetism
The electromagnetic field is described by an anti-symmetric tensor Fab made up
of electric and magnetic field components. The micro level laws are (i) Maxwells
equations for the electromagnetic field, including the Gauss law, with particles such
as electron and protons as sources; (ii) the Lorentz force law for the effect of electric
and magnetic fields on charged particles such as protons and electrons. [29]. The
lower level dynamics is the way the micro level electric and magnetic fields interact
with each other according to Maxwells equations, with the charged particles as
34
G. Ellis
sources, and these fields in turn exert forces on those particles via the Lorentz force
law, that cause them to move if they are free to do so. Bottom up effects are the way
that billions of electrons in motion at the micro level generate measurable electric
currents at the macro level, such as in a generator, and the way that forces on billions
of charged particles at the micro level can add up to generate macroscopic forces,
such as in an electric motor.
Top down effects are for example the way that electric coils (macro entities, not
describable in terms of micro level variables) constrain the motion of electrons in
specific directions. This is why the micro fields they generate add up to macro level
fields ([29, pp. 13-5 and 13-6]). Without the constraints exerted by the wires, no such
macro fields would be generated. Similarly the constraints generated by the design
of the electric motor ensure that the individual forces on electrons and protons are
channelled so as to add up to measurable macro level forces. This constraining effect
is top down action from the level of machine components to the level of protons and
electrons. If one looks for example at the Feynman derivation of the magnetic field
due to electrons flowing in a wire coil, the wire is represented as a structureless macro
entity, even though it is made up of atoms and electrons. We just take this macro
structure (the physical wire) for granted in all such derivations, without enquiring
how it is made up of micro entities.
Thus the causal effectiveness of macro entitiesthe way the wire channels the
flow of electrons along the direction of the wireis taken for granted. Yes of course
it is made up of atoms and electrons at the micro level, but that is irrelevant to its
function in the experiment, which role is due to the macro organisation embodied in
this structure. These structural constraints act down to organise micro events, as is
very clear in the case of the wire: its physical structure prevents electrons moving
sideways out of the wire. This top-down aspect of what is going on is hidden because
we take it for granted. Its just part of what we assume to be the case, so we dont
notice it.
Quantum Physics
Top down effects occur in the context of quantum physics as well [22]. Here are
some examples.
Band Structure and Quantum Cooperative Effects

The periodic crystal structure in a metal leads (via Blochs theorem) to lattice waves,
and an electronic band structure depending on the particular solid involved, resulting
in all the associated phenomena resulting from the band structure [67] such as electrical resistivity and optical absorption, as well as being the foundation of processes
in solid-state devices such as transistors and solar cells. The entire machinery for
35
describing the lattice periodicity refers to a scale much larger than that of the electron,
and hence is not describable in terms appropriate to that scale.
Thus these effects all exist because of the macro level properties of the solid
the crystal structureand hence represent top-down causation from that structure
to the electron states. This can lead to existence of quasiparticles such as phonons
(collective excitations in a periodic, elastic arrangement of atoms or molecules) that
result from vibrations of the lattice structure. Because these are all based in top-down
action, they are emergent phenomena in the sense that they simply would not exist
if the macro-structure did not exist, and hence cannot be understood by a purely
bottom-up analysis, as emphasized strongly by Laughlin [42].
Caldeira-Leggett Model
The Caldeira-Leggett model is a model for a system plus heat reservoir, used for the
description of dissipation phenomena in solid state physics, as discussed in [9]. Here
the Lagrangian of the composite system T consisting of the system S of interest and
a heat reservoir B takes the form
LT = LS + LB + LI + LCT ,
(3.7)
where LS is the Lagrangian for the system of interest, LB that for the reservoir (a set
of non-interacting harmonic oscillators), and LI that for the interaction between
them. The last term LCT is a counter term, introduced to cancel an extra harmonic
contribution that would come from the coupling to the environmental oscillators.
This term represents a top-down effect from the environment to the system, because
LI completely represents the lower-level interactions between the system and the
environment. The effect of the heat bath is more than the sum of its parts when
LCT = 0, because the summed effect of the parts on each other is given by LI .
The bottom up effects of lower level forces acting on the components of the system
are completely described by LS and LI ; they were carefully constructed that way.
Similarly the bottom up dynamics of the environment is fully described by LB . The
term LCT is thus not an outcome of the bottom up effects alone.
Superconductivity and Superfluidity

In superconductivity, the electronsdespite their repulsion for each otherform
pairs (Cooper pairs) which are the basic entities of the superconducting state. This
happens by a cooperative process: the negatively charged electrons cause distortions
of the lattice of positive ions in which they move, and the real attraction occurs
between these distortions. It also can lead to superfluidity. These effects can only
occur because of the specific nature of the lattice structure, which is thus the source
of the existence of the Cooper pairs. The Nobel lecture by Laughlin [42] emphasizes
the implications:
36
G. Ellis
One of my favourite times in the academic year occurs in early spring when I give my
class of extremely bright graduate students, who have mastered quantum mechanics but are
otherwise unsuspecting and innocent, a take-home exam in which they are asked to deduce
superfluidity from first principles. There is no doubt a very special place in hell being reserved
for me at this very moment for this mean trick, for the task is impossible. Superfluidity, like
the fractional Hall effect, is an emergent phenomenon - a low-energy collective effect of
huge numbers of particles that cannot be deduced from the microscopic equations of motion
in a rigorous way, and that disappears completely when the system is taken apart.... The
world is full of things for which ones understanding, i.e. ones ability to predict what will
happen in an experiment, is degraded by taking the system apart, including most delightfully
the standard model of elementary particles itself.
The claim made here is that this dynamics is possible because of top-down causation.
Proton-Coupled Electron Transfer

Proton coupled electron transfer (PCET) describes the joint transfer of an electron
and proton to or from a substrate [65], hence it can only occur in the presence of
suitable substrate. This substrate is a higher level structure, i.e. at a larger scale than
the electron and proton: what happens at the electrons level would not happen if the
specific substrate structure were not there. Photosynthesis is an example of PCET,
with the transfer of 24 e and 24 H + driven by at least 48 photons.
State Vector Preparation

State vector preparation is key to experimental set-ups, and is a non-unitary process
because it can produce particles in a specific eigenstate from a stream of particles
that are not in such a state. Indeed it acts rather like state vector reduction, being a
transition that maps a mixed state to a pure state.
How can this non-unitary process happen in a way compatible with standard
unitary quantum dynamics? The crucial feature is pointed out by Isham [37]: the
outcome states are drawn from some collection Ei of initial states by being selected
by some suitable apparatus, being chosen to have some specic spin state in the SternGerlach experiment; the other states are discarded. This happens in two basic ways:
separation and selection (as in the Stern Gerlach experiment), which is unitary up to
the moment of selection when it is not, or selective absorption (as in the case of
wire polarisers), which continuously absorbs energy. This top-down effect from the
apparatus to the particles causes an effective non-unitary dynamics at the lower levels,
which therefore cannot be described by the Schrdinger or Dirac equations.
In such situations, selection takes place from a (statistical) ensemble of initial
states according to some higher level selection criterion, which is a form of topdown causation from the apparatus to the particles. This is a generic way one can
create order out of a disordered set of states, and so generate useful information by
throwing away what is meaningless. The apparatus is specifically designed to have
37
this non-unitary effect on the lower level [22]. This is the essential process not only
in state vector preparation but in all purification processes: which are the core of
chemistry and chemical engineering. they are the foundation of technology.
Room at the Bottom

Given this evidence for top-down causation, the physicist asks, how can there be room
at the bottom for top-down causation to take place? Isnt there over-determination
because the lower level physics interactions already determine what will happen from
the initial conditions?
There are various ways that top down causation can be effective, based in lower
level physical operations:
Setting Constraints on Lower Level Interactions

Structural constraints break symmetries, and so for example create more general
possibilities than are available to cellular automata. This has been explained above
in the context of digital computers. It is also the case where effective potentials occur
due to local matter structuring [23]. These are crucially dependent on the nature of
the higher level structure (e.g. transistors connected together in integrated circuits in
a computer, or networks of neurons connected by synapses in a brain) that cannot be
described in terms of lower level variables.
It is the higher level patterns that are the essential causal variable in solid state
physics by creating a specific band structure in solids (hence for example the search
for materials that will permit high temperature superconductivity). Which specific
lower level entities create them is irrelevant: you can move around specific protons
and electrons while leaving the band structure unchanged. This multiple realisability
is always the case with effective potentials.
Changing the Nature of the Constituent Entities

The billiard ball model of unchanging lower level entities underlying higher level
structure is wrong. Hydrogen in a water molecule has completely different properties
than when free; electrons bound in atom interact with radiation quite differently than
when free; neutrons bound in nucleus have a half life of billions of years but they decay
in 11 1/2 minutes when free. Assuming that the nature of an entity is characterised
by the way it interacts with others, in each case the higher level context has changed
the nature of the underlying entities (an effect that is commonplace in biology). In
the case of string theory, the nature of fundamental particles depends on the string
38
G. Ellis
theory vacuum a non-local higher context for their existence. Their properties are
not invariant, they depend on this vacuum.
Creating Constituent Entities

In many cases the lower level entities would not even exist without the higher
level structure. For example, the very possibility of existence of phonons is a
result of the physical structure of specific materials (section Superconductivity
and Superfluidity). This structure is at a higher level of description than that of
electrons. Top down effects clearly occur when lower level entities cannot exist outside their higher level context (again, a common effect in biology, where symbiosis
is rife).
This is however foundational in quantum field theory, where even the existence
of particles is context dependent;
One of the lessons learner from the development of the subject [quantum field theory] has
been the realisation that the particle concept does generally not have universal significance.
Particles may register their presence on some detectors but not others, so there is an essential
observer-dependent quality about them. One is still free to assert the presence of particles,
but without specifying the state of motion of the detector, the concept is not very useful.
([6, p. 49]).
Deleting Lower Level Entities

There is not a fixed set of lower level entities when selection creates order out of
disorder by deleting unwanted lower level entities or states: top-down action selects
what the lower elements will be. This selective top-down process is what underlies
state vector preparation in quantum physics [23, 37]. It is crucial in evolutionary
biology and in the mind [22]. It is the way order is created for chaos and information
is garnered from a jumble of incoming data.
Statistical Fluctuations and Quantum Uncertainty

Lower level physics is not determinate: random fluctuations and quantum indeterminism are both in evidence. What happens is not in the throes of iron determinism:
random events take place at the micro level. These can get amplified to macro scales.
In complex systems, this unpredictable variability can result in an ensemble of
lower level states from which a preferred outcome is selected according to higher
level selection criteria. Thus top-down selection leading to increased complexity [25]
is enabled by the randomness of lower level processes.
39
Overall, there are often not fixed lower level entities: their nature, or indeed their
existence, is dependent on their higher level context.
The Big Picture: The Nature of Causation

The view put here, in agreement with [64], is that top-down effects is the key to the
rise of genuine complexity (such as computers and human beings) in the universe.
Emergence
We need to explain genuine emergence, with all its complexity, and the reality of all
the levels:
We seek reality, but what is reality? The physiologists tell us that organisms are formed of
cells; the chemists add that cells themselves are formed of atoms. Does this mean that these
atoms or these cells constitute reality, or rather the sole reality? The way in which these cells
are rearranged and from which results the unity of the individual, is not it also a reality much
more interesting than that of the isolated elements, and should a naturalist who had never
studied the elephant except by means of the microscope think himself sufficiently acquainted
with that animal?Henri Poincare [57].
Each higher level is real with its own autonomy. This is in accord with the view put
by Denis Noble [51].
The Main Thesis

How does this come into being? The main thesis of this paper is
Hypothesis: bottom up emergence by itself is strictly limited in terms of the complexity it can
give rise to. Emergence of genuine complexity is characterised by a reversal of information
flow from bottom up to top down.
The degree of complexity that can arise by bottom-up causation alone is strictly
limited. Sand piles, the game of life, bird flocks, or any dynamics governed by a
local rule [3] do not compare in complexity with a single cell or an animal body.
Spontaneously broken symmetry is powerful [2], but not as powerful as symmetry
breaking that is guided top-down by adaptive selection to create ordered structures
(such as brains and computers).
Some kind of coordination of effects is needed for such complexity to emerge.
David Deutsch has a classic comment on the topic in his book The Fabric of
Reality [17]
40
G. Ellis
For example, consider one particular copper atom at the tip of the nose of the statue of Sir
Winston Churchill that stands in Parliament Square in London. Let me try to explain why
that copper atom is there. It is because Churchill served as prime minister in the House of
Commons nearby; and because his ideas and leadership contributed to the Allied victory
in the Second World War; and because it is customary to honor such people by putting up
statues of them; and because bronze, a traditional material for such statues, contains copper,
an so on. Thus we explain a low-level physical observation the presence of a copper atom
at a particular location through extremely high-level theories about emergent phenomena
such as ideas, leadership, war and tradition. There is no reason why there should exist, even
in principle, any lower-level explanation of the presence of that copper atom than the one I
have just given.
Another example is particle collisions at the LHC at CERN: these are the result of
the top down effect of abstract thoughts in the minds of experimenters to the particle
physics level. Without these thoughts, there would be no such collisions.
The Overall View

A key point I make in my essay is that top down effects dont occur via some mysterious non-physical downward force, they occur by higher level physical relations
setting constraints on lower level interactions, which not only can change the nature
of lower level entities (as in the case of the chameleon particles that might be the
nature of dark matter), they can even lead to the very existence of such entities (e.g.
phonons or Cooper pairs). So it is indeed a two-way causal flow which enables abstract entities to be causally effective (as in the case of digital computers) but does
not violate normal physics. It is a largely unrecognised aspect of normal physics.
There are some great discussions of the nature of emergent phenomena in physics
[2, 36, 42, 60] but none of them specifically mention the issue of top down causation.
This paper proposes that recognising this feature will make it easier to comprehend
the physical effects underlying emergence of genuine complexity, and may lead to
useful new developments, particularly to do with the foundational nature of quantum
theory.
A bottom-up view is taken as an underlying principle of faith by somehard core
reductionists, who simply ignore the contextual effects that in fact occur: for example
claiming that biology is controlled bottom up by genes alone, thereby ignoring all
the discoveries of epigenetics, which prove this false [34, 50]. But such reductionism
is always a cheat, because it is always only partial. An example is that Francis Crick
famously wrote [15]
You, your joys and your sorrows, your memories and your ambitions, your sense of personal
identity and free will, are in fact no more than the behavior of a vast assembly of nerve cells
and their associated molecules.
But nerve cells and molecules are made of electrons plus protons and neutrons, which
are themselves made of quarks so why not
41
You, your joys and your sorrows, your memories and your ambitions, your sense of personal
identity and free will, are in fact no more than the behavior of a vast assembly of quarks and
electrons
And these themselves are possibly vibrations of superstrings. Why does he stop
where he does?because thats the causal level he understands best!hes not a
particle physicist. But if he assumes that the level of cells and molecules is real, its
an arbitrary assumption unless all levels are realwhich is my position. Its the only
one that makes sense.
What I am pointing out in my essay is that physics does not by itself determine
what happens in the real world. Physics per se cannot account for the existence of
either a teapot or a Jumbo jet airliner, for example. At no point have I in any ways
denied that the laws of physics and chemistry apply at the lower levels. Of course
they do. The point is that they do not by themselves determine what happens. That
is determined by top down causation, as is abundantly clear for example in the case
of the computer. Some physical processes are not emergent but are entailed in a
top-down way. For example there is no bottom up process by which the computer
memory states embodying a Quicksort algorithm can emerge from the action of the
underlying physics acting in a purely bottom-up way. Indeed the same is true of the
processes leading to creation of a teacup or a pair of spectacles (see [21]).
If you believe this is wrong, please advise me of a physical law or process that
unambiguously determines how a tea cup can be created in a purely bottom-up way.
You will not be able to do soit does not exist.
Appendix: Equivalence Classes

The key idea is that of functional equivalence classes. Whenever you can identify
existence of such equivalence classes, that is an indication that top-down causation
is taking place [4]. Indeed this is essentially the ontological nature of the higher level
effective entity: a computer program is in its nature the same as the set of all possible
implementations of the set of logical operations it entails. These are what enter into
the higher level effective relations; they can be described in many different ways,
and implemented in many different ways; what remains the same in those variants
is the core nature of the entity itself.
An equivalence relation is a binary relation satisfying three properties:
1. For every element a in X, a a (reflexivity),
2. For every two elements a and b in X, if a b, then b a (symmetry), and
3. For every three elements a, b, and c in X, if a b and b c, then a c
(transitivity).
The equivalence class of an element a is denoted [a] and may be defined as the set
of elements that are related to a by .
42
G. Ellis
??
H1
H '2
H '1
L1
H '3
L' 1
L2
L' 2
L3
L' 3
Fig. 3.1 The lower level dynamics does not lead to coherent higher level dynamics when the lower
level dynamics acting on different lower level states corresponding to a single higher level state,
give new lower level states corresponding to different higher level states
H1
H '1
L1
L' 1
L2
L' 2
L3
L' 3
Fig. 3.2 The lower level dynamics leads to coherent higher level dynamics when the lower level
dynamics acting on different lower level states corresponding to a single higher level state, give
new lower level states corresponding to the same higher level state
The way this works is illustrated in (Figs. 3.1 and 3.2) higher level states H1 can
be realised via various slower level states L1. This may or may not result in coherent
higher level action arising out of the lower level dynamics.
When coherent dynamics emerges, the set of all lower states corresponding to a
single higher level state form an equivalence class as far as the higher level actions
are concerned. They are indeed the effective variables that matter, rather than the
specific lower level state that instantiates the higher level one. This is why lower
level equivalence classes are the key to understanding the dynamics (Figs. 3.1).
References
1. D.Z. Albert, Time and Chance (Harvard University Press, Cambridge, 2000)
2. P.W. Anderson, More is different. Science 177, 393396 (1972)
3. P. Arrighi, J. Grattange, The quantum game of life. Phys. World (2012), http://physicsworld.
com/cws/article/indepth/2012/may/31/the-quantum-game-of-life
4. G. Auletta, G.F.R. Ellis, L. Jaeger, Top-down causation by information control: from a philosophical problem to a scientific research program. J. R. Soc. Interface 5, 11591172 (2008)
43
5. P.L. Berger, T. Luckmann, The Social Construction of Reality: A Treatise in the Sociology of
Knowledge (Penguin, London, 1991)
6. N.D. Birrell, P.C.W. Davies, Quantum Fields in Curved Space (Cambridge University Press,
Cambridge, 1982)
7. H. Bondi, Cosmology (Cambridge University Press, Cambridge, 1960)
8. G. Booch, Object-Oriented Analysis and Design with Applications (Addison-Wesley Professional, 1993)
9. H.P. Breuer, F. Petruccione, The Theory of Open Quantum Systems (Clarendon Press, Oxford,
2006)
10. C. Callender, Thermodynamic asymmetry in time (2011), http://plato.stanford.edu/entries/
time-thermo/
11. N.A. Campbell, J.B. Reece, Biology (Benjamin Cummings, San Francisco, 2005)
12. S. Carroll, From Eternity to Here: The Quest for the Ultimate Arrow of Time (Dutton, New
York, 2010)
13. J.Y. Chiao, Cultural neuroscience. Prog. Brain Res. 178, 287304 (2009)
14. B.J. Copeland, The Essential Turing (Clarendon Press, Oxford, 2004)
15. F. Crick, Astonishing Hypothesis: The Scientific Search for the Soul (Scribner, New York, 1995)
16. J.P. Crutchfield, Between order and chaos. Nat. Phys. 8, 1724 (2011)
17. D. Deutsch, The Fabric of Reality (Penguin Books, London, 1998)
18. S. Dodelson, Modern Cosmology (Academic Press, New York, 2003)
19. A.S. Eddington, The Nature of the Physical World (Cambridge University Press, Cambridge,
1928)
20. G.F.R. Ellis, Cosmology and local physics. Int. J. Mod. Phys. A17, 26672672 (2002)
21. G.F.R. Ellis, Physics, complexity and causality. Nature 435, 743 (2005)
22. G.F.R. Ellis, Top down causation and emergence: some comments on mechanisms. J. Roy Soc
Interface Focus 2, 126140 (2012)
23. G.F.R. Ellis, On the limits of quantum theory: contextuality and the quantumclassical cut. Ann.
Phys. 327, 18901932 (2012)
24. G.F.R. Ellis, Multi-level selection in biology (2013), arXiv:1303.4001
25. G.F.R. Ellis, Necessity, purpose, and chance: the role of randomness and indeterminism in
nature from complex macroscopic systems to relativistic cosmology (2014). Available at http://
www.mth.uct.ac.za/~ellis/George_Ellis_Randomness.pdf
26. G.F.R. Ellis, D. Ellis, D. Noble, T. OConnor (eds.), Top-down causation. Interface Focus 2(1)
(2012)
27. D.J. Felleman, D.C. Van Essen, Distributed hierarchical processing in the primate cerebral
cortex. Cereb. Cortex 1(1), 147 (1991)
28. R.P. Feynman, R.B. Leighton, M. Sands, The Feynman Lectures on Physics: Volume III. Quantum Mechanics (Addison-Wesley, Reading, 1965)
29. R.P. Feynman, R.B. Leighton, M. Sands, Feynman Lectures in Physics Volume II: Electromagnetism (Addison-Wesley, Reading, 1964)
30. V. Frankl, Mans Search for Meaning (Rider and Co., London, 2008)
31. K. Friston, A theory of cortical responses. Philos. Trans. R. Soc. (Lond.) B Biol. Sci. 36,
360815 (2005)
32. C. Frith, Making Up the Mind: How the Brain Creates Our Mental World (Blackwell, Malden,
2007)
33. C.D. Gilbert, M. Sigman, Brain states: top-down influences in sensory processing. Neuron 54,
677696 (2007)
34. S. Gilbert, D. Epel, Ecological Developmental Biology (Sinauer, Sunderland, 2009)
35. G. Greenstein, A.G. Zajonc, The Quantum Challenge: Modern Research on the Foundations
of Quantum Mechanics (Jones and Bartlett, Sudbury, Mass, 2006)
36. S. Hartmann, Effective field theories, reductionism and scientific explanation. Stud. Hist. Philos.
Sci. Part B 32, 267304 (2001)
37. C.J. Isham, Lectures on Quantum Theory (Imperial College Press, London, 1995)
38. M. Kac, Can one hear the shape of a drum? Am. Math. Mon. 73, 123 (1966)
44
G. Ellis
39. E.R. Kandel, The Age of Insight (Random House, New York, 2012)
40. J. Kurose, Computer Networking: A Top-Down Approach (Addison Wesley, Upper Saddle
River, 2010)
41. R. Lafore, Data Structures and Algorithms in Java (SAMS, Indiana, 2002)
42. R.W. Laughlin, Fractional quantisation. Rev. Mod. Phys. 71, 863874 (1999)
43. J. MacCormick, 9 Algorithms that Changed the Future (Princeton University Press, New Jersey,
2012)
44. E.Z. Macosko, N. Pokala, E.H. Feinberg, S.H. Chalasani, R.A. Butcher, J. Clardy, C.I.
Bargmann, A hub-and-spoke circuit drives pheromone attraction and social behaviour in C.
elegans. Nature 458, 11711175 (2009)
45. M.M. Mano, C.R. Kine, Logic and Computer Design Fundamentals (Pearson, 2008)
46. V. Mante, D. Sussillo, K.V. Shenoy, Context-dependent computation by recurrent dynamics in
prefrontal cortex. Nature 503, 7884 (2013)
47. M. Martnez, A. Moya, (2011) Natural selection and multi-level causation. Philos. Theory Biol.
(2011), http://dx.doi.org/10.3998/ptb.6959004.0003.002. 3 May 2011
48. C. Mitchell, J. Hobcraft, S.S. McLanahan, S.R. Siegel, A. Berg, J. Brooks-Gunn, I. Garfinkel,
D. Nottermand, Social disadvantage, genetic sensitivity, and childrens telomere length. Proc.
Nati. Acad. Sci. 111(16), 59445949 (2014)
49. S. Mukhanov, Physical Foundations of Cosmology (Cambridge University Press, Cambridge,
2005)
50. D. Noble, From the HodgkinHuxley axon to the virtual heart. J. Physiol. 580, 1522 (2007)
51. D. Noble, A theory of biological relativity: no privileged level of causation. Interface Focus 2,
5564 (2012)
52. M.J. Page et al., The suppression of star formation by active galactic nuclei. Nature 485,
213216 (2012)
53. J. Pearl, Causal diagrams for empirical research. Biometrika 82, 669688 (1995)
54. J. Pearl, Causality: Models, Reasoning and Inference (MIT Press, 2000)
55. R. Penrose, Cycles of Time: An Extraordinary New View of the Universe (Knopf, New York,
2011)
56. M. Pigliucci, An extended synthesis for evolutionary biology. Ann. N. Y. Acad. Sci. 1168,
218228 (2009)
57. H. Poincare, The Value of Science. Originally published in 1913 (trans: G.B. Halstead), Cosimo
Classics, 21(1913). ISBN: 978-1-60206-504-8
58. D. Purves, W. Wojtach, W.C. Howe, Visual ilusions. Scholarpedia (2007)
59. D. Purves, Brains: How They Seem to Work (FT Press Science, Upper Saddle River, 2010)
60. S. Schweber, Physics, community, and the crisis in physical theory. Phys. Today 46, 3440
(1993)
61. M. Sigman, H. Pan, Y. Yang, E. Stern, D. Silbersweig, C.D. Gilbert, Top-down reorganization
of activity in the visual pathway after learning a shape identification task. Neuron 46, 823835
(2005)
62. A.S. Tanenbaum, Structured Computer Organisation (Prentice Hall, Englewood Cliffs, 1990)
63. R. van den Bos, J.W. Jolles, J.R. Homberg, Social modulation of decision-making: a crossspecies review. Front. Hum. Neurosci. 7, 301 (2013). www.frontiersin.org
64. S.I. Walker, L. Cisneros, P.C.W, Davies, Evolutionary transitions and top-down causation, in
Proceedings of Artificial Life XIII (2012), http://xxx.lanl.gov/abs/1207.4808
65. D.R. Weinberg, C.J. Gagliardi, J.F. Hull, C.F. Murphy, C.A. Kent, B. Westlake, A. Paul, D.H.
Ess, D.G. McCafferty, T.J. Meyer, Proton-coupled electron transfer. Chem. Rev. 107, 5004
5064 (2007)
66. S. Weinberg, The Quantum Theory of Fields. Volume 1: Foundations (Cambridge University
Press, Cambridge, 2005)
67. J.M. Ziman, Principles of the Theory of Solids (Cambridge University Press, Cambridge, 1979)
Chapter 4
On the Foundational Assumptions

of Modern Physics
Benjamin F. Dribus
Abstract General relativity and the standard model of particle physics remain our
most fundamental physical theories enjoying robust experimental confirmation. The
foundational assumptions of physics changed rapidly during the early development of
these theories, but the subsequent challenges of their refinement and the exploitation
of their explanatory power turned attention away from foundational issues. Deep
problems and anomalous observations remain unaddressed. New theories such as
string theory seek to resolve these issues, but are presently untested. In this essay, I
evaluate the foundational assumptions of modern physics and propose new physical
principles. I reject the notion that spacetime is a manifold, the existence of static
background structure in the universe, the symmetry interpretation of covariance,
and a number of related assumptions. The central new principle I propose is the
causal metric hypothesis, which characterizes the observed properties of the physical
universe as manifestations of causal structure. More precisely, the classical causal
metric hypothesis states that the metric properties of classical spacetime arise from
a binary relation on a set, representing direct influences between pairs of events.
Rafael Sorkins maxim, order plus number equals geometry is a special case. The
quantum causal metric hypothesis states that the phases associated with directed
paths in causal configuration space, under Feynmans sum-over-histories approach to
quantum theory, are determined by the causal structures of their constituent universes.
The resulting approach to fundamental physics is called quantum causal theory.
Introduction
Relativity and Quantum Theory. Relativity and quantum theory emerged from
mathematical and philosophical seeds in the works of Gauss, Riemann, Cayley,
Hilbert, and others; were incorporated as physical theories by Einstein, Heisenberg,
Schrdinger, Weyl, and their contemporaries; and matured as definitive predictive
systems in the form of modern general relativity and the standard model of particle
B.F. Dribus (B)
William Carey University, Hattiesburg, USA
e-mail: [email protected]; [email protected]
45
46
B.F. Dribus
physics in the second half of the twentieth century. Among theories enjoying robust
experimental confirmation, these two theories represent our deepest understanding of
fundamental physics. The rapid alteration of foundational assumptions characterizing
the early development of these theories later diminished as their fruit was harvested.
Satisfactory unification of relativity and quantum theory proved to be an immense and
forbidding challenge, resisting numerous optimistic early attempts, and an abundance
of new experimental results amenable to description within the developing framework
of quantum field theory decreased motivation for radical new departures.
Foundational Problems; New Theories. Recently the triumphs of quantum field
theory have slowed, and unexplained phenomena such as dark matter and dark energy
hint at new physics. In this environment, long-acknowledged foundational problems
have gained new urgency. The fundamental structure of spacetime, the nature and
significance of causality, the quantum-theoretic description of gravity, and unified
understanding of physical law, have all attracted increased scrutiny. Untested new
theories seek to address these issues, often incorporating new assumptions as alien to
established physics as the assumptions of relativity and quantum theory were to the
Newtonian paradigm. Among these new theories, string theory [1] abolishes point
particles and introduces new dimensions, symmetries, and dualities; loop quantum
gravity [2] undertakes the quantization of relativistic spacetime; noncommutative
geometry [3] interprets spacetime as a noncommutative space; entropic gravity [4]
attributes gravitation to the second law of thermodynamics; and causal set theory
[5] discards manifold models of classical spacetime in favor of discrete partially
ordered sets. While limited, this list represents a reasonable cross-section of the
general approaches to new physics under active investigation.
Overview and Organization of This Essay. In this essay, I evaluate the foundational assumptions of modern physics and offer speculative new principles, partially
overlapping some of the new theories mentioned above. These principles cannot, to
my present knowledge, claim definitive experimental confirmation, but their consideration is reasonable alongside other untested theories. Among the assumptions
I reject are the manifold structure of spacetime, the evolution of physical systems
with respect to a universal time parameter, the existence of a static background structure serving as an arena for dynamical processes, the symmetry interpretation of
covariance, the transitivity of the binary relation encoding causal structure, and the
commutativity of spacetime. The central new principle I propose is the causal metric
hypothesis, which characterizes the observed properties of the physical universe as
manifestations of causal structure. For purposes of precision, it is convenient to formulate classical and quantum versions of the causal metric hypothesis. The classical
version states that the properties of classical spacetime are manifestations of a binary
relation on a set. Rafael Sorkins maxim, order plus number equals geometry, is
a special case. The quantum version states that the phases associated with directed
paths in causal configuration space are determined by the causal structures of their
constituent universes. These ideas are explained in more detail below. The resulting
approach to fundamental physics is called quantum causal theory.
This essay is organized as follows: in the section Identifying the Foundational
Assumptions, I identify and discuss the foundational assumptions of modern
4 On the Foundational Assumptions of Modern Physics
47
physics, focusing on assumptions enjoying wide recognition in the mainstream

scientific community. I introduce three different classes of assumptions: general
principles, formal postulates, and ancillary assumptions. I isolate six general principles of particular importance, and briefly cite a few others worthy of mention.
I then discuss assumptions specific to relativity, nonrelativistic quantum theory,
and quantum field theory. In the section Vignette of Unexplained Phenomena, I
briefly mention some empirical phenomena unexplained by these theories. In the
section Rejected Assumptions, I reject several existing assumptions, with motivation provided by the previous two sections. In the section New Principles, I
propose new physical principles, with particular focus on the fundamental structure
of spacetime. In the section Practical Considerations, I remark on the status of these
assumptions and principles in light of the current state of experimental and theoretical physics, and suggest how the ideas presented in this essay might find their way
into the laboratory.
Identifying the Foundational Assumptions

Three Classes of Assumptions. In the world of scientific thought, ideas from mathematics, philosophy, and the empirical realm converge in the form of general physical
principles, which further crystallize into formal postulates of specific physical theories, while remaining colored and often distorted by ancillary assumptions involving
issues of interpretation and biases from the prevailing intellectual environment. These
principles, postulates, and ancilla are all foundational assumptions in the sense that
basic science depends critically, and to some degree independently, on each. However, they also possess important distinguishing characteristics.
General physical principles represent attempts to capture deep physical truths
that are often difficult to quantify. As a result, such principles often survive, via an
evolutionary process of refinement and reinterpretation, through multiple scientific
revolutions, while formal postulates and ancillary assumptions often die along with
the specific theories built around them. For example, the general principle of covariance, which at its root involves an assertion of the observer-independence of physical
law, has motivated a succession of mutually contradictory formal invariance postulates, such as Galilean invariance and Lorentz invariance, along the historical path
from Newtonian physics through special relativity to general relativity and beyond.
Parallel to these invariance postulates have followed a succession of mutually contradictory ancillary assumptions regarding the interpretation of time and related issues.
Covariance itself, meanwhile, remains relevant even to nonmanifold models of spacetime.
Formal postulates of dubious aspect sometimes persist due to a lack of suitable
alternatives, even when they contradict widely acknowledged general principles.
For example, the general principle of background independence is usually taken for
granted, at a philosophical level, in the modern physics community; yet the formal postulates underlying the standard model of particle physics, as well as many
48
B.F. Dribus
newer theories, fail to satisfy this principle. Ancillary assumptions can be particularly
troublesome because of their tendency to escape serious scrutiny. Examples include
the luminiferous aether in pre-relativistic physics, and some of the assumptions
related to Bells inequalities in the foundations of quantum theory. General principles and formal postulates are safer in this regard, since they attract the conscious
focus of theorists.
Six General Physical Principles. Six crucial general principles of modern physics
are symmetry, conservation, covariance, the second law of thermodynamics, background independence, and causality. These principles are intimately interrelated.
Results such as Noethers theorem tie symmetries to conservation laws, and relativistic covariance is understood in terms of symmetry, at least locally. More generally,
covariance may be interpreted in terms of generalized order theory. Both viewpoints
involve isolating privileged information; either that fixed by a particular group action,
or that contained in a distinguished suborder. Entropy, and thence the second law of
thermodynamics, may also be expressed via partitioning of information: in ordinary
statistical thermodynamics, entropy involves microscopic refinements of macroscopic states; while in discrete causal theory, it may be measured in terms of the
cardinality of certain Galois groups of generalized order morphisms.
Background independence is usually understood as a statement about spacetime;
that it is a dynamical entity subject to physical laws, such as Einsteins field equations, rather than a static object. Philosophically, background independence provides
an example of the use of parsimony to achieve explanatory and predictive power;
the less a theory assumes, the more it can potentially reveal. Background independence is one of the strengths of general relativity; relativistic spacetime geometry
is determined via dynamics, not taken for granted. Improvement beyond relativity
is conceivable. For example, Einsteins equations do not predict the dimension of
spacetime; a theory providing a dynamical explanation of dimension would be superior in important ways. Causality is of central importance to physics, and to science
in general, principally because prediction relies upon the discovery of causal relationships, together with the assumption of reproducibility. Classically, causality is
often formalized as an irreflexive, acyclic, transitive binary relation on the set of
spacetime events. It is related to covariance via order theory, to the second law of
thermodynamics via the arrow of time, and to background independence via the
general criteria of explanatory and predictive power. However, the deep meaning of
causality, and its appropriate role in quantum theory, remain controversial.
Other General Principles. Other general principles deserving mention include
symmetry breaking, physical versions of superposition including Feynmans sum
over histories, action principles, cluster decomposition and other versions of locality, Einsteins equivalence principle, scale-dependence and independence, the holographic principle, dualities such as S-duality, and various principles involved in the
interpretation of quantum theory. Untested modern theories rely on further principles, or refinements of principles already mentioned, whose importance is tied to
their success. For example, Maldacenas AdS/CFT correspondence [6] is much more
important if string theory is physically relevant than it would be otherwise. Pure
mathematics, such as number theory, also offers general principles, and conjectured
49
principles, with deep connections to physics. For example, zeta functions, and hence
the Riemann hypothesis, are connected to quantum field theory via noncommutative
geometry and the theory of motives [7]. The Langlands program is connected to
physical symmetry and duality via representation theory and conformal field theory,
and thence also to string theory [8].
Assumptions of Relativity and Quantum Theory. The following formal postulates and ancillary assumptions apply to general relativity and quantum theory,
although some of them also survive in newer theories. General relativity postulates
a four-dimensional pseudo-Riemannian manifold of Lorentz signature, interpreted
as spacetime, whose curvature, interpreted as gravitation, is determined dynamically
via interaction with matter and energy according to Einsteins field equations, and
whose metric properties govern its causal structure. Singularities arise in the generic
case, as noted by Penrose, Hawking and Ellis, and others.
Multiple approaches to nonrelativistic quantum theory exist. I will describe two,
equivalent under suitable restrictions. The Hilbert space approach postulates complex Hilbert spaces whose elements represent probability amplitudes, self-adjoint
operators whose eigenvalues represent the possible values of measurements, and
time evolution according to Schrdingers equation. In the simplest context, these
amplitudes, operators, et cetera, represent the behavior of point particles. Feynmans
sum-over-histories approach [9] postulates probability amplitudes given by complex
sums over spaces of paths, interpreted as spacetime trajectories of point particles in
the simplest context. In a path sum, each path contributes equally in magnitude,
with phase determined by the classical action, given by integrating the Lagrangian
along the path with respect to time. This version generalizes easily to relativistic and
post-relativistic contexts.
Quantum field theory postulates operator fields that create and annihilate state
vectors in complex Hilbert spaces. States corresponding to particular particle species
are associated with particular representations of symmetry groups. The properties of
Minkowski spacetime impose external symmetries encoded by the Poincar group.
Internal symmetries, such as those encoded by gauge groups, also play a critical role.
The standard model of particle physics is expressed via the nonabelian Yang-Mills
gauge theory, and includes particles, fields, and symmetry groups in remarkable
accord with the observations of particle physicists over the last century.
Vignette of Unexplained Phenomena

Overview. Since the ascendancy of general relativity and the standard model, a
variety of unexplained physical phenomena have been recognized. The large-scale
dynamical anomalies attributed to dark matter and dark energy, the absence of a large
cosmological constant arising from vacuum energy, and the apparent asymmetry
between matter and antimatter in the observable universe, are a few of the most prominent examples. These phenomena suggest the promise of physical models that naturally incorporate scale-dependence, and that offer statistical or entropic explanations
50
B.F. Dribus
of small nonzero constants and inexact symmetries. Discrete order-theoretic and

graph-theoretic models tend to perform well by these criteria.
Dark Matter. The dark matter hypothesis is based on the failure of astrophysical
systems on the scale of galaxies to obey relativistic dynamics, assuming only the
matter content detectable by non-gravitational means. In contrast, objects near the
stellar scale seem to verify relativistic predictions remarkably well. The dark matter
hypothesis has been compared unfavorably to the luminiferous aether, and various
new dynamical laws have been proposed to account for observed behavior without
invoking missing mass. However, this phenomenon does behave like ordinary matter
in many respects, as observed in the collision of galaxies and in certain examples
of gravitational lensing. If the dark matter hypothesis is valid, the matter involved
seems unlikely to be accounted for by the standard model. Claims have been made
of laboratory observations of new particles consistent with dark matter, but these are
not broadly accepted at present.
Dark Energy. Dark energy is the entity invoked to explain the phenomenon interpreted as acceleration of the expansion of the universe. The cosmological constant
appearing in the modified form of Einsteins equations is one possible type of dark
energy. Predictions based on quantum field theory and the Planck scale yield a value
for the cosmological constant roughly 120 orders of magnitude greater than observation implies. Interestingly, causal set theory suggests a fluctuating cosmological constant close to the observed value, based on a simple argument involving discreteness
and the size of the Hubble radius in Planck units. Nonconstant models of dark energy,
such as quintessence, have also been proposed, but any fluctuations in dark energy
appear to occur on scales much larger than those of dark matter or ordinary matter
and energy. Apparent anomalies in the motion of certain large galactic clusters, called
dark flow, might reflect such fluctuations. Dark matter and dark energy extend the
scale-dependence of phenomena already observed in conventional physics. Strong
and weak interactions, electromagnetism, ordinary gravity, dark matter, and dark
energy all dominate on different scales, each covering roughly equivalent ranges
in a logarithmic sense. The extent of this scale-dependence was unknown during
the development of relativity and quantum theory, and should command significant
attention in the development of new models.
Matter-Antimatter Asymmetry. Our present understanding of antimatter comes
almost entirely from quantum field theory, and it is reasonable to ask if matterantimatter asymmetry in the observable universe might indicate a problem with
quantum field theory itself, or at least with the standard model. Unexpected asymmetries have been successfully handled by quantum field theory in the past; the
prototypical example is CP violation, which is itself related to the matter-antimatter
problem. However, potential sources of matter-antimatter asymmetry in the standard
model seem either too weak, or too strong, to account for observation. Interesting experimental issues regarding antimatter remain to be resolved. Until recently,
little direct evidence existed to demonstrate that antimatter interacts gravitationally in the same way as matter, and it had even been suggested that local matterantimatter asymmetry might result from a type of gravitational segregation. More
conventionally, experiments designed to investigate matter-antimatter asymmetry
51
have recently produced data suggesting rates for certain decay processes different
than those predicted by the standard model. It seems too early to render judgment
on the significance or meaning of these results, however.
Rejected Assumptions
Structural Assumptions; Metric Emergence. Some of the physical assumptions I
reject in this essay are already widely doubted, but survive in old and new theories
alike due to the unfamiliarity or intractability of their principal alternatives. Among
these are the basic structural assumptions that spacetime is a real manifold, that
physical systems evolve with respect to a universal time parameter, and that the
universe possesses a static background structure serving as an immutable arena
for dynamical processes. This last assumption is, of course, merely the negation of the
general principle of background independence. General relativity includes the first of
these assumptions, and the standard model includes all three. Since these assumptions
are retained largely for operational reasons, their rejection is not very revolutionary.
However, a successful theory abstaining from them would be revolutionary indeed. I
reject them partly on general mathematical and philosophical grounds, and partly for
the specific physical reason that they are incompatible with discrete quantum causal
theory.
Another basic structural assumption I reject is that spacetime is commutative. This
statement should be understood in the sense of Connes noncommutative geometry
[3]. Though this assumption is less-widely doubted in mainstream physics than those
mentioned above, it has recently become the subject of justified scrutiny. A number of existing proposals about fundamental spacetime structure lead naturally to
noncommutative spaces. For example, such spaces arise via the deformation theory of Hopf algebras, and in certain category-theoretic approaches to physics. Even
classical spaces such as Minkowski spacetime may be recognized as possessing
noncommutative structures in useful ways.
Along with these assumptions perish a number of corollaries. Spacetime dimension
becomes an emergent property, no longer assumed to be constant, static, or an integer. Properties previously ascribed to a metric, in the sense of differential geometry,
must either be discarded or assigned different structural origins. For example, given
a geodesic between two events in relativistic spacetime, there exist many other neargeodesics between them; however, a nonmanifold model of spacetime might admit a
unique short path between two events, with every other path being much longer.
Such reflections prompt reconsideration of the notions of distance and locality. Other
metric properties could be similarly reexamined, but most important is to investigate
what mechanisms supply the appearance of a metric at ordinary scales. This may be
called the problem of metric emergence.
Assumptions About Causality. The answer I will propose to the problem of metric emergence involves reinterpreting the general physical principles of causality
and covariance. This requires rejection of some common ancillary assumptions
52
B.F. Dribus
about these specific principles. First, I reject the assumption that the apparent metric
properties of classical spacetime involve any information other than a set of events
and a binary relation, the causal relation, encoding causal structure. This rejection
amounts to a negative version of the classical causal metric hypothesis; the corresponding positive version is stated below. Theorems of Stephen Hawking [10] and
David Malament [11] in the late 1970s hinted at this conclusion in a relativistic context, by demonstrating that most of the metric properties of relativistic spacetime
may be recovered from its causal structure. Causal set theory already incorporates a
version of this idea.
Second, I reject the assumption that the causal relation is transitive. This oddseeming statement merely acknowledges the physical relevance of information about
direct versus indirect causation. The usual transitive causal order may be recovered
by closing the causal relation under transitivity. Third, I reject the assumption that
the causal relation is acyclic. This rejection permits the existence of causal cycles,
which already arise as closed timelike curves in certain solutions of general relativity.
Causal cycles need raise no paradoxes; if they exist, they are properties of a binary
relation, not self-contradictory inhabitants of a background structure.
Assumptions About Covariance. Turning to covariance, I reject the assumption
that it is an instance of group symmetry, even locally; rather, it should be viewed in
order-theoretic terms. For example, different frames of reference in relativity assign
different time-orders to events separated by spacelike intervals; these orders correspond to different classes of refinements of the causal relation. This rejection is
notable because progress in physics has historically involved invoking new symmetry
principles, rather than rejecting existing ones. Since the time of Weyl, group representation theory has permeated theoretical physics as the mathematical expression
of symmetry, and remains perhaps the most promising technical vehicle for shortterm progress beyond the standard model. Over the long term, however, analogous
constructs from order theory, and perhaps other notions more primitive than groups,
will likely replace much of group representation theory in this role. Alternative
approaches to covariance involving category theory and noncommutative geometry
have already been proposed.
New Principles
Overview: Quantum Causal Theory. New principles I propose in this essay include
the causal metric hypothesis, iteration of structure as a quantization principle, and
co-relative histories. These principles, explained in more detail below, form the backbone of quantum causal theory, which is a general term I use to describe approaches
to quantum spacetime and quantum gravity that take causal structure to be fundamental. Technical tools necessary to implement these ideas include a synthesis of
multicategory theory and categorification in abstract algebra, involving interchangeability of objects, morphisms, elements, and relations; a refined version of random
graph dynamics; and the theory of semicategory algebras. In particular, path alge-
53
bras encode the properties of both individual causal universes and their configuration
spaces, while providing convenient methods of computation. Details of many of these
ideas appear in my paper [12]. Here I focus only on the basic concepts.
Causal Metric Hypothesis. Foremost among the new principles I propose is the
causal metric hypothesis. The philosophical content of this hypothesis is that the
observed properties of the physical universe are manifestations of causal structure.
To crystallize this idea into a precise, quantitative approach to physics, it is convenient
to first state a classical version of the hypothesis, which serves as a precursor to the
corresponding quantum version, just as classical notions form the building blocks
of quantum theory in Feynmans sum-over-histories approach. The classical causal
metric hypothesis may be stated as follows:
The properties of classical spacetime arise from a binary relation on a set S, where elements
of S represent spacetime events, and elements of represent direct influences; i.e., causal
relations, between pairs of events.
Figure 4.1 illustrates the classical causal metric hypothesis, and demonstrates how
it differs from the paradigm of general relativity. Figure 4.1a shows a region of relativistic spacetime, with distinguished events marked by nodes. In general relativity,
the geometry of spacetime governs the scope of causal influence. For example, event
x may have been influenced by all events in its geometric past, shown in dark
gray, and may influence all events in its geometric future, shown in light gray. The
classical causal metric hypothesis turns this picture on its head, taking spacetime
geometry to be nothing more than a way of describing actual influences. Figure 4.1b
shows a family of events, with direct influences indicated by edges running up the
page. Under the classical causal metric hypothesis, the geometric past and future
are a posteriori constructions. Rafael Sorkins causal set maxim, order plus number
equals geometry, is a special case of the classical causal metric hypothesis.
(a)
(b)
x
w
x
w
BDCT
TGGA
Fig. 4.1 a In general relativity, spacetime geometry governs the scope of causal influence; b under
the classical causal metric hypothesis, spacetime geometry is merely a way of describing actual
influences
54
B.F. Dribus
The Causal Relation. The binary relation referenced in the classical causal metric
hypothesis is a mathematical way of encoding direct influences between pairs of
events, represented by edges in Fig. 4.1b. Such a relation, which I will call the causal
relation in this context, may be viewed as a generalized partial order, with the word
order indicating precedence and succession. For example, event x in Fig. 4.1b precedes event y; this is written x y. In Sorkins causal set theory, the causal relation is
a partial order in the technical sense, but there are good reasons to generalize this picture; for example, by abstaining from transitivity and acyclicity, as already indicated
above. However, the most interesting versions of causal theory I know of do impose
reasonable assumptions on the causal relation; for example, local finiteness. More
generally, assumptions about local structure are usually more reasonable to impose
than their nonlocal counterparts, due to our ignorance of the global structure of the
universe.
Recovery of Lorentzian manifold structure from the causal relation is necessary
at some level of approximation, owing to the large-scale success of general relativity.
The metric recovery theorems of Hawking and Malament, mentioned above, demonstrate that specifying appropriate volume data, as well as order data, is sufficient to
recover continuum geometry. According to the classical causal metric hypothesis,
this volume data should derive in some way from the pair (S, ). The simplest dependence is the trivial one, in which a single unit of volume is assigned to each element
of S, irrespective of ; this is the causal set approach, as encapsulated by Sorkin.
However, the causal metric hypothesis allows for alternative methods of specifying
volume data that depend on the causal relation in more complicated ways.
Iteration of Structure as a Quantization Principle. Feynmans sum-over-histories
approach to quantum theory [9] is perhaps the most promising general approach
under the causal metric hypothesis. Significant efforts have already been made to
adapt this approach to causal set theory, although technical problems such as the
permeability of maximal antichains complicate the picture. For this reason, and many
others, it is preferable to work in relation space, as described in section 5 of my paper
On the Axioms of Causal Set Theory [12]. Sums in this context involve paths in a
configuration space of classical universes, each represented by a pair (S, ). I refer
to such a space as a causal configuration space. For example, the causal configuration
space of causal set theory is the space of all acyclic, transitive, interval-finite universes
admitting an order embedding into the natural numbers. Causal configuration space
inherits a directed structure induced by special morphisms between pairs of universes,
called transitions. This directed structure may be viewed as a higher-level analogue
of the directed structures on the individual universes (S, ), encoded by the causal
relations . This emergence of higher-level directed structure on causal configuration
space is a prototypical example of a recurring principle in quantum causal theory
that I refer to as iteration of structure. In particular, quantization consists of passage
from individual universes to causal configuration space. Mathematically, this may
be viewed in terms of a generalized version of categorification/decategorification,
in which structure is added or ignored by promoting elements or demoting objects.
Co-relative Histories; Kinematic Schemes. For technical reasons, transitions are
too specific to be physically fundamental; they carry gauge-like information.
55
BDCT
TGGA
Fig. 4.2 A co-relative history. Gray indicates new structure in the target universe
Appropriate equivalence classes of transitions, which I call co-relative histories, are

the physically significant building blocks of higher-level structure in causal configuration space, providing a refined version of iteration of structure. Figure 4.2 illustrates
a co-relative history.
Co-relative histories replace the notion of time evolution in quantum causal theory.
The target universe of a co-relative history may be viewed as a later stage of development of its source universe. A suitable choice of co-relative histories, providing
evolutionary pathways for every possible universe, yields a special substructure of
causal configuration space that I call a kinematic scheme.
Figure 4.3 modified from a similar figure in my paper [12], shows a portion of a
kinematic scheme S that I refer to as the positive sequential kinematic scheme. The
word sequential means that each co-relative history in S adds a single element to
its source universe. The word positive means that the elements of each universe in
S may be labeled by positive integers. The generations indicated by the large numbers 0, 1, 2, 3, 4 in Fig. 4.3 correspond to such labeling. The inset in Fig. 4.3 shows
a portion of the abstract underlying directed graph corresponding to S; comparison
of this graph to S itself illustrates iteration of structure. Each upward-directed path
in S represents a kinematic account of the evolution of its terminal universe. The
path terminating at the universe U in Fig. 4.3 is an example. Note that the spacelike
hypersurface; i.e., maximal antichain, in S, represented by the three universes in
double circles, is permeated by the path from to U . This indicates that the relation
space over S; i.e., the corresponding space of co-relative histories, provides a superior viewpoint in a structural sense. This is an example of the relative viewpoint
advocated by Alexander Grothendieck, in which one studies relationships between
mathematical objects, rather than studying each object individually. Gray indicates
universes whose causal relations are intransitive. These universes distinguish S from
the configuration spaces arising in causal set theory [13].
The theory of kinematic schemes provides a precise realization of the principle
advocated by Robert Spekkens in his essay The paradigm of kinematics and dynamics
must yield to causal structure, also appearing Chap. 2. Different kinematic schemes
lead to different dynamical equations, all equally valid. For example, kinematic
56
B.F. Dribus
U
1 0
BDCT
TGGA
Fig. 4.3 Portion of the positive sequential kinematic scheme S ; inset shows the underlying directed
structure; large-font numbers indicate generations; double circles represent a maximal antichain;
dark path represents a permeating chain; gray indicates intransitive universes
schemes in which sources and targets differ by entire generations of elements govern
discrete causal analogues of relativistic dynamics.
Dynamics; Quantum Causal Metric Hypothesis. The sum-over-histories approach
to quantum theory, suitably adapted, assigns amplitudes to families of co-relative histories in a kinematic scheme. The sources of these co-relative histories are viewed as
initial universes, and the corresponding targets are viewed as terminal universes.
In ordinary quantum theory, such amplitudes are complex-valued, but the complex
numbers cannot be taken for granted in the discrete causal context. Finite algebraic
structures provide interesting alternatives. These amplitudes may be interpreted as
encoding probabilities of reaching given families of terminal universes from given
families of initial universes. They are computed by summing quantities called phases
over paths between pairs of families of universes. The values of these phases are of
great interest; they supply the specific physical content of the theory, just as choosing
a Lagrangian supplies the physical content of a typical conventional theory, via
the corresponding action principle. The quantum causal metric hypothesis states that
these phases arise from causal structure in an appropriate sense:
The properties of quantum spacetime arise from a kinematic scheme S . In particular, the
phases associated with directed paths in S , under the sum-over-histories approach to quantum
theory, arise from the causal relations on the constituent universes of S .
57
As mentioned above, technical advantages result from working in terms of the

relation space over S; i.e., the corresponding space of co-relative histories.
Causal Schrdinger-type Equations. Given a suitable choice of phases, adapting
and generalizing Feynmans reasoning to the quantum causal context enables the
derivation of dynamical equations, which I refer to as causal Schrdinger-type equations. A special case of such an equation is
R; (r ) = (r )
R; (r ),
r r
where R is a subspace of the space of co-relative histories over a kinematic scheme,

r and r are consecutive co-relative histories in R, is the binary relation on R
induced by iteration of structure, is the phase map, and
R; is the past causal wave
function, defined by summing phases over the set of maximal irreducible directed
paths in R terminating at r .
Practical Considerations
Current Status of Rejected Assumptions and New Principles. The rejected
assumptions and new principles discussed in this essay occupy a variety of positions
with respect to theory and experiment, some more precarious than others. Manifold structure of spacetime remains tenable, but the existence of a universal time
parameter and static background structure have been doubtful ever since the first
observations supporting general relativity. The idea that noncommutative geometry
is essential to quantum spacetime is still conjectural. Consideration of the negative
version of the causal metric hypothesis may be omitted in favor of the stronger
positive version, to which I return below. Intransitivity of the causal relation is
obvious at large scales; for example, it is uncommon to be directly related to ones
grandparents. At the fundamental scale, the issue may be treated technically by examining whether or not physical predictions depend on including intransitive universes
in the sum over histories. A priori, the answer is yes, but special choices of phase
maps might annul this. Regarding causal cycles, I know of no solid evidence of their
existence; however, certain interesting interpretations of well-known phenomena do
incorporate them. Inadequacy of the symmetry interpretation of covariance might be
demonstrated only in conjunction with breakdown of manifold structure.
Turning to new principles, the causal metric hypothesis is most compelling in the
discrete setting, due to the metric recovery theorems. There is at present no convincing experimental evidence of spacetime discreteness, but it is thus far infeasible
to experimentally probe most regimes where such evidence might present itself. In
this regard at least, the plausibility of the causal metric hypothesis must be judged
indirectly at this time. The theory of co-relative histories can be neither right
nor wrong; it represents a viewpoint, more useful in some contexts than others.
The idea itself is quite general, but since relationships in category-like settings
58
B.F. Dribus
generally involve directed structure, the theory is most natural in the causal context.
The same is true of iteration of structure; moreover, the conceptual utility of this idea
seems greatest in the discrete causal setting. To demonstrate the contrast, Einstein
manifolds possess directed structure, but configuration spaces of Einstein manifolds
are generally nothing like Einstein manifolds themselves.
Recovery of Established Physics at Appropriate Scales. The parsimony of the new
principles proposed in this essay renders recovery of established physics from these
principles a substantial challenge, with a correspondingly great compensation if this
challenge can be met. The metric emergence problem for flat Minkowski spacetime
is the obvious first step toward both relativity and the standard model in this context,
since along with it will emerge the usual algebraic notions regarding coordinate
transformations and particle states. Note, however, that while the standard model adds
particle states as separate ingredients to Minkowski spacetime, both must emerge
together in the quantum causal context. Treating matter and energy as auxiliary data
would defeat the purpose of the program by violating the causal metric hypothesis,
as well as the principle of background independence. Based on our best guesses
about the fundamental scale, the simplest elementary particle interactions currently
accessible to observation might easily involve Avogadros number of fundamental
causal elements, or its square, or its cube. This is encouraging in the sense that such
magnitudes allow for familiar mechanisms such as entropy, and novel ones such as
graph-dynamical phase transitions, to produce sharp behavior and select for precise
quantities. However, it is discouraging in the sense that interactions large enough to
observe might be difficult to model.
Implications of Recent Observations. Last year, the Large Hadron Collider (LHC)
at CERN detected a new particle with energy near 125 GeV and properties similar
to the predicted properties of the standard model Higgs boson. Work is ongoing
to analyze possible deviations from these predictions, but concern exists that the
observed particle may match the standard model Higgs so precisely that the results
will provide little or no help in pointing to new physics. Whether or not this is true,
new high-energy particle physics may soon become technologically or economically
infeasible in laboratory settings. This sharpens the need for creative ideas regarding
the general problem of what experimental phenomena to search for and how to
search for them. In the context of quantum causal theory, results one might look for
experimentally include inexactness of symmetries, variation or small nonzero values
of physical constants, and new kinds of scale-dependence. Quantities such as the
emergent dimension of spacetime might vary with energy density, though such
effects might be extremely small.
Opportunities for observational physics exist beyond those afforded by traditional
laboratory experiments, particularly in cosmological contexts. Shortly before publication of this volume, the BICEP experiment, which measures polarization in the cosmic microwave background, reported detection of so-called B-modes of primordial
gravitational waves. This observation has been widely regarded as evidence in favor
of the inflationary hypothesis in cosmology, which is based primarily on the apparent
communication in the early universe of regions now widely separated. Inflation is
thus rooted in causal considerations. In my paper [12], I propose a quantum-causal
59
alternative to inflation, in which causal structure grew abruptly sparser in the early
universe, due to a graph-dynamical phase transition. I am presently trying to connect
this idea to experiment.
Connections to Quantum Information Theory. An intriguing possibility is that
quantum circuits might provide relatively large-scale windows into fundamentalscale physics. Such circuits may be represented by small causal universes whose
relations are weighted by single-qubit unitary transformations. In traditional quantum theory, important restrictions on such universes arise from results such as the
no-cloning theorem. Such circuits are small at ordinary scales, but they are many
orders of magnitude larger than the Planck scale. Only very simple quantum circuits
have been constructed to date, but complex circuits may be built in the near future.
The behavior of quantum circuits might be related to fundamental-scale behavior
in at least two different ways. First, and most optimistically, if spacetime possesses
a sufficiently simple structure, appropriate quantum circuits might serve as virtual
fundamental-scale laboratories easily accessible to future technology. Computations involving such circuits might then suggest unforeseen phenomena that could
be detected independently at reasonable scales. Alternatively, breakdown of manifold structure at the fundamental scale might lead to detectable deviations from
ideal behavior in quantum circuits. In particular, in the discrete context, the algebraic objects involved in standard quantum information theory, such as complex
Lie groups, would require replacement by complicated discrete objects. Due to the
information-theoretic sensitivity involved in the physical implementation of quantum circuits, quantum computing might provide an ideal setting in which to detect
the deviations associated with such objects.
References
1. K. Becker, M. Becker, J. Schwarz, String Theory and M-Theory: A Modern Introduction
(Cambridge University Press, Cambridge, 2007)
2. T. Theimann, Modern Canonical Quantum General Relativity. Cambridge Monographs on
Mathematical Physics (Cambridge University Press, Cambridge, 2007)
3. A. Connes, Noncommutative Geometry (InterEditions, Paris, 1990). English version of
Gomtrie non commutative
4. E. Verlinde, On the origin of gravity and the laws of Newton. J. High Energy Phys. 2011,
29 (2011)
5. R. Sorkin, Light, links and causal sets. J. Phys. Conf. Ser. 174, 012018 (2009)
6. J. Maldacena, The large N limit of superconformal field theories and supergravity. Int. J. Theor.
Phys. 38(4), 11131133 (1999)
7. A. Connes, M. Marcolli, Noncommutative Geometry, Quantum Fields and Motives (Colloquium Publications, 2007)
8. E. Frenkel, Lectures on the Langlands Program and Conformal Field Theory (Springer, Berlin,
2005). (Based on lectures given by the author at the DARPA workshop Langlands Program
and Physics at the Institute for Advanced Study, March 2004)
9. R.P. Feynman, Space-time approach to non-relativistic quantum mechanics. Rev. Mod. Phys.
20, 367 (1948)
10. S.W. Hawking, A.R. King, P.J. McCarthy, A new topology for curved space-time which incorporates the causal, differential, and conformal structures. J. Math. Phys. 17(2), 174181 (1976)
60
B.F. Dribus
11. D.B. Malament, The class of continuous timelike curves determines the topology of spacetime.
J. Math. Phys. 18(7), 13991404 (1977)
12. B.F. Dribus, On the axioms of causal set theory. Preprint. arXiv: http://arxiv-web3.library.
cornell.edu/pdf/1311.2148v3.pdf
13. D. Rideout, R. Sorkin, Classical sequential growth dynamics for causal sets. Phys. Rev. D61(2),
024002 (2000)
Chapter 5
The Preferred System of Reference

Reloaded
Israel Perez
Abstract According to Karl Popper assumptions are statements used to construct

theories. During the construction of a theory whether the assumptions are either true
or false turn out to be irrelevant in view of the fact that, actually, they gain their
scientific value when the deductions derived from them suffice to explain observations. Science is enriched with assumptions of all kinds and physics is not exempted.
Beyond doubt, some assumptions have been greatly beneficial for physics. They are
usually embraced based on the kind of problems expected to be solved in a given
moment of a science. Some have been quite useful and some others are discarded
in a given moment and reconsidered in a later one. An illustrative example of this
is the conception of light; first, according to Newton, as particle; then, according
to Huygens, as wave; and then, again, according to Einstein, as particle. Likewise,
once, according to Newton, a preferred system of reference (PSR) was assumed;
then, according to Einstein, rejected; and then, here the assumption is reconsidered.
It is claimed that the assumption that there is no PSR can be fundamentally wrong.
Introduction
One of the main objectives of present day physics is to formulate the so-called
theory of everything (TOE), a unifying theory that will be capable of describing
physical phenomena at all spatial and energy scales. For the past fifty years, legions
of physicists have worked relentlessly on this problem without arriving at satisfactory
results. This lack of success tells us two important things: first, that the problem is
much more complex than originally thought; and, second, that we may have arrived
at a dead end. Indeed, many researchers are realizing that theoretical physics is
falling into a deep crisis. Such crisis is mirrored in an ever increasing number of
FQXi 2012 Contest, Which of our basic physical assumptions are wrong?
I. Perez (B)
Department of Physics and Engineering Physics, University of Saskatchewan,
Saskatoon, SK, Canada
61
62
I. Perez
publications proposing bold alternatives and yet no clear answers have been found.
Actually, as more experimental evidence piles up, the puzzle aggravates. When we
found ourselves in such situation, the most natural way to proceed is to revise the
foundations and discard whatever that is blinding our sight. Here we shall review
one of the most prominent principles in the history of physics that, if seriously
reconsidered, can help to heal the current problems in physics, namely, the existence
of a preferred system of reference (PSR).
Epistemological Background
Before we move on to our central topic, it would be appropriate to discuss the
epistemology behind scientific theories. In this section we shall inform the reader
that certain kind of scientific hypothesis are natural components of physical theories
and that they are legitimate insofar as they reinforce the theory. In doing this, we
shall present a series of arguments that would be pivotal to understand why the PSR
can still play a major role in the future of physics.
Assumptions: True, False or Useful?

For the benefit of this work, we shall see as synonyms the words: postulate, axiom,
and principle. By these we understand a statement or proposition that serves as the
foundation for a chain of reasoning in the construction of a theory. Generally, it is
understood that this kind of propositions are true because they are so self-evident
that no proof is demanded to demonstrate their veracity. For instance, the proposition
A straight line segment can be drawn joining any two points is generally accepted
as true. However, when giving a deeper thought, some propositions turn out to be
uncertain. To illustrate this point, let us consider the following assertion: Space and
time are continuous. If we now ask: is this statement true or false? Needless to say,
the reply would be a shrug. In the face of the lack of certainty, the theorist can still
proceed and argue that during the conception of a theory whether the statement is
false or true turns out to be irrelevant provided that the predictions derived from
such assumption reproduce the experimental evidence at hand. In other words, the
assumption will gain its scientific value not from the preliminary judgement of the
statement but from the experimental verification of the predictions which are derived
from it. Only then, the theorist can vigorously contend that the assumption is not only
true but also scientific. Thus, despite that the truthfulness or falsehood of a statement
can be debatable, a theorist can presume, for practical purposes, that the proposition
has the potential to be really true.
Theories, on the other hand, can also rest on the basis of false statements as
long as these are helpful to strengthen the theory. Examples of this sort can be found
anywhere in physics. One typical case is the possibility of reversible processes in
5 The Preferred System of Reference Reloaded
63
thermodynamics. After all, we all are aware that this notion was invented having
in mind that, in practice, all processes are irreversible. Despite this, the principle
has been highly beneficial for this branch of physics. Other commonly used false
assumptions are: point particles or rigid bodies. Therefore, at the end, for the theorist
what is crucial is not the truthfulness or falsehood of the assumptions but their
usefulness in solving particular problems.
Physical Theories
In a wide sense, we understand by physical theory a rational and logical construct
composed of principles, concepts and definitions aim at explaining experimental
observations. Theories are formulated by seeking correlations and symmetries among
the observations. The summary of this search is then etched in the so-called laws of
nature [1]. These are a set of statements written in mathematical language where the
information of physical phenomena has been, so to speak, codified. According to
Max Tegmark [2, 3], theories have two components: mathematical structures (equations) and baggage. Baggage are words that explain how the theories are connected
to what we humans observe and intuitively understand, i.e., ontologies, physical
concepts and notions. And since physics is mainly written in mathematical language,
Tegmark goes beyond and asserts that the universe is actually a complex mathematical structure. He remarks that as the theory becomes more fundamental, the level
of abstraction increases up to the point that only mathematics would be capable of
describing the universe and, therefore, the baggage would be gradually replaced by
mathematics. According to him the TOE will have no baggage at all. At first sight,
this position seems to be extremist but a deeper reflexion shows us that it may not be
the case. To grasp the significance of this, first, one should ask what a mathematical
structure is. The minimalistic view is that a mathematical structure is no other thing
that a set of abstract objects connected by logical relations and operations. For our
purposes, this definition suffices since the task of physics is to seek for physical
correlations. From this standpoint, one is then allowed to assert that the description
of the universe can be reduced to a set of logical relations, i.e., physical laws. If
we agree, this means that what can be said of the universe in terms of baggage can
also be said with mathematics. Mathematics is, so to speak, a language in which our
intuitive perceptions can be expressed more effectively.
Now, since physical theories use mathematical structures, their structure should
be axiomatic. The axiomatization of physics allows us to apply the deductive method
from which theorems and quantitative predictions are derived. Such predictions are
usually tested in the light of experiments and when the predictions are corroborated,
one says that the model has shown its mettle. On the contrary, if the model is incapable of reproducing the data, it should be discarded. In this sense, the job of a
theoretical physicist is to single out the mathematical structures or models that fit the
observations.
64
I. Perez
Physical Objects
By analogy with the case of physical assumptions, during the construction of the
theory, whether physical objects really exist or not turns out to be irrelevant (because
ontologies can have a metaphysical source). This is in view of the fact that the
proposed concepts and objects will acquire their physical meaning once the model is
faced with experimental evidence. This could be the case of strings, loops, taquions,
axions, etc. In some other cases, the experimental observations mold the shape of the
theory as well as the properties of its physical objects. For instance, the conception
of electron was figured out from observations on electrolysis which suggested a
minimum quantity of charge. Later, electrons were conceived as an intrinsic part of
the atom and new physical properties such as spin were assigned. In brief, the notion
of a physical object strongly depends on the structure of the observations and the
theoretical framework where the object is interpreted.
Assumptions and Principles in the History of Physics

Hidden Assumptions
Since ancient times people have built theories based on principles which were
considered to be absolute truths, but as time went by some of them have been proven
to be actually false. One particular case is the famous assumption that heavier objects
fall faster than lighter ones. This principle was held as a physical law for hundreds
of years but, as more theoretical and experimental evidence accumulated, Galileo
showed that it could no longer be held. In some other cases, theories convey some
hidden assumptions. For instance, classical mechanics is based on the three laws of
motion and some definitions. In addition to these elements, the theory tacitly presupposes some unnoticed or disregarded assumptions such as: (a) measurements do
not affect the physical system under study; (b) the speed of propagation of physical
entities can have any velocity, even infinite; (c) physical quantities are continuous.
Assumption (a) fails in the microscopic realm. This issue was rectified by quantum mechanics (QM) with the introduction of a powerful postulate: the uncertainty
principle. The principle states, among other things, the probabilistic character of a
measurement due to the fact that the measuring instrument considerably influences
the response of the system under study. Assumption (b), on the other hand, finds
its restriction within the context of relativity theory in which there is a maximum
speed for the propagation of physical entities. And finally, assumption (c) also finds
limitations in QM where some physical magnitudes are discrete. If we extrapolate
this reasoning, we can figure out that surely our modern theories are still incomplete
and may need a deep revision.
65
Some Physical Assumptions in the History of Physics

In view of our previous discussion, it is worth extracting from our theories a short list
of some of the most typical assumptions. This will help us to be aware of how physics
erects and sculpts its reality. It is not the intention of this section to discuss either
the truthfulness or the falsehood of each example; for it is evident that some can be
true, false or uncertain. As we discussed above, what is of real value for physics is
their usefulness in solving the problems that physics has at a given moment of its
history. Some of the assumptions are:
Time flows equally for all observers regardless of their state of motion or its
position in a gravitational field.
The earth/sun is the center of the universe.
Rigid bodies, ther, vacuum, and fundamental particles (atoms) exist.
There is no speed limit for the propagation of physical entities.
The measuring process does not affect the system under study.
Space, time, and all other physical magnitudes are continuous/discrete.
Light is a wave/corpuscle.
Space is static/dynamic; space is a condensed state of matter, space is a network
of relationships.
Energy, charge, torque, linear, and angular momenta are conserved.
The laws of physics were created along with the creation of the universe and they
do not evolve with time.
All particles of a particular class are identical (e.g., electrons, quarks, positrons,
etc.).
The principle of: relativity, general covariance, uncertainty, equivalence, causality,
exclusion, cosmology, etc. hold.
There is an absolute system of reference.
Some of these assumptions have been definitely discarded, some are still in use, and
some others have been reconsidered several times in several epochs. From these three
cases, we would like to deal with the last one. One of the most famous cases is the
reintroduction of the heliocentric model, first proposed by Aristarchus of Samos in
the third century B.C. The model remained in the shadow for many centuries until
it was revived by Copernicus et al. in the XV century. The history of science tells
us that this model was by far more superior in describing the celestial mechanics
than the Ptolemaic system. Another famous case is related to the nature of light. In
1905 Einstein reintroduced into physics the almost forgotten notion that light can be
a particle, just as Newton had put forward more than two centuries ago. Armed with
this idea, Einstein built a rational explanation of the photoelectric effect discovered
in 1887 by Heinrich Hertz. These two examples teach us that no matter how old
or controversial an assumption might be, its potential to solve problems justifies its
reestablishment as a scientific hypothesis.
Unquestionably, some assumptions have caused a great impact more than others,
not only to the structure of a given theory but also to the whole evolution of physics.
66
I. Perez
Due to their preeminent influence, this kind of proposals deserve both a special attention and a scrupulous assessment; for their arbitrary rejection could be detrimental
for the progress of physics. In what follows, we shall discuss the last assumption
from the list above. I shall argue that this is one of the principles that physics should
revive if physics wishes to make considerable headway for years to come. To this
end, I shall try to dissipate some of the misconceptions that have been appended to
it for more than a century.
The Principle of Relativity Is Not at Variance

with the Preferred System of Reference
Newtons Absolute Space
When Newton developed his laws of motion, he thought that they were valid in
absolute space (AS). He contended that the water inside the famous bucket was
rotating relative not to the bucket but to AS [4]. This experiment gave him confidence
that any body possesses not only apparent (or relative) motion but also genuine motion
and such motion can only be relative to space itself, or generally speaking, relative to
a PSR. From this, it follows that if bodies move relative to AS, then this entity has to
be something endowed with physical properties and our disquisition would reduce
to identify them. For Newton, AS was an homogenous and isotropic background in
which material bodies were embedded, some sort of rigid container mathematically
represented by Euclidean space. What we all learn at school, on the other hand, is that
this entity is not composed of a material substance, rather, it is total emptiness. It is not
clear if Newton agreed with this view, but we have evidence that he actually thought
that there was an ethereal and pervading material substance conveying gravitational
interactions. In a letter sent to Bentley in 1692, Newton wrote:
It is inconceivable that inanimate brute matter should, without the mediation of something
else which is not material, operate upon and affect other matter, without mutual contact, as it
must do if gravitation in the sense of Epicurus be essential and inherent in it. And this is one
reason why I desired you would not ascribe innate gravity to me. That gravity should be
innate, inherent, and essential to matter, so that one body may act upon another at a distance,
through a vacuum, without the mediation of anything else, by and through which their action
and force may be conveyed from one to another, is to me so great an absurdity, that I believe
no man who has in philosophical matters a competent faculty of thinking can ever fall into it.
Newtons theory of gravitation assumes that between two celestial bodies there is
absolutely nothing mediating the interaction, instead, the alleged interaction occurs
by gravitational fields acting at a distance. We note from this letter, however, that
his theory does not reflect his actual view. His words seem to imply that he did not
even believe in total emptiness, since in Newtons time by the word vacuum people understood devoid of matter. In spite of this, what is relevant for us is that he
established in his theory that space was immovable. This assumption was precisely
67
what the philosopher Ernst Mach disliked [5]. For he questioned the scientific utility
of an entity that exists but is not affected by the matter it contains. Mach replied
to the bucket experiment arguing that the water moved relative not to AS but to
the stellar matter surrounding the bucket, because for him only relative motion was
possible. Although Machs argument is weighty, it is not clear what physical substance mediates the gravitational force. It was the insight of Einstein that shed light
on the problem some years later. Starting in 1905, Einstein rejected the ther as the
medium for the propagation of light and, by doing this, he left physical space again
absolutely empty, just as in Newtons theory. Einstein immediately realized this flaw
and from 1907 to 1916 he embarked in a historical journey to try to materialize
Machs ideas [69]. In the Einsteinian vision of the universe, space is mathematically represented by a pseudo-Riemannian manifold characterized by the metrical
field g which contains the gravitational potentials. As a consequence, one has to
conclude that the water in the bucket moves relative to the gravitational fields (GF).
Einstein then finally replaced the material substance, conceived by both Newton and
Maxwell, by the metric field [10]. Since then, the assumption that space can be made
up of a material substance has been ruled out from physics (we will return to this
topic below). Yet, physics has never ignored the power of intuition. In 1933 the Swiss
astronomer Fritz Zwicky discovered some anomaliesnow known as dark matter
in his studies of the Coma galaxy cluster. This evidence clearly suggests that there
is something in space by far more complex than originally thought and that it could
be indeed composed of an imponderable and invisible material substance: Newtons
substance? The ther that Einstein rejected? Unfortunately, Fritz discovery was
ignored for about forty years until Vera Rubin et al. revived it in the 1970s. Still dark
matter is one of the most puzzling problems in modern physics.
Invariance of Newtons Laws

Let us not digress from Newtons work and bear in mind henceforth the previous
discussion. It is well established that Newtons laws are invariant with respect to
Galilean transformations. This is in virtue of the Galilean Principle of Relativity
(GPR)In fact, in The Principia, Newton included this principle as corollary V and
justified it with the aid of the second law. It states that all mechanical laws are the
same in any inertial system of reference (ISR). But, what is the experimental meaning
of this principle? It simply means that no mechanical experiment can tell whether an
ISR is at rest or in motion relative to AS (this was well understood by Newton). The
understanding of this statement is vital to make clear that the GPR is not at variance
with the existence of the PSR. This being said, let us consider the following two key
questions:
1. Does the fact that the PSR cannot be experimentally detected mean that the PSR
does not exist?
68
I. Perez
2. If the PSR cannot be experimentally detected, does the assumption become a

meaningless assumption?
To grasp the deep significance of these questions, let us contemplate the following
situation borrowed from QM and that is beautifully discussed at length by Popper [11,
pp. 211217]. During the development of the atomic model, Niels Bohr imagined
that electrons follow orbital paths around the nucleus. He assumed that electrons
revolve with a given period T and from this the energy levels E n and thus the
emission spectrum of the hydrogen atom was computed. The key for the success
of this approach is the assumption that electron orbits are quantized. However, if
we take a closer look at the concept of path and scrutinize the principles of QM,
we find a serious difficulty. According to Heisenbergs principle, the experimental
determination of two correlated observables A and B is limited by the uncertainty
relation AB h/4, where h is Planck constant. In particular, if we assume the
observables to be the momentum px and the position x, the principle tells us that: the
higher the precision in the measurement of x, the higher the error in the measurement
of px and vice-versa. This means that it is impossible to experimentally determine the
particles path (i.e., the simultaneous knowledge of both px and x) with a certainty
exceeding the above expression. The reason rests in the fact that the measurement
affects the pristine state of the particle. Hence, according to Heisenberg himself,
it is meaningless to grant any real significance to the notion of path [12]. As
such, the path becomes an unobservable magnitude, i.e., a magnitude unaccessible
to experimental verification and therefore it is not useful as a basis for theoretical
predictions. The recognition of this, urges us to conclude that measurements cannot
serve as a foundation to test physical reality, so to speak, there is no such a thing as
physical reality since our instruments are not capable of revealing the true state of a
system. In reality, the information we get from our measurements is the outcome of
the interaction between the instrument and the system under study. If we conveyed
these considerations to the extreme, we would arrive at dramatic conclusions: we
would conclude that the factual character of physics is just a chimera. Fortunately,
not everything is lost and here the probabilistic and statistical interpretation of QM
comes to the rescue. Since we cannot access the precise state of a particle, at least
we can tell with some degree of certainty the probable outcome of an experiment.
On the other hand, the formalism of QM through the Schdinger equation allows
us to calculate with certainty the particles path up to a moment before the measurement, i.e., the formalism assumes that there is a path. Evidently, this is at variance
with Heisenbergs principle. So, does the path physically exist or not? The answer is
not trivial, but being physics a factual science, we understand that the experimental
data are essential to sustain the scientific credibility of a theory; for the data collected give us some information of the state of a system. But we have also learned
that our measuring processes modify the absolute state. Hence, despite that the path
cannot be exactly determined, it is scientifically legitimate to presuppose that the
particles path physically exists, just as the formalism assumes. Admittedly, the fact
that the measurement destroys the knowledge of the actual path, does not imply that
the physical notion of path has no scientific value and, at the same time, does not
69
encourage us to reject QM altogether on the basis that the theory is dealing with
unobservables (in the words of Popper, metaphysical constructs).
One more example of this type is illustrative to reinforce the view that the lack
of experimental evidence does not suffice to reject a hypothesis regardless of its
apparently unobservable character. Consider the postulate that space and time are
continuous. Here once more, we have no conclusive experimental evidence to thoroughly sustain this assumption. In spite of this, good reasons can be advanced for
trusting our postulate; actually, our theories have indeed shown that it can be true.
Thus, having in mind these two examples, the answer to the first key question is
clearly in the negative, for if one accepts the existence of a non PSR, one cannot
deny the existence of the PSR since the GPR assures the equality of the mechanical
laws in all ISR. Then, the second question is immediately answered also in the negative. From here we conclude that the GPR should not be understood as the exclusion
of only one ISR, but quite the contrary, as the inclusion of all of them. Evidently, the
arbitrary rejection of the PSR can be, in the long term, detrimental for the advancement of physics because we would deprive our theories from elements indispensable
for their logical consistency.
Invariance of the Laws of Physics

It is unquestionable that the PSR assumption in classical mechanics resulted highly
beneficial for the progress of physics for more than 200 years. The extension of
this assumption to electromagnetic phenomena was also very fruitful. It achieved its
highest peak with the development of electrodynamics. By the mid 1860s Maxwell
predicted that light was some kind of electromagnetic wave that travels through the
ther. Some years later, in 18878, Hertz could generate Maxwells waves, leaving
no doubt that Maxwell was in the right way [13]. Nonetheless, the mere corroboration of electromagnetic radiation did not suffice to establish the existence of the
medium. Maxwell was aware that in his equations the ther did not appear, that is,
they conserve the same form whether there was ther or not. Yet, for him, the material
substance was indispensable to avoid the action at a distance that dominated gravitational interactions. Besides, all known waves hitherto required a medium to propagate
and it was natural to assume that light waves could not be the exception. By the middle
of the 1870s, Maxwells theory was still under construction and many experiments
were in line waiting for a satisfactory explanation. In the following years, Maxwell
expanded the scope of the theory but, unfortunately, he prematurely died in 1879.
During the next decade, a new generation of physicists resumed Maxwells work. For
this reason, Oliver Heaviside, Oliver Lodge, and George FitzGerald were called The
Maxwellians. These brilliant scientists shaped Maxwells theory nearly as we know
it today [14, 15]. But in spite of the great advances, they did not solve the ther issue,
still, the equations had no explicit link to the ther. Fortunately, both Hertz and Paul
Dirac (six decades later) also realized Maxwells problem and promptly modified
the equations [13, 16, 17]. With the aim of accounting for effects of charged bodies
70
I. Perez
in motion relative to the ther, Hertz replaced the partial time derivatives by total
(also known as convective, Lagrangian, material or substantial) time derivatives. At
that time, his formulation did not attract much attention because some of the predictions were in disagreement with experiments on insulators. Incidentally, modern
investigations have revealed that Hertz formulation was not incorrect at all and that
the observed discrepancies can be attributed to quantum effects [18]. Indeed, Dirac
in 1951 also proposed a new electrodynamics and discussed that quantum principles
play an important role for reviving the ther concept and when considering the topic
seriously, the ther was crucial to build a satisfactory theory of the electron (now
known as quantum electrodynamics).
The problem of the ther was not only theoretical but also experimental. It was
imperative to show that the ubiquitous substance was not a mere idea. To prove its
existence, physicists engaged in an epic hunt by the end of the XIX century. In 1887,
Michelson and Morley carried out their famous interferometric experiment which,
according to the thinking of that time, would tell them whether the PSR existed or not
(below we dispel some misconceptions about these kind of experiments). As is well
known, the results were negative, and by analogy with the experimental implications
of the GPR, later, from 1900 to 1905, Larmor [19, 20], Lorentz [21, 22], and Poincar
[23] realized that no electromagnetic experiment can tell whether an ISR is at rest
or in motion relative to the ther. Such discovery was called simply the Principle of
Relativity (PR) and it is considered as a generalization of the GPR. Thus, in spite of
its undetectability, Larmor, Lorentz, and Poincar answered the above key questions
in the negative, whilst Einstein held the opposite opinion; he was actually appealing
to the principle of parsimony [6]. Since no experiment can tell whether an ISR is at
real rest or in real motion, Einstein declared that these statements are meaningless
(cf. with Heisenbergs opinion above). For him, just as Mach, only relative motion is
measurable and hence has real meaning. Nonetheless, if we strictly follow this line
of thought, motion would adopt a fictitious character. These theoretical perplexities
were apparently overlooked by Einstein but exposed years later by H. Ives and
G. Stilwell when they experimentally corroborated time dilation [24, 25]We shall
discuss in the following sections the importance of the PSR on this issue.
To comply with the PR, physicists were prompted to construct a new dynamics
which is now known as Relativistic Dynamics [26]. Both Maxwells laws and the new
kinematical and dynamical laws are said to be Lorentz invariant. The new symmetry
assures that not only the form of the laws of physics (LP) but also the values of
their constants remain the same in all ISR. This inevitably leads us to ask again: Is
then the PR at variance with the existence of the PSR? Certainly, the answer goes
in the negative [16, 17] for no experiment forces us to reject the PSR [22, 23, 27].
By analogy with the GPR, the PR should be understood not as the discrimination of
the PSR, but quite the opposite, as the inclusion of the PSR for the description of
physical phenomena. Within this context, Lorentz invariance experimentally means
that any experiment carried out in the PSR will lead to the same LP that can be found
in any other ISR. The history of physics tells us however that modern theories have
discarded it following Einsteins canon [6, 10]. But if one upholds the opinion that
the PSR is not an issue of parsimony but of usefulness and logical consistency in
71
the physics, one can claim that the assumption that there is no PSR is fundamentally
wrong. Let us make some other considerations to support this claim.
Experimental and Theoretical Considerations

in Favor of the PSR
Immediately after the development of the special theory of relativity (SR) a hot debate
not only on the existence of the PSR but also on the constancy of the speed of light
set in both on theoretical and experimental grounds. Even today many researchers
in the fields of physics and the philosophy of physics have kept alive these topics
from an epistemological perspective. Thanks to their perseverance, the good news is
that substantial advances have been made in the last decades. Although not widely
known, specially among the physics community, now it has been understood some
key aspects that can be fundamental that could be fundamental for the future of
physics.
Misinterpretation of Experiments: The Michelson-Morley

Experiment
In the first place, experimental arguments against the PSR have been misleading since
the advent of SR. Interferometric and non-interferometric experiments performed
during the XIX and XX centuries have been considered as proofs that the ther
does not exist. For example, it is common to find in textbooks statements such as:
if the ther existed the Michelson-Morley experiment (MMX) would have shown
any variation in the fringe shift N of the interference pattern. This is evidently
misleading because, as we discussed above, in virtue of the PR, no electromagnetic
experiment can tell about the existence of the PSR. This is quite clear from the PR
and it would be worth dissipating the misconceptions around the experiment because
most interpretations suffer from the same drawback.
The experiment had the purpose of measuring the absolute speed of the Earth relative to the therto illustrate my point, I will use the words vacuum and ther
interchangeably. In Fig. 5.1 we show the MMX as seen from both systems of reference, the vacuum (S) and the Earth (S ) that moves with speed v < c relative to S. For
the sake of illustration, the extension of the wave fronts for the four electromagnetic
waves have been accentuated to visualize the Doppler effect and their propagation
vectors are displayed as well. According to Michelson and Morley the determination
of v would be obtained by measuring changes in the phase difference produced at
the point P by the interference of two electromagnetic waves. Then, from the theory
of interference as calculated in S, the phase difference is
72
I. Perez
-v
(c)
(a)
t1
t1
t3
HSM
-v
t9
t2
Light
Source
(b)
t0
t0
t8
t2
t4
2 1/2
(c -v )
t5
l0
c-v
t6
Aether
wind
t7
t3
t4
-c-v
P -v
l0
Fig. 5.1 The Michelson-Morley experiment. a Forward and b backward advance of light waves
as seen from the observer at rest in the vacuum. Arrows represent the propagation vectors of the
four wave fronts. Solid arcs are for longitudinal wave motion and dashed arcs for transversal wave
motion. c As judged from Earth, the vacuum is passing by. Using Galilean transformations, the
one-way speed of light becomes anisotropic
= k(s s ) = ks = t,
(5.1)
here s=ct is the difference in the optical path length (OPL) for the waves in the longitudinal and transversal journeys, respectively; and t = t t the corresponding
time difference. In these expressions we have used the relations: c = /k = for
the one-way speed of light in vacuum, with k = 2/ the wave number, the wavelength of light and = 2 the angular frequency. From the preceding formulation
it is evident that a fringe shift exists whenever d/dt = 0. This is achieved in practice
by varying the OPL or changing the speed of the beams of light. Recall also that the
experiment was designed to revolve, so we have to consider an angular dependence
in the phase representing the revolution of the plane of the interferometer with
respect to the motion of the Earth (there is still another angle to be considered but
to convey our idea we do not need to include it here [2830]). Ignoring the length
contraction effect, the expression for the phase in the system S is:

2
2
2 1/2
2
2 1/2
,
l0 (1 sin ) (1 cos )
=
c
(5.2)

here = v/c, = 1/ 1 2 , and l0 is the length of the arms as measured at rest
in S. This equation tells us that = (, ), but if the apparatus is not rotated we
arrive, to a first approximation, at the traditional expression found in most textbooks:
(l0 /c) 2 . If we assume the Earth as an ISR then d/dt = 0 and no fringe
shift will be observed. So, the rotation of the apparatus is indispensable to observe
a fringe shift. The maximum phase occurs after a /4 rotation and N (the number
of fringes) is calculated by the difference before (B) and after (A) rotation; then we
have N = A B (2l0 /c) 2 . However, when we consider length contraction in
Eq. (5.2), we find that the OPL is the same for both light beams, so = 0 and hence
no fringe shift is observed regardless of the variations of and/or . This justifies
why the experiment failed to observe a positive result. Given this outcome we ask:
73
does this mean that there is no vacuum? To give a definite answer, let us now discuss
the physics from the perspective of an observer in S .
First, we emphasize that in Eq. (5.1) we have made used of the relation c =
to express in terms of time. This change is possible because in the solutions of
the wave equation, and k have a linear dispersion relation, i.e., the group and the
phase velocities V /k = /k = c are the same in all directions (isotropy of
the one-way speed of light in vacuum). That the one-way speed of light is isotropic
in at least the system S, does not follow from SR but it is a direct consequence of
electrodynamics. Our problem is then to find out if this is also true in any other ISR in
motion relative to the vacuum. Immediately, some closely related questions come to
our mind: Where do the alleged anisotropy of the one-way speed of light find in most
relativity textbooks come from? What is the physical basis for postulating that the
speed of light is independent of the state of motion of the source or the observer? In
the last statement it is implied that the velocity of light can depend on the velocity of
the source or the observer. But what is the rational source that prompted physicists
to conceive such possibility?
Before the discovery of the Lorentz transformations (LT) the only known
transformations relating two ISR, moving with relative velocity v, were the Galilean
transformations:
t = t.
(5.3)
r = r vt;
Note that the time relation expresses that time flows equally for all inertial observers,
meaning that there is a unique rate of flow in all ISR. As a consequence of these
transformations, physicists were induced to believe that the speed of light could
acquire different numerical values in frames in motion relative to the ther and,
in consequence, the wave numbers ks or the frequencies s for each of the light
beams involved in an experiment would not take on, in general, the same values.
This can be easily shown by applying these transformations to the phase of the
wave function = a exp[2i(n r ct)/], where a is the amplitude and n a
unit vector. After a straightforward calculation, the phase in the system S becomes:
2i[n r (c n v)]/. Therefore, in S , the speed of light is c n v, which
is anisotropic. Evidently, the error in this prediction is the misapprehension that
electrodynamics and Galilean relativity are compatible formulations. This is what
Einstein spotted in his famous Gedankenexperiment about the race with light rays.
While Maxwells theory states that the one-way speed of light is a constant relative
to the ther, the Galilean addition of velocities dictates that the speed of light must
be velocity dependent.
Although we have already identified our nave mistake, let us further proceed
with our analysis. As seen from S (refer back to Fig. 5.1c), the speed of energy
flow (or energy flux given by the Poynting vector) for the longitudinal and
oblique
beams is c, therefore the velocity of the energy flow in the y-direction is c2 v 2 .
According to the observer in S , the vacuum is passing by with velocity v = v x ,
and if we apply Galilean relativity to light propagation, the velocities of the energy
flow for the four beams in the frame S must take on the values:
74
I. Perez

c
= (c v) x ; c
= c2 v 2 y ,
(5.4)
where stands for forward and backward directions, respectively. Thus, for the
parallel direction the wave fronts travel the OPL: s1 = l0 = t1 (c v) and s2 = l0 =
t2 (c + v); for the forward and backward journeys, respectively. Accordingly, the
time spent in the longitudinal journey is t = (2l0 /c) 2 . The time for the transversal
journey is calculated as follows. The transversal distance in one direction is s3 =
t /2. Solving for t and taking the difference t , we obtain to a first
s4 = l0 = c
approximation: (2l0 /c) 2 . Since the Earth is in motion relative to the vacuum,
then t = t 1 and l0 = l0 1 . Taking into account these effects in our previous
calculations, we also find that = 0. Showing once more that the experiment cannot
determine the Earths velocity relative to the vacuum.
We must remark, that the absolute speed of light waves never changes regardless
of the speed S relative to S, because the light waves travel through the vacuum and its
speed is determined by the properties of the medium. If we assume that the medium
is static, isotropic, homogeneous and its temperature remains constant, we have no
reason to believe that the speed of light would change. The alleged anisotropy of the
speed of light is just a fictitious effect caused by the relative motion between the Earth
and the vacuum. This automatically means that the speed of the waves is independent
of the motion of the source or the observer (second postulate of SR). Thus, the null
result of these kind of experiments does not prove that there is no medium. Some
experimentalists that concluded that there is no medium, made the same mistake
of convoluting Galilean relativity and electrodynamics [31, pp. 518524].
If we have made clear that no experiment of this kind rules out the PSR, we are
faced again with the two key questions above and, therefore, the issue may become
only a matter of usefulness and coherence in the logic of a theory. Einstein rejected
the ther, first, because, from the theoretical viewpoint, SR could not make special
distinctions among ISR; actually, for him the ther assumption was not wrong but
appeared to be superfluous. And second, because, from the experimental viewpoint,
there was no unambiguous evidence of its existence. Nevertheless, according to the
discussion of the previous section, the first argument is weak, for if one follows such
line of thought then Newtons AS would have been rejected as well from classical mechanics since the GPR guarantees that all ISR are equivalent. In Einsteins
epoch, the second argument had a great weight, however, the discussion given above
and the experimental evidence accumulated after the 1930s, strongly disagrees with
Einsteins view. The experimental evidence we are referring to is this. Consider the
following hypothetical situation. Imagine that before the discovery of relativity, particle accelerators had been already developed. And assume that the ALICE, ATLAS
and CMS collaborations at the large hadron collider had released the news, wellknown today, that the quantum vacuum is actually a perfect fluid [32]. If this fluid
were assumed to be at rest and not significantly affected by the presence of material
particles it would immediately be identified as the ther or AS; just in the same way
as in 2012 many physicists sympathized with the discovered boson at the LHC and
identified it as the Higgs boson despite that they did not know yet its other physical
75
properties (spin, etc.). So, if by 1905, physicists had already discovered the presence
of dark matter, the background radiation, the presence of a perfect fluid and the
Casimir effect, would physicists, despite the success of relativity, have good reasons
to discard the medium for light and thus the PSR? Indeed, the answer would be in
the negative. The concept would be maintained because the experimental evidence
would have suggested its presence.
On the Experimental Corroboration of the Second

Postulate of Special Relativity
In second place, research on experimental methods used for the measurement of
the speed of any physical entity, shows that when the paths of the physical entities
involved in the measurement form a closed circuit, what the experiment measures is
an average speed, i.e., a harmonic mean of the speed or the so-called two-way speed.
This implies that it is not feasible to measure the one-way speed of light [33]. Since
the second postulate of SR tacitly states that there is a finite isotropic speed c, the
studies reveal that this postulate has never been experimentally corroborated. Under
such scenario, one can raise sharp objections against either SR or electrodynamics
similar to those raised against the PSR. But despite that nature conspires to hide from
us this crucial knowledge, there are enough reasons to hold the postulate. And, once
more, this lack of experimental evidence does not impel us to reject it.
Misinterpretation of Newtons Theory

The structure of a physical theory is fixed and cannot be modified at will. A genuine
scientific theory cannot, when faced with an irreconcilable fact, be rectified a little
to make it agree. There are some colleagues (see for instance [34, p. 6]) who have
claimed that Newtonian mechanics is just as relativistic as SR under the argument
that Galilean transformations leave invariant Newtons laws. They even contend, led
by the relativity school, that Galilean invariance demonstrates that there is no PSR.
This is, of course, nonsense for AS is a principle of the theory and it cannot be
arbitrarily eliminated. In fact, we shall see below that Minkowski space plays the
role of absolute space in SR.
Relative Motion Leads to Quandaries

It is worth discussing briefly how quandaries arise in SR due to the idea that only
relative motion is significant. Here I shall consider the case of time dilation which is
usually confused with the so-called clock paradox. I do not treat the original clock
paradox [6] since it has shown to be misleading [35]. Instead, I slightly modify the
76
I. Perez
situation to identify where the perplexing part of SR is. The problem is related to the
topic of relative motion versus absolute motion. Imagine three synchronized clocks
placed in line at three equidistant points A, B, and C. Consider that clocks at A and
C are moved simultaneously (that is, at t A = t B = tC = 0) towards B with the same
constant velocity v (and by symmetry, the same initial acceleration if you wish). Now
we wonder whether time really dilates or not for clocks in motion. (i) According to
SR, an observer at rest next to clock B will figure out that, since both clocks A and C
are moving towards B at the same speed, they will
arrive at B synchronized among
each other but lagging behind clock B by the factor 1 (v/c)2 . So far so good, but
this is not the end of the story. Relative motion strictly dictates that the two clocks
A and C are not only moving relative to each other at constant speed V , but also
relative to the clock B at speed v. Since according to Einstein there is no PSR, this
means that absolute motion is meaningless absolute motion is meaningless. On the
basis of this theoretical restriction, it is equally legitimate to judge the situation from
the standpoint of an observer in the ISR of clock A. (ii) From this perspective, clocks
B and C are approaching clock A at speeds v and V , respectively. And by symmetry,
when the three clocks meet, the observer at A will find that, bothclocks B and C,
will lag behind clock A in proportion to their relative velocities, 1 (v/c)2 and

1 (V /c)2 , respectively. Moreover, since V > v, he will assert that clock C will
be lagging behind clock B. (iii) With the same right and by the same argument, a third
observer in the ISR of clock C will claim that when the three clocks meet, clocks
B and A will be lagging behind clock C in proportion to their relative velocities,
etc. Certainly, according to the view that only relative motion is meaningful, the
tree options are equally legitimate, although it is obvious that if the experiment is
performed the three options cannot be true. In view of these baffling conclusions, it
is impossible to decide solely on the grounds of the principles of the theory itself,
what would be the actual outcome of an experiment like this (a similar situation
occurs with the stellar aberration). By 1937, Ives and Stilwell realized about these
quandaries and discussed the topic at length [24, 25, 36]. They carried out a series
of experiments to test time dilation and pointed out that the source of the problem is
the omission of the ther. If we reintroduce the PSR in our picture, we will have a
logical criterion to decide what will be the actual outcome of the experiment since, in
this case, only absolute motion is meaningful (below we discuss how to distinguish
absolute from relative motion). Even if we were not able to determine the real state
of motion of an ISR, we can still theoretically assume that either the ISR of clock B
is at rest or moving at speed w relative to the PSR. Under this scenario, we realize
that the flow of time of clock B will remain constant at all times whereas the flow
of time for the clocks A and C will be altered since they are absolutely moving (for
detail calculations on this view see Ives and Stilwell works [24, 25, 36]). Therefore,
from the absolute point of view, options (ii) and (iii) are nave and can be discarded at
once. We are left then with option (i). Whether this option is true or not would depend
on the adopted clock synchronization convention, topic which is outside the scope
of present work [33]. This example constitutes a logical justification to reconsider
the PSR. Einstein rejected it because he considered it superfluous, now we see that
77
parsimony leads to logical predicaments. What we learn here is that parsimony is not
always the best choice; for if a theory A, assuming the PSR, explains the same amount
of observations as another theory B, in which no PSR is assumed, one should chose
theory A because it is free from perplexities. The theory A we refer to is not SR with
a PSR, but Lorentz ther theory [21] [not to be confused with FitzGerald-Lorentz
hypothesis about length contraction].
The Law of Inertia and the Conservations Laws

When we work within the context of Newtonian mechanics, we are usually unaware
how the law of inertia was defined. The law states that: Every body persists in its
state of rest or of uniform motion in a right line, unless it is compelled to change that
state by forces impressed thereon. But with respect to what system of reference is
this true? To answer, let us imagine that we have two systems of reference, S and S ,
with S moving along the x direction with velocity V and acceleration W relative to
S. We now have a particle that moves also along the x direction with velocity v and
acceleration w relative to S as well. According to Galilean relativity, the velocity v
and the acceleration w in S are given, respectively, by: v = v V , w = w W .
If the particle moves by inertia relative to S, we have in S that w = W = 0, that
is, the observer in S cannot figure out that the particle is moving by inertia. For this
reason the law of inertia loses its meaning if we do not specify to what system of
reference the law refers to. In consequence, in order for this law to be meaningful,
we need to define a special system where the law of inertia holds. Following Newton,
such system is AS and by virtue of the Galilean transformations, the law of inertia
is also true in any other system moving uniformly relatively to AS. It follows that a
system in which Newtons laws hold is an ISR.
The recognition of Newtons laws and AS invites us to accept their consequences
since space (Euclidean) and time are isotropic and homogeneous. From these properties, as we know from Noethers first theorem, the conservation of momentum and
energy follow. Because Minkowski space-time has these properties, the law of inertia
and the conservation of energy-momentum hold also true in SR. In his article on cosmological considerations, [37] Einstein, led by Mach, manifested his disagreement
with Newton enunciating the so-called relativity of inertia: In a consistent theory
of relativity there can be no inertia relatively to space, but only inertia of masses
relatively to one another. This statement is just another version of Machs principle.
To incorporate this principle in General Relativity (GR), the law of inertia and the
sacred conservations laws had to be redefined. Lorentz, Klein, Einstein, et al. soon
realized this inconvenience and tried to amend it. The reason is quite evident: for
Einstein, inertia is due to the presence of other masses and relativistic space is, in general, dynamic (non-Euclidean). In an attempt to save energy conservation, Einstein
introduced the pseudo-tensor tik and claimed that the total momentum and energy,

Ji = (T + ti4 )d V (with T the stress-energy tensor), of the closed system are,
78
I. Perez
to a large extent, independent of the coordinate system [38, 39]. The contemporary
version acknowledges, however, that the conserved quantities are not in general
conserved in GR and other diffeomorphism covariant theories of gravity [40, 41].
The problem consists in that in GR, gravitation is represented by the metric tensor
g (that underlies the geometry of space) and the gravitational energy contained in
the geometrical part cannot be, in practice, localized. In mathematical terms, this is
implied in the divergence of the stress-energy tensor
T ; = 0,
(5.5)
which expresses the exchange of energy-momentum between matter and the gravitational field. For asymptotically flat and stationary spacetimes at infinity (i.e., spaces
that tend to Minkowski space), one can always find an energy conservation law by
integration of Eq. (5.5). But, it is no longer possible for general spacetimes.
As for the law of inertia, Einstein worked out a cosmological model where he
first considered the scenario of an open and expanding universe [37]. To solve the
gravitational field equations, one needs to provide the boundary conditions. By a
suitable choice of a reference system, the g in spatial infinity tends to Minkowski
metric . He rejected this possibility because he first realized that the reference
system would represent a PSR contradicting the spirit of relativity; and, secondly,
because this choice would discriminate Machs principle. He then opted for avoiding
boundary conditions at infinity and considered the possibility of a finite and closed
universe. For this purpose, he modified his field equations introducing the famous
cosmological constant (for details on the physical meaning of , see A. Harvey
et al. [42]). The modified equations read:
R
1
Rg = T + g ,
2
(5.6)
where R is the Ricci tensor and R the scalar curvature. With this new term, he
thought he had succeeded not only in satisfying Machs principle but also in removing the boundary conditions. His joy, however, did not last much because, in the
same year, the Dutch astronomer de Sitter found a vacuum solution in which matter
(T = 0) was not necessary to define an ISR [43]. Three decades later, Pauli recognized that the problem was still open [39, p. 182] and Steven Weinberg expressed in
his book of 1972 that the answer given by GR to this problem through the equivalence principle lies somewhere between that of Newton and Mach [44, pp. 8688].
More recently, some physicists claim that the Lense-Thirring effect contemplates
some effects of Machs principle [45], although, most specialists agree that GR is
neither completely Machian nor absolutely relativistic [46, p. 106], implying that,
after all, both the PSR and Newtons law of inertia are still very alive. In addition to
this failure, a closer inspection of the ontology of space in GR reveals peculiarities
that require a careful examination.
79
Are Space and Vacuum the Same Physical Entities?

The understanding of the nature of the vacuum might be crucial to achieve the TOE
because this problem is closely related to the energy density of the vacuum vac
and the cosmological constant problem. The nave calculations for vac obtained
from quantum field theory (QFT), yield a value that differs by about 120 orders of
magnitude when compared to the value obtained from cosmological observations
where it is believed that GR is the correct theory. Up to now, physicists are still
perplexed for such a big difference [43, 47, 48]. This can indicate that either GR,
or QFT, or both are in need of serious revision. After many years of study, some
researchers suspect (as I do) that the geometrical representation of space may not
be the best choice for the future of physics [49]. In fact, this notion seems to be at
variance with the notion of vacuum in QFT. In this theory the vacuum has a ground
state energy different from zero, the so-called zero point energy. Nevertheless, it
appears that GR posses a distinct understanding. Taking a closer look at Eq. (5.6),
we notice the following peculiarities: The left hand side represents the geometry of
space where the energy of the GF is included. On the right hand side, we find the
T where we include matter and the -term that can be understood as a repulsive
force due to the vacuum energy (known as dark energy). Since the latter can be put on
the left hand side too, one can interpret it as gravitational energy rather than vacuum
energy (this is still under debate). If we leave it on the right side, the vacuum is
viewed, in its rest frame, as a perfect fluid with energy density vac and isotropic
pressure pvac , both related by w = Pvac /vac . For this fluid the stress-energy
tensor reads
vac
= (vac + pvac )u u + pvac g ,
(5.7)
T
where u is the fluid four-velocity. If we assume a motionless fluid, the first term in
this expression is zero. Hence, vac is proportional to and it is legitimate to assume
that vac = = /(8G). Now, Einstein equations for empty space with no
vacuum energy ( = 0) read R = 0. Solutions for these equations are Minkowski
and Euclidean spaces. It is worth noting that Einstein considered Minkowski metric
as a special case of metric with constant gravitational potentials . Since =
constant, there is no GF in this space. Moreover, from the geodesic equation
d2x
dx dx
,
=

ds 2
ds ds
(5.8)
where is the Christofel symbol and s a scalar parameter of motion, it follows

that a test particle moves in straight line (Newtons law of inertia). This rationale
also applies to Euclidean space, implying also that, in this space, there is a constant
gravitational potential and, therefore, zero GF. Nonetheless, it is difficult to reconcile ourselves with this interpretation given that there are no sources of gravitation
and given that, in Newtonian mechanics, Euclidean space represents AS, i.e., total
emptiness. We are thus tempted to think that if Euclidean space represents the PRS
80
I. Perez
in Newtons theory, Minkowski space represents the PSR in SR. Naturally, for GR,
Minkowski space is not a realistic space, although our analysis is exposing the substantival character of space in GR. Considering now that the vacuum has nonzero
energy, Einsteins equations read R 21 Rg = g . If > 0, vac > 0 and
pvac < 0, one of the solutions is the de Sitter flat space. If, < 0, vac < 0 and
pvac > 0 we have the anti-de Sitter space. In the former, space is open and expands;
and in the latter, space is close and the expansion decelerates. So, tests particles will
move apart in a de Sitter space; implying inertia without matter (in contradiction to
Machs principle). Finally, if T = 0, the field equations are of the form (5.6) and
one of the solutions is the well-known Friedmann-Walker-Robertson space which
represents our real expanding universe. In any case, we see that regardless of our
considerations on the right hand side of Eq. (5.6), there is always space (except
when all components of g are zero) which is subsequently filled, according to our
considerations, with stuff. In this sense, GR represents space as a container (the
pseudo-Riemannian manifold) that responds according to the energy-matter content. Is there any problem with this? Indeed, in GR we can have space seen as perfect
emptiness, and space seen as an energetic perfect fluid (the vacuum). We see that, just
as the Euclidean manifold plays the role of the background in Newtonian mechanics, the pseudo-Riemannian manifold along with the metric is playing the role of a
substratum, since space can exist even if T = 0 and = 0.
The fact that in GR is possible to have space without sources of any kind seems to
be in contradiction with the notion of vacuum as seen from QM and electrodynamics.
Whilst geometrical spaces have no electromagnetic properties per se (compare to
the Reissner-Nordstrm metric), the vacuum of electrodynamics has intrinsic finite
electric permittivity and magnetic permeability . The assumption of the existence
of a perfectly empty space is, just as the assumption of the existence of rigid bodies,
a false but useful assumption. That the vacuum is an actual physical entity can
be demonstrated even from the perspective of electrodynamics. To get the feeling
of this, consider the following situation [50]. Suppose that a coil with n turns is
energized and carries a current I . Accordingly, the magnetic induction of the coil is
B = 0 n I + 0 M, where n I is the magnetic intensity and M is the magnetization
induced in the coil. If we carry out an experiment where we keep the current constant
and reduce the density of matter, B decreases. As we continue to eliminate all
matter then M = 0 and B = 0 n I . This result experimentally demonstrates that the
vacuum is a paramagnetic medium with magnetic permeability 0 = 4107 N/A2 .
And because this property is exclusive of matter, the experiment tells us that the
vacuum is not deprived of material substance at all. In contrast, if physical space
were totally empty, one would expect null electromagnetic properties.
On the other hand, the field of condensed matter has made important advances,
particularly, in the field of Bose-Einstein condensates and superfluids. Giving the
mathematical analogies of these systems with the quantum vacuum, some physicists
have suggested that the vacuum can be a condensed state of matter [51]. One of the
consequences of this approach is that perhaps the equivalence principle and some
other symmetries such as Lorentz invariance and gauge invariance may not be fundamental at all but emergent from the low-energy sector of the quantum vacuum.
81
Indeed, F. Witenberg showed that assuming the vacuum as a kind of plasma composed of positive and negative massive particles interacting by the Planck force over
a Planck length, one can derive QM and Lorentz invariance as asymptotic approximations for energies small compared to the Planck energy [52]. He finally concluded
that Minkowski spacetime is superfluous for physics provided that Lorentz invariance
is assumed as a dynamic symmetry, just as conceived in Lorentz ther theory where
length contraction and time dilation are explained in terms of atomic deformations
and motions through the medium. Certainly, this would imply that electromagnetic
fields, and no less particles, are states and excitations, respectively, of the vacuum.
Following a similar line of thought, M. Urban et al. recently showed that the origin
of the speed of light (and the permeability 0 and permittivity 0 constants) is the
result of interaction of photons with fermions pairs spontaneously produced in the
quantum vacuum. This implies, again, that the vacuum is the medium for light and
that the speed of light is strictly defined relative to it. As for the law of inertia, B.
Haisch et al. put forward a quantum mechanism to justify its origin. They showed that
inertia can originate from the quantum vacuum without alluding to Machs principle
[53]. Admittedly, all this evidence strongly suggests that the vacuum is some sort of
diluted material fluid. If we trust this view, the energy density vac found from QFT
might be downright correct. Admitting the vacuum as a material substance capable
of transmitting gravitation, just as Newton devised it, prompts us to deeply revise
the geometrical interpretation of space and gravitation in GR.
General Relativity Is Not Fully Relativistic

and the Speed of Light Is Not Constant
Absolute Motion Versus Relative Motion
In his celebrated scholium, [4] Newton taught us how to distinguish false motion
from real motion; there he wrote:
The causes by which true and relative motions are distinguished, one from the other, are
the forces impressed upon bodies to generate motion. True motion is neither generated nor
altered, but by some force impressed upon the body moved; but relative motion may be
generated or altered without any force impressed upon the body...
Then, at the end, he topped off:

But how we are to collect the true motions from their causes, effects, and apparent differences;
and viceversa, how from the motions, either true or apparent, we may come to the knowledge
of their causes and effects, shall be explained more at large in the following tract. For to this
end it was that I composed it.
Newtons Principia was completely devoted to demonstrate that absolute motion

exists and can be distinguished from relative one by forces. In Newtons theory,
an observer in a non-inertial system (NIS), say the Earth, that rotates with angular
82
I. Perez
relative to an ISR will observe a series

velocity and translates with acceleration R
of forces acting on a particle of mass m:
m r m ( r) 2m v,
F = S + mg m R
(5.9)
where S is the sum of external forces, g the gravitational acceleration and v the
particles velocity as measured on Earth. The fourth term appears in case is not
constant, the fifth term is the centrifugal force and the last one is the Coriolis force. All
these additional forces are known as inertial, fictitious or pseudo forces. The adjective
fictitious and the prefix pseudo speak for themselves. In this theory, these are not
real forces because their nature arise from relative motion. Not convinced, a relativist
will claim that the Earth observer feels, i.e., measures these forces and, hence, they
are real for him; consequently, the adjective fictitious fades away. A Newtonian in
turn will reply that if motion were purely relative, the Earth could be considered as
static frame subjected to the pseudo-forces of Eq. (5.9) and, in consequence, the view
that the world rotates around the Earth would be equally true. This line of reasoning
will send us back to the idea of the Earth as the center of the universe and one would
not be able to decide whether the Earth really rotates or not (similar to the time
dilation quandary discussed in section Relative Motion Leads to Quandaries). For
a Newtonian, the relativist view is, needless to say, nave. For if an experiment could
be conceived to measure the effects of the pseudo-forces, we would be demonstrating
that AS exists. The Focault pendulum is a beautiful example that Newton was right.
Exploiting the effects of the Coriolis force, the experiment not only gives geocentrism
a coup de grce, but also informs us that absolute rotation can be measured even if
we were enclosed in a laboratory without observing the fixed stars. The experiment
shows that the Earths angular velocity relative to AS (Euclidean space) can be
determined by just measuring the rotation of the oscillation plane as function of
time. Likewise, the Michelson-Gale experiment shows clear evidence that, without
looking at the sky, the Earth absolutely rotates relative to the vacuum [54]. This
experiment not only measures but also teaches us that the vacuum is the medium
for light. If we now judge these experiments from the standpoint of SR, the Earth
revolves relative to a system either in motion or at rest relative to the Minkowskian
background (physically speaking the vacuum). If is small, the calculations from
both SR and Newtons theory agree. And what does GR have to say about this? For
GR, as in the case of the Newtons bucket, the Earth rotates relative to its GF so that
the fictitious forces become genuine GF (see the Kerr field and the Lense-Thirring
effect). In the case of the Focualt experiment, GR includes tiny corrections that,
in practice, cannot be distinguished from Newtons results. We thus see once more
that g plays the role of background for the rotation of the Earth, by analogy with
the Euclidean metric in Newtonian theory. But just as one cannot place a system of
reference at absolute rest relative to AS, one cannot place a system of reference at
rest relative to the GF. Thus, to determine astronomers use Eq. (5.4) and assume
a special ISR, the so-called fixed-space system or International Celestial Reference
System. Such system, evidently, is an ideal candidate for a PSR.
83
In the development of GR, Einstein sought to justify inertia, and therefore

rotational motion, relative to the masses of the universe through both Machs principle
and the equivalence principle. We saw above that he did not succeed. Furthermore,
Einstein did not succeed either in creating a fully relativistic theory. This means that
not all systems of reference are equivalent. That this is the case can be seen from the
principle of general covariance [44, pp. 9193]. In 1917, Kretschmann [55] recognized in a critical study of GR, that the principle does not imply that the LP most
be relativistic, but only that their form must be the same under general coordinate
transformations. In fact, even Newtons laws can be written in covariant form. Thus,
general covariance is not a PR but a principle that imposes restrictions between matter and geometry [56, 57]; for this reason, John Wheeler suggested that Einsteins
theory should be called, instead of general relativity, Geometrodynamics. Today,
some physicists still look for a fully relativistic theory where Machs principle could
be embraced [46]. This, indeed, indicates that GR is not hermetic to accept a PSR,
even going against its own spirit. Both astronomy and cosmology have always been
in need of a special system of reference to assess the celestial dynamics and define a
cosmic time. The cosmic microwave background radiation also strongly demands a
special system. It seems to me that the PSR is valuable to satisfactorily account for
physical phenomena at all scales [5860].
Covariance and the Variation of the Speed of Light

Before we close this treatise, it is worth elucidating the fact that SR has actually
only one postulate, i.e., the PR, since the second one is already tacitly included in
electrodynamics. This postulate is valid insofar as one deals with ISR, but invariance
no longer holds for NISor appealing to the equivalence principle for systems of
reference in GF. This means that the value of their constants and physical quantities
may acquire different values in different NIS. As we showed in Eq. (5.9), the same
occurs in Newtons theory. Covariance, by contrast, only demands that the form of
the LP must remain the same. As early as 1911, Einstein was aware of this [7].
He knew, for instance, that the only cause that could change the path of light is by
varying the speed of the different parts of a wave front. During the development of
GR, he emphasized that the assumption of the constancy of the speed of light must
be abandoned for NIS [79]. However, the principle of general covariance (also
known as diffeomorphism covariance or general invariance) demands that the metric
tensor g must change whereas all constants must remain the same under general
coordinate transformations. Since then it is widely believed that the speed of light is
a universal constant at any point of a GF. This could be true insofar as we understand
space as GR does, but we have shown above that the vacuum can be seen as a diluted
material fluid. Under this assumption, we can reinterpret the bending of light just
as a simple refraction phenomenon. One can keep the vacuum static and assume
84
I. Perez
it as a inhomogenous medium with degraded refraction index that vary as function

of position in the GF. The gradient depends on the gravitational potential which, in
turn, will make the speed of light function of position. Thus, within this context, the
warping of space can be physically understood as the change in the density of the
medium [61]. Certainly, this will not account for the perihelion of Mercury or other
gravitational phenomena, but it gives us a hint on how to build a unified theory and
reinterpret gravitational effects.
Final Remarks
Throughout the course of this treatise I briefly reviewed the role played by the PSR
in physics. In doing so, I presented a series of epistemological, experimental, and
theoretical arguments to dispel the series of misconceptions around this central tenet,
and, at the same time, I gave weighty reasons to champion its reintroduction into
physics. I also pointed out that the geometrization of space may not be the most
appropriate for the future of physics. Instead, the experimental evidence at hand
suggests that space is a dynamical condensed state of matter. Due to the lack of
space, I cannot discuss here the progress that has already been advanced based on
these radical ideas and I prefer to leave it for a future contribution. The purpose
of this work is to show that the PSR is not in conflict with physics and that the
vacuum can be understood in a different way. Once we accept this, the next step
is to unify the concepts of particle and wave using the notion of quasiparticles. In
this sense, a field would become a state of the vacuum and a particle an excitation.
The implications of this insight may impact physics at all scales leading to the TOE
without invoking exotic assumptions (multiverses, extra dimensions, etc.). In my
opinion, there are enough experimental and theoretical elements for a new revolution
in physics. Thomas Kuhn taught us that a paradigm shift might be a thorny episode in
the evolution of science [62]. The PSR assumption constitutes a paradigm shift that
would request a drastic change in the way of understanding reality. Some established
facts such as the expansion of the universe and the big bang model may need to be
revised in the light of this new paradigm. Inevitably, this will lead us at some point
to the bucket problem. And just as Newton held, here it is claimed that the water
moves relative to the vacuum, provided that we understand elementary particles as
quasiparticles and the vacuum as a dynamical material fluid.
Acknowledgments This research was supported by the Natural Sciences and Engineering Research
Council of Canada, the Canada Research Program and CONACYT Mexico under grant 186142.
The author is thankful to the University of Saskatchewan, Prof. Alex Moewes for his support in this
project, and the FQXi organizers of the 2012 contest for opening the doors to new and fresh ideas
fundamental for the progress of physics.
85
References
1. P. Weingartner, P. Mittelstaedt, Laws of Nature (Springer, Berlin, 2005)
2. M. Tegmark, Is the theory of everything merely the ultimate ensemble theory? Ann. Phys. 270,
1 (1998)
3. M. Tegmark, The mathematical universe. Found. Phys. 38, 101 (2008)
4. Sir Isaac Newton, Newtons Principia, The Mathematical Principles of the Natural Philosophy,
1st English edn. (Daniel Adee, New York, 1846)
5. E. Mach, The Science of Mechanics, 5th English edn. (The Open Court Publishing Company,
La Salle, 1942)
6. A. Einstein, Zur elektrodynamik bewegter krper. Ann. Phys. 17, 891 (1905)
7. A. Einstein, ber den einflutss der schwerkraft auf die ausbreitung des lichtes. Ann. Phys. 35,
898 (1911)
8. A. Einstein, Die grundlage der allgemeinen relativittstheorie. Ann. Phys. 49, 769 (1916)
9. A. Einstein, Hamiltons principle and the general theory of relativity. S. B. Preuss. Akad. Wiss.
(1916)
10. A. Einstein, Aether and the theory of relativity. Lectured delivered at Leyden (1920)
11. K. Popper, The Logic of Scientific Discovery, 1st edn. (Routledge Classics, London, 2009)
12. W. Heisenberg, The physical principles of the quantum theory. Z. Phys. (English Translation
1930, C. Eckart, F.C. Hoyt, 33 879 (1925))
13. H. Hertz, Electric Waves (MacMillan, London, 1900)
14. J.G. OHara, W. Pricha, Hertz and the Maxwellians (Peter Peregrinus Ltd., London, 1987)
15. B.J. Hunt, The Maxwellians (Cornell University Press, Ithaca, 1991)
16. P.A.M. Dirac, Is there an ther? Nature Lett. Editor 168, 906 (1951)
17. P.A.M. Dirac, Is there an ther? Nature Lett. Editor 169, 702 (1952)
18. M.J. Pinheiro, Do Maxwells equations need revision?a methodological note (2005),
arXiv:physics/0511103
19. J. Larmor, On a dynamical theory of the electric and luminiferous medium. Philos. Trans. R.
Soc. 190, 205300 (1897)
20. J. Larmor, Aether and Matter (C. J. Clay and Sons, Cambridge, 1900)
21. H.A. Lorentz, Simplified theory of electrical and optical phenomena in moving systems.
Zittingsverslag Akad. v. Wet 7(507), 427 (1899)
22. H.A. Lorentz, Electromagnetic phenomena in a system moving with any velocity less than that
of light. Proc., Amsterdam. 6, 809831 (1904)
23. H. Poincar, Sur la dynamique de llectron, on the dynamics of the electron. C. R. Acad. Sci
140, 1504 (1905)
24. H.E. Ives, The aberration of clocks and the clock paradox. J. Opt. Soc. Am. 27, 305 (1937)
25. H.E. Ives, G.R. Stilwell, An experimental study of the rate of a moving clock. J. Opt. Soc. Am.
28, 215 (1938)
26. G. Granek, Poincars contributions to relativistic dynamics. Stud. Hist. Philos. Mod. Phys.
31, 15 (2000)
27. J.S. Bell, Speakable and Unspeakable in Quantum Mechanics, 2nd edn. (Cambridge University
Press, Cambridge, 2004)
28. G.A. Articolo, The Michelson-Morley experiment and the phase shift upon a 90o rotation. Am.
J. Phys. 37, 215 (1969)
29. P. Mazur, On the Michelson-Morley experiment. Am. J. Phys. 37, 218 (1969)
30. M.M. Capria, F. Pambianco, On the Michelson-Morley experiment. Found. Phys. 24, 885
(1994)
31. J.D. Jackson, Classical Electrodynamics, 3rd edn. (Wiley, New York, 1999)
32. L. James, I. Nagle, G. Bearden, W.A. Zajc, Quark-gluon plasma at the RHIC and the LHC:
perfect fluid too perfect? New J. Phys. 13, 075004 (2011)
33. I. Prez, On the experimental determination of the one-way speed of light. Eur. J. Phys. 32,
9931005 (2011)
86
I. Perez
34. L.D. Landau, E.M. Lifshitz, Mechanics, Volume 1 of Course of Theoretical Physics, 3rd edn.
(Butterworth-Heinemann, Oxford, 2000)
35. A. Grnbaum, The clock paradox in the special theory of relativity. Philos. Sci. 21(3), 249253
(1954)
36. H. Ives, G.R. Stilwell, An experimental study of the rate of a moving clock II. J. Opt. Soc. Am.
31, 369 (1941)
37. A. Einstein, Kosmologische betrachtungen zur allgemeinen relativittstheorie. S.B. Preuss,
Akad. Wiss. 142157, (1917)
38. A. Einstein, S.B. Preuss, Akad. Wiss. 448, (1918)
39. W. Pauli, Theory of Relativity (Dover Publications, New York, 1958)
40. Robert M. Wald, Andreas Zoupas, General definition of conserved quantities in general
relativity and other theories of gravity. Phys. Rev. D 61, 084027 (2000)
41. L.B. Szabados, Quasi-local energy-momentum and angular momentum in general relativity.
Living Rev. Relativ. 12, 4 (2009)
42. A. Harvey, E. Schucking, Einsteins mistake and the cosmological constant. Am. J. Phys. 68,
723 (2000)
43. S. Weinberg, The cosmological constant problem. Rev. Mod. Phys. 61, 1 (1989)
44. S. Weinberg, Gravitation and Cosmology Principles and Applications of the General Theory
of Relativity (Wiley, New York, 1972)
45. H. Bondi, J. Samuel, The Lense-Thirring effect and Machs principle. Phys. Lett. A 228,
121126 (1997)
46. J. Barbour, H. Pfister, Machs Principle: From Newtons Bucket to Quantum Gravity, vol. 6
(Birkhuser, Boston, 1995)
47. S. Carroll, The cosmological constant. Living Rev. Relativ. 4, 1 (2001)
48. T. Padmanabhan, Cosmological constantthe weight of the vacuum. Phys. Rep. 380, 235320
(2003)
49. D. Dieks (ed.), The Ontology of Spacetime, vol. I & II (Elsevier, Amsterdam, 2006, 2008)
50. M. Urban, The quantum vacuum as the origin of the speed of light. Eur. J. Phys. D 67, 58
(2013)
51. G. Volovik, The Universe in a Helium Droplet (Oxford University Press, New York, 2003)
52. F. Winterberg, Relativistic quantum mechanics as a consequence of the Planck mass plasma
conjecture. Int. J. Theor. Phys. 46, 32943311 (2007)
53. B. Haisch, A. Rueda, Y. Dobyns, Inertial mass and the quantum vacuum fields. Ann. Phys. 10,
393414 (2001)
54. A.A. Michelson, H.G. Gale, F. Pearson, The effect of the earths rotation on the velocity of
light part II. Astrophys. J. 61, 140 (1925)
55. E. Kretschmann, ber den physikalischen Sinn der Relativitatspostulate. Ann. Phys. 53, 575
614 (1917)
56. J.D. Norton, General covariance and the foundations of general relativity: eight decades of
dispute. Rep. Prog. Phys. 56, 791858 (1993)
57. D. Dieks, Another look at general covariance and the equivalence of reference frames. Stud.
Hist. Philos. Mod. Phys. 37, 174191 (2006)
58. E. Constanzo, M. Consoli, From classical to modern ether-drift experiments: the narrow window
for a preferred frame. Phys. Lett. A 333, 355363 (2004)
59. T. Jacobson, D. Mattingly, Gravity with a dynamical preferred frame. Phys. Rev. D 64, 024028
(2001)
60. V. Scarani, The speed of quantum information and the preferred frame: analysis of experimental
data. Phys. Lett. A 276, 17 (2000)
61. X.-H. Ye, L. Qiang, Inhomogeneous vacuum: an alternative interpretation to curved spacetime.
Chin. Phys. Lett. 25, 15711573 (2008)
62. T. Kuhn, The Structure of Scientific Revolutions (University of Chicago Press, Chicago, 1962)
Chapter 6
Right About Time?

Sean Gryb and Flavio Mercati
Abstract Have our fundamental theories got time right? Does size really matter?
Or is physics all in the eyes of the beholder? In this essay, we question the origin
of time and scale by reevaluating the nature of measurement. We then argue for a
radical scenario, supported by a suggestive calculation, where the flow of time is
inseparable from the measurement process. Our scenario breaks the bond of time
and space and builds a new one: the marriage of time and scale.
Introduction
Near the end of the 19th century, physics appeared to be slowing down. The mechanics
of Newton and others rested on solid ground, statistical mechanics explained the
link between the microscopic and the macroscopic, Maxwells equations unified
electricity, magnetism, and light, and the steam engine had transformed society. But
the blade of progress is double edged and, as more problems were sliced through,
fewer legitimate fundamental issues remained. Physics, it seemed, was nearing an
end.
Or was it? Among the few remaining unsolved issues were two experimental
anomalies. As Lord Kelvin allegedly announced: The beauty and clearness of the
dynamical theory [...] is at present obscured by two clouds [1]. One of these clouds
was the ultra-violet catastrophe: an embarrassing prediction that hot objects like the
S. Gryb (B)
Institute for Theoretical Physics, Utrecht University, Leuvenlaan 4,
3584 CE Utrecht, The Netherlands
S. Gryb
Institute for Mathematics, Astrophysics and Particle Physics,
Radboud University, Huygens Building, Heyendaalseweg 135, 6525 AJ
Nijmegen, The Netherlands
F. Mercati
Perimeter Institute for Theoretical Physics, 31 Caroline Street North,
Waterloo, AJ N2L 2Y5, Canada
87
88
S. Gryb and F. Mercati
sun should emit infinite energy. The other anomaly was an experiment by Michelson
and Morley that measured the speed of light to be independent of how an observer
was moving. Given the tremendous success of physics at that time, it would have
been a safe bet that, soon, even these clouds would pass.
Never bet on a sure thing. The ultra-violet catastrophe led to the development of
quantum mechanics and the MichelsonMorley experiment led to the development
of relativity. These discoveries completely overturned our understanding of space,
time, measurement, and the perception of reality. Physics was not over, it was just
getting started.
Fast-forward a hundred years or so. Quantum mechanics and relativity rest on
solid ground. The microchip and GPS have transformed society. These frameworks
have led to an understanding that spans from the microscopic constituents of the
nucleus to the large scale structure of the Universe. The corresponding models have
become so widely accepted and successful that they have been dubbed standard
models of particle physics and cosmology. Resultantly, the number of truly interesting
questions appears to be slowly disappearing. In well over 30 years, there have been
no experimental results in particle physics that cannot be explained within the basic
framework laid out by the standard model of particle physics. With the ever increasing
cost of particle physics experiments, it seems that the data is drying up. But without
input from experiment, how can physics proceed? It would appear that physics is,
again, in danger of slowing down.
Or is it? Although the number of interesting fundamental questions appears to be
decreasing, the importance of the remaining questions is growing. Consider two of the
more disturbing experimental anomalies. The first is the naturalness problem, i.e., the
presence of unnaturally large and small numbers in Nature. The most embarrassing
of these numbersand arguably the worst prediction of scienceis the accelerated
expansion of the Universe, which is some 120 orders of magnitude smaller than its
natural value. The second is the dark matter problem that just under 8590 % of the
matter content of our Universe is of an exotic nature that we have not yet seen in the
lab. It would seem that we actually understand very little of what is happening in our
Universe!
The problem is not that we dont have enough data. The problem is that the data we
do have does not seem to be amenable to explanation through incremental theoretical
progress. The belief that physics is slowing down or, worse, that we are close to a
final theory is just as as unimaginative now as it would have been before 1900. Our
thesis here will be that the lesson to take from that period is that the way forward is
to question the fundamental assumptions of our physical theories in a radical way.
This is easier said than done: one must not throw out the baby with the bath water.
What is needed is a careful examination of our physical principles in the context of
real experimental facts to explain more data using less assumptions.
The purpose of this work is to point out three specific assumptions made by our
physical theories that might be wrong. We will not offer a definite solution to these
problems but suggest a new scenario, supported by a suggestive calculation, that puts
these assumptions into a new light and unifies them. The three assumptions we will
question are
6 Right About Time?
89
1. Time and space are unified.

2. Scale is physical.
3. Physical laws are independent of the measurement process.
We will argue that these three assumptions inadvertently violate the same principle:
the requirement that the laws of physics depend only on what is knowable through
direct measurement. They fall into a unique category of assumptions that are challenged when we ask how to adapt the scientific method, developed for understanding
processes in the lab, to the cosmological setting. In other words, how can we do science on the Universe as a whole?
We will not directly answer this question but, rather, suggest that this difficult
issue may require a radical answer that questions the very origin of time. The flow
of time, we will argue, may be fundamentally linked to the process of measurement.
We will then support this argument with an intriguing calculation that recovers the
black hole entropy law from a simple toy model. Before getting to this, let us explain
the three questionable assumptions.
Three Questionable Assumptions

Many of our most basic physical assumptions are made in the first week of physics
education. A good example is one of the first equations we are taught: the definition
of velocity,
v=
x
.
t
(6.1)
It is perhaps a bit over-dramaticbut, at the same time, not inaccurateto say that
to give this equation a precise operational meaning has been an outstanding issue in
physics for its entire history. This is because, to understand this equation, one has
to have an operational definition of both x, t, and . Great minds have pondered
this question and their insights has led to scientific revolutions. This includes the
development of Newtonian mechanics, relativity, and quantum mechanics.1 Recently,
the meaning of x and, in particular, t, have been the subject of a new debate whose
origin is in a theory of quantum gravity. This brings us to our first questionable
assumption.
Time and Space Are Unified

The theory of relativity changed our perception of time. As Minkowski put it in 1908
[2], space by itself, and time by itself, are doomed to fade away into mere shadows,
1
A lot to digest in the first week!
90
and only a kind of union of the two will preserve an independent reality. Nowhere is
this more apparent than in the main equation physicists use to construct the solutions
of general relativity (GR):

SEinstein-Hilbert =
d 4 x (R + Lmatter )
g .
(6.2)
Can you spot the t? Its hidden in the 4 of d 4 x. But there are important structures
hidden by this compact notation.
We will start by pointing out an invisible minus sign in Eq. (6.2). When calculating
spacetime distances, one needs to use
x 2 + y2 + z2 t 2,
(6.3)
which has ain front of the t 2 instead of Pythagoras +. The minus sign looks
innocent but has important consequences for the solutions of Eq. (6.2). Importantly,
the minus sign implies causal structure, which means that only events close enough
to us so that light signals sent from these events can make it to us now can effect
what is going on now. This, in turn, implies that generic solutions of GR can only be
solved by specifying information at a particular time and then seeing how this information propagates into the future. Doing the converse, i.e., specifying information
at a particular place and seeing how that information propagates to another place, is,
in general, not consistent.2 Thus, the minus sign already tells you that you have to
use the theory in a way that treats time and space differently.
There are other ways to see how time and space are treated differently in gravity.
In Julian Barbours 2009 essay, The Nature of Time [3], he points out that Newtons
absolute time is not absolute at all. Indeed, the Newtonian notion of duration
that is, how much time has ticked by between two distinct instantscan be inferred
by the total change in the spatial separations of particles in the Universe. He derives
the equation
t 2
di2 ,
(6.4)
where the di are inter-particle separations in units where the masses of the particles
are one. The factor of proportionality is important, but not for our argument. What
is important is that changes in time can be inferred by changes in distances so that
absolute duration is not an input of the classical theory. This equation can be generalized to gravity where it must be solved at every point in space. The implications
for the quantum theory are severe: time completely drops out of the formalism.
Expert readers will recognize this as one of the facets of the Problem of Time [4].
The fact that there is no equivalent Problem of Space can be easily traced back to
the points just made: time is singled out in gravity as the variable in terms of which
2
Technically, the difference is in the elliptic versus hyperbolic nature of the evolution equations.
6 Right About Time?
91
the evolution equations are solved. This in turn implies that local duration should
be treated as an inferred quantity rather than something fundamental. Clearly, time
and space are not treated on the same footing in the formalism of GR despite the
rather misleading form of Eq. (6.2). Nevertheless, it is still true that the spacetime
framework is incredibly useful and, as far as we know, correct. How can one reconcile
this fact with the space-time asymmetry in the formalism itself? We will investigate
this in Sect. Time from Coarse Graining.
Scale Is Physical
Before even learning the definition of velocity, the novice physicist is typically introduced to an even more primary concept that usually makes up ones first physics
lesson: units. Despite the rudimentary nature of units, they are probably the most
inconsistently understood concept in all of physics. If you ask ten different physicists for the physical meaning of a unit, you will likely get ten different answers. To
avoid confusion, most theoreticians set all dimensionful constants equal to 1. However, one cant predict anything until one has painfully reinserted these dimensionful
quantities into the final result.
And yet, no one has ever directly observed a dimensionful quantity. This is because
all measurements are comparisons. A meter has no intrinsic operational meaning,
only the ratio of two lengths does. One can define an object A to have a length of
one meter and make a measurement that reveals that some other object B has twice
the length of object A. Then, we can deduce that object B has a length of 2 meters.
This, however, tells you nothing about the intrinsic absolute length of object A for
if a demon doubled the intrinsic size of the Universe, the result of the experiment
would be exactly the same. So, where do units come from?
Some units, like the unit of pressure, are the result of emergent physics. We
understand how they are related to more fundamental units like meters and seconds.
However, even our most fundamental theories of Nature have dimensionful quantities
in them. The standard model of particle physics and classical GR require only a
singe unit: mass. Scale or, more technically, conformal invariance is then broken by
only two quantities with the units of mass. The first is the recently observed Higgs
mass, which can be related to all the masses of the particles in the standard model.
The second is the Plank mass, which sets the scale of quantum gravity. As already
discussed, there is a naturalness problem associated with writing all other constants
of nature as dimensionless quantities but this will not bother us to much here.
The presence of dimensionful quantities is an indication that our fundamental
theories are not fundamental at all. Instead, scale independence should be a basic principle of a fundamental theory. As we will see in Sect. Time from Coarse
Graining, there is a formulation of gravity that is nearly scale invariant. We will try
to address the nearly with the considerations of the next section.
92
Physical Laws Are Independent of the Measurement Process

There is one assumption that is so fundamental it doesnt even enter the physics
curriculum. This is the assumption that the scientific method is generally applicable
for describing everything in the Universe taken together. We know that the scientific method can be applied in the laboratory where external agents (i.e., scientists)
carefully control the inputs of some subsystem of the Universe and observe the subsystems response to these inputs. We dont know, however, whether it is possible to
apply these techniques to the Universe as a whole. On the other hand, when it comes to
quantum mechanics, we do know whether our formalism can be consistently applied
to the Universe. The answer is NO! The reasons are well understoodif not disappointingly under appreciatedand the problem even has a name: the measurement
problem.
The measurement problem results from the fact that quantum mechanics is
a framework more like statistical physics than classical mechanics. In statistical
physics, one has practical limitations on ones knowledge of a system so one takes
an educated guess at the results of a specific experiment by calculating a probability distribution for the outcome using ones current knowledge of the system. In
quantum mechanics, one has fundamental limitations on ones knowledge of the
systemessentially because of the uncertainty principleso one can only make an
educated guess at the outcome of a specific experiment by calculating a probability
distribution for the outcome using ones current knowledge of the system. However,
it would be strange to apply statistical mechanics to the whole Universe because
the Universe itself is only given once. It is hard to imagine what an ensemble of
Universes, for which one can calculate and give meaning to a probability distribution, would even mean.3 The same is true in quantum mechanics, but the problem is
worse. The framework itself is designed to give you a probability distribution for the
outcome of some measurement but how does one even define a measurement when
the observer itself is taken to be part of the system? The answer is not found in any
interpretation of quantum mechanics, although the problem itself takes a different
form in a given interpretation. The truth is that quantum mechanics requires some
additional structure, which can be thought of as describing the observer, in order
for it to make sense. In other words, quantum mechanics alone, without additional
postulates, can never be a theory of the whole Universe.
As a consequence of this, any approach to quantum gravity that uses quantum
mechanics unmodifiedincluding all major approaches to quantum gravityis not,
and can never be a theory of the whole Universe. It could still be used for describing
quantum gravity effects on isolated subsystems of the Universe, but that is not the
3 This is one of the goals of the Many Worlds interpretation of quantum mechanics whose proponents
believe that it is possible to make sense of such an ensemble using the standard axioms of classical
probability theory (see [5] for a popular account). Whether it is sensible to apply these axioms to the
Universe as a whole, however, is unclear. Furthermore, having to believe in an infinite number of
unobservable parallel Universes is a big price to pay just to make sense of probabilities in quantum
mechanics.
6 Right About Time?
93
ambition of a full fledged quantum gravity theory. Given such a glaring foundational
issue at the core of every major approach to quantum gravity, we believe that the
attitude that we are nearing the end of physics is unjustified. The shut-up and
calculate era is over. It is time for the quantum gravity community to return to these
fundamental issues.
One approach is to change the ambitions of science. This is the safest and, in some
ways, easiest option, but it would mean that science is inherently a restricted framework. The other possibility is to try to address the measurement problem directly. In
the next section, we will give a radical proposal that embraces the role of the observer
in our fundamental description of Nature. To understand how this comes about, we
need one last ingredient: renormalization, or the art of averaging.
A Way Forward
The Art of Averaging
It is somewhat unfortunate that the great discoveries of the first half of the 20th century
have overshadowed those of the second half of the century. One of these, the theory of
renormalization, is potentially the uncelebrated triumph of twentieth century physics.
Renormalization was born as rather ugly set of rules for removing some undesirable
features of quantum field theories. From these humble beginnings, it has grown into
one of the gems of physics. In its modern form due to Wilson [6], renormalization has
become a powerful tool for understanding what happens in a general system when
one lacks information about the details of its fine behavior. Renormalizations reach
extends far beyond particle physics and explains, among other things, what happens
during phase transitions. But, the theory of renormalization does even more: it helps
us understand why physics is possible at all.
Imagine what it would be like if, to calculate everyday physics like the trajectory
of Newtons apple, one would have to compute the motions of every quark, gluon, and
electron in the apple and use quantum gravity to determine the trajectory. This would
be completely impractical. Fortunately, one doesnt have to resort to this. High-school
physics is sufficient to determine the motion of what is, fundamentally, an incredibly
complicated system. This is possible because one can average, or coarse grain,
over the detailed behavior of the microscopic components of the apple. Remarkably,
the average motion is simple. This fact is the reason why Newtonian mechanics is
expressible in terms of simple differential equations and why the standard model is
made up of only a couple of interactions. In short, it is why physics is possible at all.
The theory of renormalization provides a framework for understanding this.
The main idea behind renormalization is to be able to predict how the laws of
physics will change when a coarse graining is performed. This is similar to what
happens when one changes the magnification of a telescope. With a large magnification, one might be able to see the moons of Jupiter and some details of the structure of
94
their atmospheres. But, if the magnification, or the renormalization scale, is steadily

decreased, the resolution is no longer good enough to make out individual moons and
the lens averages over these structures. The whole of Jupiter and its moons becomes
a single dot. As we vary the renormalization scale, the laws of physics that govern the
structures of the system change from the hydrodynamic laws governing atmospheres
to Newtons law of gravity.
The theory of renormalization produces precise equations that say how the laws
of physics will change, or flow, as we change the renormalization scale. In what
follows, we will propose that flow under changes of scale may be related to the flow
of time.
Time from Coarse Graining

We are now prepared to discuss an idea that puts our three questionable assumptions
into a new light by highlighting a way in which they are connected. First, we point out
that there is a way to trade a spacetime symmetry for conformal symmetry without
altering the physical structures of GR. This approach, called Shape Dynamics (SD),
was initially advocated by Barbour [7] and was developed in [8, 9]. Symmetry trading
is allowed because symmetries dont affect the physical content of a theory. In SD,
the irrelevance of duration in GR is traded for local scale invariance (we will come
to the word local in a moment). This can be done without altering the physical
predictions of the theory but at the cost of having to treat time and space on a different
footing. In fact, the local scale invariance is only an invariance of space, so that local
rodsnot clockscan be rescaled arbitrarily. Time, on the other hand, is treated
differently. It is a global notion that depends on the total change in the Universe.
The equivalence between SD and GR is a rather remarkable thing. What can
be proved is that a very large class of spacetimes that are solutions of GR can be
reproduced by a framework that does not treat spacetime as fundamental. Instead,
what is fundamental in SD is scale-invariant geometry. Recently [10], it has been
discovered that some solutions of SD do not actually correspond to spacetimes as
all, although they are still in agreement with experiment. These are solutions that
describe certain kinds black holes in SD. In these solutions, there is no singularity
where the curvature of spacetime becomes infinite. Rather, there is a traversable
worm hole that connects the event horizon of a black hole to another region of space.
This exciting discovery could pave the way to a completely different understanding
of black holes.
Symmetry trading is the key to understanding how GR and SD are related. In 2
spatial dimensions, we know that this trading is possible because of an accidental
mathematical relationship between the structure of conformal symmetry in 2 dimensions and the symmetries of 3 dimensional spacetime [11].4 We are investigating
4
Technically, this is the isomorphism between the conformal group in d spatial dimensions and the
deSitter group in d + 1 dimensions.
6 Right About Time?
95
whether this result will remain true in 3 spatial dimensions. If it does, it would mean
that the spacetime picture and the conformal picture can coexist because of a mere
mathematical accident.
We now come to a key point: in order for any time evolution to survive in SD, one
cannot eliminate all of the scale. The global scale of the Universe cannot be traded
since, then, no time would flow. Only a redistribution of scale from point to point
is allowed (this is the significance of the word local) but the overall size of the
Universe cannot be traded. In other words, global scale must remain for change to
be possible. How can we understand this global scale?
Consider a world with no scale and no time. In this world, only 3 dimensional
Platonic shapes exist. This kind of world has a technical name, it is a fixed point of
renormalizationfixed because such a world does not flow since the renormalization scale is meaningless. This cannot yet be our world because nothing happens
in this world. Now, allow for something to happen and call this something a measurement. One thing we know about measurements is that they can never be perfect.
We can only compare the smallest objects of our device to larger objects and coarse
grain the rest. Try as we may, we can never fully resolve the Platonic shapes of the
fixed point. Thus, coarse graining by real measurements produces flow away from
the fixed point. But what about time? How can a measurement happen if no time has
gone by? The scenario that we are suggesting is that the flow under the renormalization scale is exchangeable with the flow of time. Using the trading procedure of SD,
the flow of time might be relatable to renormalization away from a theory of pure
shape.
In this picture, time and measurement are inseparable. Like a diamond with many
faces, scale and time are different reflections of a single entity. This scenario requires
a radical revaluation of our notions of time, scale, and measurement.
To be sure, a lot of thought is still needed to turn this into a coherent picture.
A couple of comments are in order. Firstly, some authors [12, 13] have investigated a similar scenario, called holographic cosmology using something called
gauge/gravity duality. However, our approach suggests that one may not have to
assume gauge/gravity duality for this scenario but, instead, can make use of symmetry trading in SD. Furthermore, our motivation and our method of implementation
is more concrete. Secondly, in the context of scale-invariant particle toy models,
Barbour, Lostaglio, and one of the authors [14] have investigated a scenario where
quantum effects ruin the classical scale invariance. In these models, the quantum
theory has an emergent scale, which can then be used as a clock that measures the
quantum time evolution of the scale invariant shapes of the system. This simple model
illustrates one way in which the radical scenario discussed here could implemented
into a concrete theory. Finally, why should we expect that there is enough structure in
a coarse graining of pure shapes to recover the rich structure of spacetime? A simple
answer is the subject of the next section.
96
The Size that Matters

In this section, we perform a simple calculation suggesting that the coarse graining
of shapes described in the last section could lead to gravity. This section is more
technical than the others but this is necessary to set up our final result. Brave souls
can find the details of the calculations in the Technical Appendix.
We will consider a simple toy model that, remarkably, recovers a key feature of
gravity. Before getting into the details of the model, we should quickly point out that
this model should be taken more as an illustration of one way in which it is possible
to define the notion of a coarse graining on shape space. The model should not be
taken as a literal model for gravity or black holes, even though some of the results
seem suggestive in this regard. Certainly much more work would be needed to flesh
this out in a convincing way.
The model we will consider is a set of N free Newtonian point particles. To
describe the calculation we will need to talk about two spaces: Shape Space and
Extended Configuration Space (ECS). Shape Space is the space of all the shapes of
the system. If N = 3, this is the space of all triangles. ECS is the space of all Cartesian
coordinates of the particles. That is, the space of all ways you can put a shape into
a Cartesian coordinate system. The ECS is larger than Shape Space because it has
information about the position, orientation, and size of the shapes. Although this
information is unphysical, it is convenient to work with it anyway because the math
is simpler. This is called a gauge theory. We can work with gauge theories provided
we remove, or quotient, out the unphysical information. To understand how this is
done, examine Fig. (6.1) which shows schematically the relation between the ECS
and Shape Space. Each point on Shape Space is a different shape of the system, like
Fig. 6.1 Each point in Shape Space is a different shape (represented by triangles). These correspond
to an equivalence class (represented by arrows) of points of the Extended Configuration Space
describing the same shape with a different position, orientation, and size
6 Right About Time?
97
a triangle. All the points along the arrows represent the same shape with a different
position, orientation, or size. By picking a representative point along each arrow, we
get a 1to1 correspondence between ECS and Shape Space. This is called picking
a gauge. Mathematically, this is done by imposing constraints on the ECS. In our
case, we need to specify a constraint that will select a triangle with a certain center
of mass, orientation, and size. For technical reasons, we will assume that all particles
are confined to a line so that we dont have to worry about orientation. To specify
the size of the system, we can take the length of the system, R, on ECS. This is
the moment of inertia. By fixing the center of mass and moment of inertia in ECS,
we can work indirectly with Shape Space. The main advantage of doing this is that
there is a natural notion of distance in ECS. This can be used to define the distance
between two shapes, which is a key input of our calculations.
To describe the calculation, we need to specify a notion of entropy in Shape
Space. Entropy can be thought of as the amount of information needed to specify a
particular macroscopic state of the system. To make this precise, we can use the notion
of distance on ECS to calculate a volume on Shape Space. This volume roughly
corresponds to the number of shapes that satisfy a particular property describing the
state. The more shapes that have this property, the more information is needed to
specify the state. The entropy of that state is then related to its volume, m , divided
by the total volume of Shape Space, tot . Explicitly,
S = kB log
m
,
tot
(6.5)
where kB is Boltzmanns constant.

We will be interested in states described by a subsystem of n < N particles
that have a certain center of mass x 0 and moment of inertia, r . To make sense
of the volume, we need a familiar concept: coarse graining. We can approximate
the volume of the state by chopping up the ECS into a grid of size . Physically,
the coarse graining means that we have a measuring device with a finite resolution given by . Consider a state that is represented by some surface in ECS. This
is illustrated in Fig. (6.2) by a line. The volume of the state is well approximated
by counting the number of dark squares intersected by the line. In the Technical
Appendix, we calculate this volume explicitly. The result is
Fig. 6.2 Left Approximation

of a line using a grid. Right
Further approximation of the
line as a strip of thickness
equal to the grid spacing
98

m 2 r n2

R2 r 2 1 +
m
M m
m 2
x
M 0
N n2
2
(6.6)
where M and R are the total mass and moment of inertia of the whole system and m
is the mass of the subsystem. We can then compare this volume to the total volume
of Shape Space, which goes like the volume of an N 1 dimensional sphere (the
1 is because of the center of mass gauge fixing). Thus,
tot R N 1 .
(6.7)
The resulting entropy is

S=
N r 2
r
1
kB
k B log + .
2
n R
R
(6.8)
Remarkably, the first term is exactly the entropy of a black hole calculated by
Bekenstein and Hawking [15, 16]. More remarkably, the second term is exactly
the first correction to the BekensteinHawking result calculated in field theory [17,
18]. However, one should be careful not to interpret this result too literally. After all,
we are considering only a very simplified case. A much more detailed analysis is
necessary to draw any conclusions from this about real black holes. Note, however,
that Erik Verlinde [19] discovered a way to interpret Newtonian gravity as an entropic
force for systems whose entropy behaves in this way. It would appear that this simple model of a coarse graining of pure shapes has the right structure to reproduce
Newtonian gravity.
Conclusions
We have questioned the basic assumptions that: (i) time and space should be treated
on the same footing, (ii) scale should enter our fundamental theories of Nature, and
(iii) the evolution of the Universe is independent of the measurement process. This
has led us to a radical proposal: that time and scale emerge from a coarse graining of
a theory of pure shape. The possibility that gravity could come out of this formalism
was suggested by a simple toy model. The results of this model are nontrivial. The
key result was that the entropy (6.8) scales like r 2 , which, dimensionally, is an area.
In three dimensions, this is the signature of holography. Thus, in this simple model,
Shape Space is holographic. If this is a generic feature of Shape Space, it would be
an important observation for quantum gravity.
Moreover, the toy model may shed light on the nature of the Plank length. In this
model, the Plank length is the emergent length arising in ECS given by
L 2Planck = G
R2
.
N
(6.9)
6 Right About Time?
99
This dimensionful quantity, however, is not observable in this model. What is

physical, instead, it the dimensionless ratio r/R. This illustrates how a dimensionful
quantity can emerge from a scale independent framework. Size doesnt matterbut
a ratio of sizes does. The proof could be gravity.
Technical Appendix
The extended configuration space is R N : the space coordinates, ri , (i = 1, . . . , N ) of
N particles in 1 dimension. To represent the reduced configuration space, or Shape
Space, we can use a gauge fixing surface. To fix the translations, we can fix the center
of mass to be at the origin of the coordinate system:
N
m i ri = 0 .
(center of mass at the origin)
(6.10)
i=1
The equation above gives three constraints selecting three orthogonal planes through
the origin whose orientation is determined by the masses m i . A natural gauge-fixing
for the generators of dilatations is to set the moment of inertia with respect to the
center of mass to a constant5 (the weak equation holds when the gauge-fixing (6.10)
is applied):
N
mi m j

mi
2
|ri |2 = R 2 .
|r
r
|
i
j
2
M
M
i< j
( f i xed moment o f iner tia)
i=1
(6.11)
The last relation defines a sphere in R N centered at the origin. Thus, Shape Space
is the intersection of the N 1-dimensional sphere (6.11) with the three orthogonal
planes (6.10).
The flat Euclidean metric, ds 2 = m i i j ab dria dr bj , is the natural metric on the
extended configuration space Q. This metric induces the non-flat metric

2
= m i i j ab dria dr bj
dsinduced
QS
(6.12)
on Shape Space.
Description of a Macrostate in Shape Space

Consider an N -particle toy Universe with an n-particle subsystem, n < N . The
particles in the subsystem have coordinates xi = ri , (i = 1, . . . , n), while the
5
We are using here

the notion of moment of inertia with respect to a point, which we rescaled by
the total mass M = i m i to give it the dimensions of a squared length.
100
coordinates of all the other particles will be called yi = rn+i , (i = 1, . . . , N n).

It is useful to define the coordinates of the center of mass of the subsystem and of
the rest of the Universe6 :
x0 =
n

mi
i=1
xi ,
y0 =
N
n

i=1
m n+i
yi ,
M m
m=
n
mi ,
(6.13)
m n+i
|yi y0 |2 .
M
(6.14)
i=1
and the center-of-mass moment of inertia of the two subsystems

r=
n

mi
i=1
|xi x0 |2 ,
r =
N
n

i=1
The relation between the moments of inertia of the total system and those of the two
subsystems is

R 2 = r 2 + (r )2 + 1 +
m
M m
m 2
x .
M 0
(6.15)
We define a macrostate as a state in which the moment of inertia of the subsystem,

r , and its center of mass, x 0 , are constant. To calculate the Shape Space volume of
such a macrostate, we must integrate over all Shape Space coordinates x i and yi that
respect the conditions (6.13), (6.14), and (6.15) using the measure provided by the
induced metric (6.12). Lets make the following change of variables:
xi =
m i (xi x0 ) ,
yi =
m n+i (yi y0 ) .
(6.16)
Our equations become

1
m
r=
1
M
n

i=1
m i xi = 0,
i=1
xi2 , r =
1
M
N
n
i=1
1
Mm
n

m n+i yi = 0,
i=1

yi2 , R 2 = r 2 + (r )2 + 1 +
m
Mm
m
M
x02 .
(6.17)
In the new coordinates, the metric is the identity matrix (it loses the m i factors
on the diagonal). The integral is over the direct product of an (n 2)-dimensional
sphere
(N n 2)-dimensional sphere of radius Mr =
of radius Mr and an

m
m 2
M R 2 r 2 1 + Mm
M x 0 whose volume (calculated with a coarse-graining
of size ) is:
Notice that the two sets of coordinates must satisfy the relation m x0 + (M m)y0 = 0 in order
to keep the total center of mass at the origin.
6 Right About Time?
101
4 (N n1)/2 (n1)/2
M N 4 r n2
((N n 1)/2)((n 1)/2)
N n2

2
m 2
m
2
2
x0
R r 1+
.
M m M
m = 2
(6.18)
The total volume of Shape Space is that of an (N 1)-dimensional sphere of

radius M R
tot =
2 N /2
M N 1 R N 1 .
(N /2)
(6.19)
Thus, the Shape Space volume per particle, in the limit 1 n N , r r , m M

reduces to
N

2/n
r 2
m x0 2 2n
r

m
1
1+
,
r
R
R
M m M R
(6.20)
and its logarithm has the expansion (remember that x0 < R)

S=
2
1 N r 2
r

kB
k B log k B log + .
2 n R
R
n
r
(6.21)
Notice that the numerical factors change in the 3 dimensions. In that case, they are
S=
3
4
N r 2
r

kB
3 k B log k B log .
2
n R
R
n
r
(6.22)
References
1. L. Kelvin, Nineteenth Century Clouds over the Dynamical Theory of Heat and Light, Philosophical Magazine, Sixth Series 2 140 (1901). From a 1900, April 27, Royal Institution
lecture
2. H. Minkowski, Space and time, The principle of relativity: a collection of original memoirs on
the special and general theory of relativity (Dover, New York, 1952), pp. 7591
3. J. Barbour, The Nature of Time. arXiv:0903.3489 [gr-qc]
4. C.J. Isham, Canonical quantum gravity and the problem of time. arXiv:gr-qc/9210011
5. D. Wallace, The Emergent Multiverse: Quantum Theory according to the Everett Interpretation
(Oxford University Press, Oxford, 2012)
6. K. Wilson, J.B. Kogut, The Renormalization group and the epsilon expansion. Phys. Rept. 12,
75200 (1974)
7. J. Barbour, Dynamics of pure shape, relativity and the problem of time, in Decoherence and
Entropy in Complex Systems: Proceedings of the Conference DICE, Piombino 2002, ed. by
H.-T Elze. Springer Lecture Notes n Physics (2003)
8. H. Gomes, S. Gryb, T. Koslowski, Einstein gravity as a 3D conformally invariant theory. Class.
Quant. Grav. 28, 045005 (2011) arXiv:1010.2481 [gr-qc]
102
9. H. Gomes, S. Gryb, T. Koslowski, The link between general relativity and shape dynamics.
Class. Quant. Grav. 29, 075009 arXiv:1101.5974 [gr-qc]
10. H. Gomes, A birkhoff theorem for shape dynamics. Class. Quant. Grav. 31, 085008 (2014)
arXiv:1305.0310 [gr-qc]
11. S. Gryb, F. Mercati, 2 + 1 gravity as a conformal gauge theory in 2 dimensions (in preparation)
12. A. Strominger, Inflation and the dS/CFT correspondence, JHEP 0111 049, (2001).
arXiv:hep-th/0110087 [hep-th]
13. P. McFadden K. Skenderis, The holographic universe. J. Phys. Conf. Ser. 222, 012007 (2010)
arXiv:1001.2007 [hep-th]
14. J. Barbour, M. Lostaglio, F. Mercati, Scale anomaly as the origin of time. Gen. Rel. Grav. 45,
911938 (2013) arXiv:1301.6173 [gr-qc]
15. J.M. Bardeen, B. Carter, S.W. Hawking, The four laws of black hole mechanics. Commun.
Math. Phys. 31, 161170 (1973)
16. J.D. Bekenstein, Black holes and entropy. Phys. Rev. 7, 23332346 (1973)
17. K.S. Gupta, S. Sen, Further evidence for the conformal structure of a schwarzschild black hole
in an algebraic approach. Phys. Lett. B526, 121126 (2002) arXiv:hep-th/0112041
18. D. Birmingham, S. Sen, An exact black hole entropy bound. Phys. Rev. D63, 047501 (2001)
arXiv:hep-th/0008051
19. E.P. Verlinde, On the origin of gravity and the laws of Newton. JHEP 04, 029 (2011)
arXiv:1001.0785 [hep-th]
Chapter 7
A Critical Look at the Standard

Cosmological Picture
Daryl Janzen
Abstract The discovery that the Universe is accelerating in its expansion has
brought the basic concept of cosmic expansion into question. An analysis of the
evolution of this concept suggests that the paradigm that was finally settled into prior
to that discovery was not the best option, as the observed acceleration lends empirical support to an alternative which could incidentally explain expansion in general.
I suggest, then, that incomplete reasoning regarding the nature of cosmic time in
the derivation of the standard model is the reason why the theory cannot coincide
with this alternative concept. Therefore, through an investigation of the theoretical
and empirical facts surrounding the nature of cosmic time, I argue that an enduring three-dimensional cosmic present must necessarily be assumed in relativistic
cosmologyand in a stricter sense than it has been. Finally, I point to a related result
which could offer a better explanation of the empirically constrained expansion rate.
Introduction
Many of our basic conceptions about the nature of physical reality inevitably turn
out to have been false, as novel empirical evidence is obtained, or paradoxical implications stemming from those concepts are eventually realised. This was expressed
well by Einstein, who wrote [1]
What is essential, which is based solely on accidents of development?Concepts that have
proven useful in the order of things, easily attain such an authority over us that we forget
their Earthly origins and accept them as unalterable facts.The path of scientific advance
is often made impassable for a long time through such errors. It is therefore by no means
an idle trifling, if we become practiced in analysing the long-familiar concepts, and show
upon which circumstances their justification and applicability depend, as they have grown
up, individually, from the facts of experience.
Or, as he put it some years later [2],

D. Janzen (B)
Department of Physics and Engineering Physics,
University of Saskatchewan, 116 Science Place, Rm 163, Saskatoon, SK S7N 5E2, Canada
103
104
D. Janzen
The belief in an external world independent of the percipient subject is the foundation of all
science. But since our sense-perceptions inform us only indirectly of this external world, or
Physical Reality, it is only by speculation that it can become comprehensible to us. From this
it follows that our conceptions of Physical Reality can never be definitive; we must always
be ready to alter them, to alter, that is, the axiomatic basis of physics, in order to take account
of the facts of perception with the greatest possible logical completeness.
And so it is in the same spirit, that I shall argue against a number of concepts in
the standard cosmological picture that have changed very little in the past century,
by making note of original justifications upon which they were based, and weighing
those against empirical data and theoretical developments that have been realised
through the intervening years.
The essay will concentrate initially on the nature of cosmic expansion, which
lacks an explanation in the standard cosmological model. Through a discussion of
the early developments in cosmology, a familiarity with the pioneering conception of
expansion, as being always driven by a cosmological constant , will be developed,
upon which basis it will be argued that the standard modelwhich cannot reconcile
with this viewaffords only a very limited description. Then, the nature of time in
relativistic cosmology will be addressed, particularly with regard to the formulation
of Weyls postulate of a cosmic rest-frame. The aim will therefore be towards a
better explanation of cosmic expansion in general, along with the present acceleration that has recently become evident, by reconceiving the description of time in
standard cosmology, as an approach to resolving this significant shortcoming of the
big bang Friedman-Lematre-Robertson-Walker (FLRW) models, and particularly
the flat CDM model that describes the data so well.
On Cosmic Expansion
The expansion of our Universe was first evidenced by redshift measurements of spiral
nebulae, after the task of measuring their radial velocities was initiated in 1912 by
Slipher; and shortly thereafter, de Sitter attempted the first relativistic interpretation
of the observed shifts, noting that the frequency of light-vibrations diminishes with
increasing distance from the origin of co-ordinates due to the coefficient of the timecoordinate in his solution [3]. But the concept of an expanding Universe, filled with
island galaxies that would all appear to be receding from any given location at rates
increasing with distance, was yet to fully form.
For one thing, when de Sitter published his paper, he was able to quote only three
reliable radial velocity measurements, which gave merely 2:1 odds in favour of his
prediction. However, in 1923 Eddington produced an updated analysis of de Sitter
space, and showed that the redshift de Sitter had predicted as a phenomenon of his
statical geometry was in fact due to a cosmical repulsion brought in by the -term,
which would cause inertial particles to all recede exponentially from any one [4]. He
used this result to support an argument for a truly expanding Universe, which would
7 A Critical Look at the Standard Cosmological Picture
105
expand everywhere and at all times due to . This, he supported with an updated
list of redshifts from Slipher, which now gave 36:5 odds in favour of the expansion
scenario.
That same year, Weyl published a third appendix to his Raum, Zeit, Materie,
and an accompanying paper [5], where he calculated the redshift for the de Sitter
cosmology,
ds = dt + e
2

3t
(dx 2 + dy 2 + dz 2 ),
(7.1)
the explicit form of which would only be found later, independently by Lematre [6]
and Robertson [7]. Weyl was as interested in the potential relevance of de Sitters
solution for an expanding cosmology as Eddington [5], and had indeed been confused
when he received a postcard from Einstein later that year (Einstein Archives: [24
81.00]), stating,
With reference to the cosmological problem, I am not of your opinion. Following de Sitter,
we know that two sufficiently separate material points are accelerated from one another. If
there is no quasi-static world, then away with the cosmological term.
Eight days after this was posted, Einsteins famous second note [8] on Friedmans
paper, which he now referred to as correct and clarifying, arrived at Zeitschrift f r
Physik. Einstein evidently had in mind that the cosmic expansion can be described
with set to zero in Friedmans solution, and he might have thought Weyl would
notice [8] and make the connectionbut the latter evidently did not, as he wrote
a dialogue the following year [9] in which the proponent of orthodox relativity1
eventually states, If the cosmological term fails to help with leading through to
Machs principle, then I consider it to be generally useless, and am for the return to
the elementary cosmologythat being a particular foliation of Minkowski space,
which, of the three cosmological models known to Weyl, was the only one with
vanishing .
At this point in the dialogue, the protagonist Paulus perseveres, citing the evidence
for an expanding Universe, and therefore the de Sitter cosmology as the most likely
of the three known alternatives. Weyls excitement over its description is evident
in Paulus final statement: If I think about how, on the de Sitter hyperboloid the
world lines of a star system with a common asymptote rise up from the infinite past
(see Fig. 7.1), then I would like to say: the World is born from the eternal repose of
Father ther; but once disturbed by the Spirit of Unrest (Hlderlin), which is
at home in the Agent of Matter, in the breast of the Earth and Man, it will never
come again to rest. Indeed, as Eq. (7.1) indicates, and as illustrated in Fig. 7.1, the
universe emerges from a single point at t = , even though slices of constant
cosmic time are infinitely extended thereafterand comoving geodesics naturally
disperse throughout the course of cosmic time.
1 The dialogue is set between Saints Peter and Paul, with the latter presenting Weyls apostatical
and heretical views against the Relativity Church. The following statement, which seems to be
loosely quoted from the postcard sent by Einstein, was made by Peter.
106
D. Janzen
Fig. 7.1 Slices of constant time in the Lematre-Robertson coordination of de Sitter space (black
lines), along with comoving world lines (red lines), drawn on a two-dimensional slice of de Sitter
space in three-dimensional Minkowski space
Thus, we have a sense of the concept of cosmic expansion that was common
amongst the main thinkers in cosmology in the 1920s, who were considering the
possibility of expansion driven by the cosmical repulsion in de Sitter space. Indeed,
Hubble was aware of this concept, as he wrote of the de Sitter effect when he
published his confirmation of cosmic expansion in 1929 [10]; and de Sitter himself,
in 1930, wrote of as a measure of the inherent expanding force of the universe
[11]. Thus, along with the evidence that our Universe actually does expand, one had
in-hand the description of a well-defined force to always drive that expansion.
It was therefore a huge blow to Eddington, e.g., when in 1932 Einstein and de Sitter
[12] finally rejected that interpretation of cosmic expansion, in favour of a model
that could afford no prior explanation for why the Universe should expand. As he
put it [13],
the theory recently suggested by Einstein and de Sitter, that in the beginning all the matter
created was projected with a radial motion so as to disperse even faster than the present rate
of dispersal of galaxies,2 leaves me cold. One cannot deny the possibility, but it is difficult
to see what mental satisfaction such a theory is supposed to afford.
To see why the big bang FLRW models with matter provide no explanation of
expansion, for the reason stated by Eddington, we need only look at Friedmans
equation,

a
p+
,
=
a
3
2
3
2
(7.2)
They do not state this in words, but it is the meaning of their mathematical formulae. [Eddingtons
footnote].
107
which describes the dependence of the scale-factor, a, on and the density, ,

and pressure, p, of matter. Since p + /3 goes like 1/a 4 for radiation or 1/a 3 for
non-relativistic matter, the decelerative force due to finite matter-densities blows up
exponentially as a 0, while the accelerative force due to vanishes; so the
inherent expanding force of the universe only contributes to the expansion of space
later on, when the relative contributions of matter and radiation have sufficiently
weakened. Therefore, aside from Weyls vacuous de Sitter cosmology, with its big
bang singularity at t = , the big bang FLRW models can never explain the
cosmic expansion they describe, which must be caused by the big bang singularity
itselfi.e., where the theory blows up.
But since the cosmic microwave background radiation (CMBR) indicates that
the Universe did begin in a hot dense state at a finite time in the past, the model
Eddington had favoured instead (in which an unstable Einstein universe that existed
since eternity would inevitably begin expanding purely due to [14]) also cant be
accepted.
The principal source of standard cosmologys great explanatory deficit is the
fact that although the non-vacuous big bang FLRW models do describe expanding
universesand in particular the flat CDM model describes the observed expansion of our Universe very well [1526]they afford no reason at all for why those
universes should expand, since that could only be due to the initial singularity; i.e.,
as we follow the models back in time, looking for a possible cause of expansion, we
eventually reach a point where the theory becomes undefined, and call that the cause
of it all. In contrast, Ive discussed two FLRW models, neither of which is empirically
supported, which would otherwise better explain the expansion they describe, as the
result of a force that is well-defined in theory.
The basic cause and nature of cosmic expansion, along with its recently-observed
acceleration, are significant problems of the standard model; so, considering the
evidence that the acceleration is best described by pure [1526], there is strong
motivation to search for an alternative big bang model that would respect the pioneering concept of expansion, as a direct consequence of the de Sitter effect in the
modified Einstein field equations. It is therefore worth investigating the axiomatic
basis of the Robertson-Walker (RW) line-element. As I will eventually argue that
the problem lies in the basic assumptions pertaining to the description of cosmic
time, Ill begin by discussing some issues related to the problem of accounting for a
cosmic present.
The Cosmic Present

The problem of recognising a cosmic present is that, according to relativity theory, it
should not be possible to assign one time-coordinate to the four-dimensional continuum of events that could be used to describe objective simultaneity, since two events
that are described as simultaneous in one frame of reference will not be described as
such by an observer in relative motion. However, as noted by Bondi [27],
108
D. Janzen
The Newtonian concept of the uniform omnipresent even-flowing time was shown by special
relativity to be devoid of physical meaning, but in 1923 H. Weyl suggested that the observed
motions of the nebulae showed a regularity which could be interpreted as implying a certain
geometrical property of the substratum . This in turn implies that it is possible to introduce
an omnipresent cosmic time which has the property of measuring proper time for every observer
moving with the substratum. In other words, whereas special relativity shows that a set of arbitrarily moving observers could not find a common time, the substratum observers move in such
a specialized way that such a public or cosmic time exists.
Although the existence of such a time concept seems in some ways to be opposed to the
generality, which forms the very basis of the general theory of relativity, the development of
relativistic cosmology is impossible without such an assumption.
In fact, as Einstein himself noted in 1917 [28],

The most important fact that we draw from experience as to the distribution of matter is that
the relative velocities of the stars are very small as compared with the velocity of light. So
I think that for the present we may base our reasoning upon the following approximative
assumption. There is a system of reference relatively to which matter may be looked upon
as being permanently at rest.
Thus, the assumption of a cosmic rest-frameand a corresponding cosmic time

was justified in the derivation of Einsteins cylindrical model.
While Einstein originally proposed this as an approximative assumption that the
empirical evidence seemed to support, the fact that he did restore absolute time when
it came to the problem of describing the Universe on the largest scale was not lost on
his peers. De Sitter was immediately critical of the absolute time variable in Einsteins
model, noting that Such a fundamental difference between the time and the spacecoordinates seems to be somewhat contradictory to the complete symmetry of the
field-equations and the equations of motion [29]. And a few years later, Eddington
wrote that an objection to Einsteins theory may be urged, since [30] absolute space
and time are restored for phenomena on a cosmical scaleJust as each limited
observer has his own particular separation of space and time, so a being coexistive
with the world might well have a special separation of space and time natural to him.
It is the time for this being that is here dignified by the title absolute. Therefore, he
concluded, Some may be inclined to challenge the right of the Einstein theoryto
be called a relativity theory. Perhaps it has not all the characteristics which have at
one time or another been associated with that name
Indeed, although the assumption of an absolute time in relativistic cosmology is
definitely not in the spirit of relativity, the theory isnt fundamentally incompatible
with such a definition. Furthermore, it is significant that despite such early criticisms,
Einstein never wavered in assuming an absolute time when he came to consider the
cosmological problem [12, 31, 32], i.e. as he always favoured the Friedman solutions
(with = 0), which begin by postulating the same.
So, we have two opposing descriptions of relativistic timeboth of which are
principally due to Einstein himself!and what Ill now argue is that developments
both in cosmology and in our understanding of relativity theory which have taken
place in the past century demand the latterthat there is one absolute cosmic time
relative to which every observers proper time will measure, as space-time will be
109
perceived differently due to their absolute motion through the cosmic present that
must be uniquely and objectively definedrather than the former implication of
Einsteins 1905 theory of relativity [33].
In the case of special relativity, a description in which space-time emerges as a
clearly defined absolute cosmic present endures, can be realised by considering fourdimensional Minkowski space as a background structure, and a three-dimensional
universe that actually flows equably though itwith the past space-time continuum
emerging as a purely ideal set of previous occurrences in the universe. Then, if we
begin in the cosmic rest-frame, in which fundamental observers world lines will
be traced out orthogonal to the cosmic hyperplane, photons can be described as
particles that move through that surface at the same rate as cosmic time, thus tracing
out invariant null-lines in space-time. In this way, the evolution of separate bodies, all
existing in one three-dimensional space, forms a graduating four-dimensional map.
The causal and inertial structures of special relativity are thus reconciled by
describing the world lines of all observers in uniform motion through the cosmic
present as their proper time axes, and rotating their proper spatial axes accordingly,
so that light will be described as moving at the same rate in either direction of proper
space. And then, so that the speed of photons along invariant null-lines will actually
be the same magnitude in all inertial frames, both the proper space and time axes in
these local frames must also be scaled hyperbolically relative to each other.
This description of the emergence of space-time in a special relativistic universe
can be illustrated in the following way. Consider a barograph, consisting of a pen,
attached to a barometer, and a sheet of paper that scrolls under the pen by clockwork.
The apparatus may be oriented so that the paper scrolls downwards, with changes
in barometric pressure causing the pen to move purely horizontally. We restrict the
speed of the pens horizontal motion only so that it must always be less than the rate at
which the paper scrolls underneath it. The trace of the barometric pressure therefore
represents the world line of an arbitrarily moving observer in special relativistic
space-time, with instantaneous velocity described in this frame by the ratio of its
speed through the horizontal cosmic present and the graph papers vertical speed, with
speed measured in either case relative to the ticking of the clockwork mechanism,
which therefore cancels in the ratio.
Now, in order to illustrate the relativity of simultaneity, we detach the pen (call it
A ) from the barometer so that it remains at rest absolutely, and add another pen, B,
to the apparatus, at the exact same height, which moves horizontally at a constant
rate thats less than the constant rate that the paper scrolls along; therefore, with
absolute velocity less than the absolute speed limit. Furthermore, we make A and B
observers, by enabling them to send and receive signals that transmit horizontally
at the same rate (in clockwork time) as absolute time rolls on (in clockwork time),
thus tracing out lines on the graph paper with unit speed.
As this system evolves, the two timelike observers can send these photons
back and forth while a special relativistic space-time diagram is traced out. If wed
rather plot the map of events in coordinates that give the relevant description from
Bs perspective, we use the Lorentz transformation equations corresponding to the
description of the map as Minkowski space-time: a spacelike line is drawn, tilted
110
D. Janzen
from the horizontal towards Bs world line by the appropriate angle, and the events
along that surface are described as synchronous in that frame, even though they take
place sequentially in real time. In particular, at the evolving present, Bs proper
spatial axis extends, in one direction, onto the empty sheet of graph paper in which
events have not yet occurred, and, in the other direction, into the past space-time
continuum of events that have already been traced onto the paperwhile the real
present hyperplane, where truly simultaneous events are occurring, is tilted with
respect to that axis of relative synchronicity.
The main difference between this interpretation of special relativity and Einsteins
original one, is that simultaneity and synchronicity have objectively different
meanings for us, which coincide only in the absolute rest framewhereas Einstein
established an operational concept of simultaneity, so that it would be synonymous
with synchronicity, in section 1, part 1 of his first relativity paper [33]. Einsteins
definition of simultaneity is a basic assumption thats really no less arbitrary than
Newtons definitions of absolute space, time, and motion; and, as Ill argue, the
evidence from cosmology now stands against Einsteins wrong assumption, as it is
really more in line with Newtons.
The distinction between simultaneity and synchronicity in this different interpretation of relativity, can be understood more clearly through our barograph example,
by adding two more observers, C and C , which remain at rest relative to B, with
C positioned along the same hyperplane as A and B, and C positioned precisely
at the intersection of C s world line (so that the world lines of C and C exactly
coincide, as they are traced out on the space-time graph) and Bs proper spatial
axis (therefore, on a different hyperplane than A , B, and C ); thus, C shall not be
causally connected to A , B, and C , since by definition information can only transmit
along the cosmic hyperplane; see Fig. 7.2.
The significant point that is clearly illustrated through the addition of C and C ,
is that although in the proper coordinate system of B (or C or C ), C appears to
exist synchronously and at rest relative to B, C which in contrast appears to exist
in Bs (spacelike separated) past or future (depending on the direction of absolute
motion; in Fig. 7.2, C appears to exist in Bs relative past)is really the causally
connected neighbour that remains relatively at rest, with which it should be able to
synchronise its clock in the usual way; i.e., the synchronisation of Bs and C s clocks
Fig. 7.2 Snapshots, in two proper reference frames, of an emergent space-time. Although the proper
times of C and B appear to coincide, C is disconnected from the causally coherent set, {A , B , C }
111
will be wrong because simultaneous noumena will not be perceived as synchronous

phenomena in any but the cosmic rest-frame.
According to this description, we should have to relinquish the concept that there
can be no priviliged observers, as well as Einsteins light-postulate in its original form.
With regard to the latter, consider that photons will still be described as travelling
at a constant speed in all directions of all reference frames, due to the invariance
of null-lines. But this wont actually be true, since an observer moving through the
universe will keep pace better with a photon in their direction of motion, and will
remain closer to that photon at all later times, on the cosmic hyperplane. Therefore,
although light actually wont recede as quickly through the universe in the direction
of absolute motion, it can always be described as such in the proper coordinate frame
because it travels along invariant null-lines.
And with regard to the former concept, it is useful to note Galileos argument
that, to a person riding in the cabin of a moving ship, everything inside the cabin
should occur just as if the ship were at rest. It was crucial for Galileo to make this
point by isolating the inertial system from its relatively moving surroundingsas
the point would have been less clear, e.g., if he had argued that when riding in the
back of a wagon one can toss a ball straight in the air and have it fall back to the
same point within the wagon. However, if one should argue that there really cant
be privileged observers in the Universe, due to the relativity of inertia, one must
go beyond this local-inertial effectviz. the relativity of inertiaand consider the
frame with respect to its cosmic surroundingsin which case the argument cant be
justified.
For consider a neutrino, created in a star shortly after the Big Bang: in the neutrinos proper frame, only minutes may have elapsed since it left the star, throughout
which time the galaxies would have formed, etc., all moving past it in roughly the
same direction, at nearly the speed of light. Clearly the most reasonable interpretation, however, is that the neutrino has really been travelling through the Universe
for the past 13.8 billion yearsand this description may be given, with the cosmic
present uniquely and objectively defined, in all frames including the neutrinos.
Furthermore, if we would assume that there are no privileged observers, it should
be noted that the consequence of describing simultaneity and synchronicity as one
and the same thing in all frames is a block universe [34]a temporally singular
absolute world [35] in which the distinction between past, present, and future has
only the significance of a stubborn illusion [36]; i.e., The objective world simply is,
it does not happen. Only to the gaze of my consciousness, crawling upward along the
life line of my body, does a section of this world come to life as a fleeting image in
space which continuously changes in time [37]; There is no dynamics within spacetime itself: nothing ever moves therein; nothing happens; nothing changes. one
does not think of particles moving through space-time, or as following along
their world lines. Rather, particles are just in space-time, once and for all, and the
world line represents, all at once, the complete life history of the particle [38].
And so Ive argued against the simultaneity of synchronicity,a reasonably
intuitive concept held in common between the theories of both Newton and Einstein.
But is there any sensible justification for the concept that the space in which events
112
D. Janzen
really take place simultaneously must be orthogonal to the proper time-axis of an

inertial observer? When our theories are interpreted in this way, is that because one
can, e.g., sit down on the floor with legs out in front, raise their right arm out to the
side and their left arm up in the air, and then stick out their tongue in the direction
in which time is flowing, for them as much as it is for their entire surroundings? Of
course not. This is no more justified for someone who thus defines a right-handed
coordinate system while sitting on solid ground, than it is for a person in the cabin
of a shipwhether that is floating on water or flying through space. Therefore, intuition justifies only existence in space that endures with the ticking of everyones
watchand relativity theory demands that this cannot be both coherently defined
and synchronous with every inertial observer!
Now, although it may be argued that the alternative assumption of cosmic time
is unobservable metaphysics, and therefore unscientific, that simply isnt truefor
cosmology does provide strong empirical evidence of an absolute rest-frame in our
Universe, as follows. As Einstein noted already in 1917 [28], there appears to be
a frame relative to which the bodies of our Universe are at rest, on average. Now,
Einstein had no idea of the scope of the Universe at that time, but already by 1923 Weyl
realised the significance of this point, which has indeed stood the test of time, when he
wrote that [5] Both the papers by de Sitter [3] and Eddington [4] lack this assumption
on the state of rest of starsby the way the only possible one compatible with
the homogeneity of space and time. Without such an assumption nothing can be
known about the redshift, of course. For it is true, even in de Sitter space, that
a cosmic time must be assumed in order to calculate redshifts; e.g., for particles
in the comoving Lematre-Robertson frame illustrated in Fig. 7.1 and described by
Eq. (7.1), the redshift will be different from that in the frame of comoving particles
in the three-sphere which contracts to a finite radius and subsequently expands (as
illustrated by the gridlines of the de Sitter hyperboloid in Fig. 7.1) according to
3
ds = dT + cosh2

2

T
3

d3 2 ,
(7.3)
where d3 describes the three-sphere. The existence of more than one formally
distinct RW cosmological model in one and the same space-time thus illustrates the
importance of defining a cosmic time.
Since 1923, a number of novel observations have strengthened the evidence for
a cosmic present, such as Hubbles confirmation of cosmic expansion, the detailed
measurement of the expansion rate that has lately been afforded through type Ia
supernovae observations, and the discovery of the CMBR, which gives a detailed
signature of the cosmic rest-frame relative to which we are in fact moving, according
to the common interpretation of its dipole anisotropy. Thus, the assumption of a
cosmic present is now very well justified by empirical evidence.
113
Implications for Cosmology

Although many points should be considered in connection to the description of an
absolute cosmic present, such as concepts of time travel, free will, and a causally
coherent local description of gravitational collapse in the Universenotwithstanding
space-time curvature in general,the one consequence that I will note pertains to
cosmology and a better explanation of cosmic expansion.
To start, note that in deriving the general line-element for the background geometry
of FLRW cosmology, Robertson required four basic assumptions [39]: i. a congruence of geodesics, ii. hypersurface orthogonality, iii. homogeneity, and iv. isotropy. i.
and ii. are required to satisfy Weyls postulate of a causal coherence amongst world
lines in the entire Universe, by which every single event in the bundle of fundamental world lines is associated with a well-defined three-dimensional set of others
with which it really occurs simultaneously. However, it seems that ii. is therefore
mostly required to satisfy the concept that synchronous events in a given inertial
frame should have occurred simultaneously, against which Ive argued above.
In special relativity, if we allow the fundamental world lines to set the cosmic
rest-frame, then the cosmic hyperplane should be orthogonalbut that shouldnt be
the case in general. Indeed, as Ive shown in my Ph.D thesis [40], in the cosmological
Schwarzschild-de Sitter (SdS) solution,
ds 2 =
3
r
r3
+ 2M r
dr 2 +
3
3r
+ 2M r 2
dt + r 2 d2 ,
r
(7.4)
for which M 2 > 1/9, r > 0 is timelike, and t is forever spacelike, the r -coordinate
should well describe the cosmic time and factor of expansion in a universe in which,
in the coordinates carried by fundamental observers, the cosmic present would not
be synchronous, and r would evolve in proper time as
r ( ) sinh2/3 [( 3/2) ],
(7.5)
which is incidentally also the flat CDM scale-factor of the standard model that has
been empirically constrained this past decade [1526]; see Appendix: Concerning
Schwarzschild-de Sitter as a Cosmological Solution for a derivation of Eq. (7.5)
beginning from Eq. (7.4), and a discussion of the results connection to cosmology.
This is the rate of expansion that all observers would measure, if distant galaxies
were themselves all roughly at rest with respect to fundamental world lines. But in
contrast to FLRW theory, this universe actually has to expandat all r > 0as
a result of the de Sitter effect; i.e., if such a universe did come to exist at any
infinitesimal time, it would necessarily expandand in exactly the manner that we
observewhich may be the closest to an explanation of that as we can achieve.
It is, of course, important to stress that this intriguing result is utterly meaningless
if simultaneity should rather be defined as synchronicity in a given frame of reference.
In that case, as Lematre noted [41], the solution describes flat spatial slices extending
114
D. Janzen
from r = 0 to , with particles continuously ejected from the origin. It is therefore

only by reconceiving the relativistic concepts of time and simultaneity that SdS can
be legitimated as a coherent cosmological model with a common originand one
with the very factor of expansion that weve measuredwhich really should expand,
according to the view of expansion as being always driven by .
Acknowledgments Thanks to Craig Callender for reviewing an earlier draft and providing thoughtful feedback that greatly improved this essay. Thanks also to the many participants who commented
on and discussed this paper throughout the contest, and FQXi for organising an excellent contest
and providing criteria that helped shape the presentation of this argument.
Appendix: Concerning Schwarzschild-de Sitter as a

Cosmological Solution
During the Essay Contest discussions, the critical remarks on this essay that were
most important for me, and were by far the most probing, were those offered by
George Ellis. Professor Ellis criticism of the final section indicated, first of all, that
the brief mention I made there of a result from my Ph.D thesis was too underdeveloped
to pique much interest in itand in fact that, stated as it was there, briefly and out
of context of the explicit analysis leading from Eq. (7.4) to (7.5), the point was too
easily missed. He wrote that the model is of course spatially inhomogeneous, when
the spatial slices are actually homogeneous, but rather are anisotropic; and when
I pointed out to him that this is so because, in the cosmological form of the SdS
solution r > 0 is timelike and t is forever spacelike, he replied that the coordinate
notation is very misleading.
So, one purpose of this appendix is to provide the intermediate calculation between
Eqs. (7.4) and (7.5), that had to be left out of the original essay due to space
limitationsand, in developing a familiarity with the common notation, through
the little calculation, to ensure that no confusion remains in regard to the use of r
as a timelike variable and t as a spacelike one. For the notation is necessary both
in order to be consistent with every other treatment of the SdS metric to date, and
because, regardless of whether r is timelike or spacelike in Eq. (7.4), it really does
make sense to denote the coordinate with an r because the space-time is isotropic
(i.e. radially symmetric) in that direction.
With these bookkeeping items out of the way (after roughly the first four pages),
the appendix moves on to address Professor Ellis two more substantial criticisms,
i.e. regarding the spatial anisotropy and the fact that the model has no dynamic matter
in it; for, as he noted, the model is interesting geometrically, but it needs supplementation by a dynamic matter and radiation description in order to relate to our
cosmic history. These important points were discussed in the contest forum, but
were difficult to adequately address in that setting, so the problem is given more
proper treatment in the remaining pages of this appendix once the necessary mathematical results are in-hand. Specifically, in the course of developing a physical picture
115
in which the SdS metric provides the description of a universe that would appear
isotropic to fundamental observers who measure the same rate of expansion that we
do (viz. as given by Eq. (7.5)), we will come to a possible, consistent resolution to
the problem of accounting for dynamic matter, which leads to a critical examination
of the consistency and justification of some of the most cherished assumptions of
modern physics, thus further questioning its foundations.
We begin by writing down the equations of motion of radial geodesics in the
SdS geometry, using them to derive a description of the SdS cosmology that would
be appropriate to use from the perspective of fundamental observers who evolve as
they do, beginning from a common origin at r = 0, always essentially because of the
induced field potential. It will be proved incidentally that the observed cosmological
redshifts, in this homogeneous universe which is not orthogonal to the bundle of
fundamental geodesicsand is therefore precluded by the a priori assumptions of
standard FLRW cosmologymust evolve through the course of cosmic time, as
a function of the proper time of fundamental observers, with the precise form of
the flat CDM scale-factori.e., with exactly the form that has been significantly
constrained through observations of type Ia supernovae, baryon acoustic oscillations,
and CMBR anisotropies [1526].
Since the Lagrangian,
r 2M
L=
r
3
3r
dt
d
2
+
r
r 2M
3
3r
dr
d
2
= 1,
(7.6)
for timelike (r, t)-geodesics with proper time in the SdS geometry is independent
of t, the Euler-Lagrange equations indicate that
r 2M
1 L
=
E
2 (dt/d )
r
3
3r
dt
d

(7.7)
is conserved (2dE/d dL/dt = 0). Substituting Eq. (7.7) into (7.6), then, we
find the corresponding equation of motion in r :

dr
d
2
= E2
r 2M
r
3
3r
(7.8)
While the value of E may be arbitrary, we want a value that distinguishes a particular set of geodesics as those describing particles that are fundamentally at resti.e.,
well distinguish a preferred fundamental rest frame by choosing a particular value of
E that meets certain physical requirements. In order to determine which value to use,
we first note that where r is spacelike, Eq. (7.8) describes the specific (i.e., per unit
rest-mass) kinetic energy of a test-particle, as the difference between its (conserved)
specific energy and the gravitational fields effective potential,
116
D. Janzen
Veff (r )
r 2M
r
3
3r
(7.9)
Then, a reasonable definition sets the fundamental frame as the one in which
the movement of particles in r and t is essentially caused by the non-trivial field
potentiali.e., so that dr/d = 0 just where the gravitational potential is identically
trivial (Veff 1), and the line-element, Eq. (7.4), reduces to that of Minkowski space.
2
From Eq. (7.8), this amounts to setting E 2 = 1; therefore,
a value E > 1 corresponds to a particle that would not come to rest at r = 3 6M/, where Veff 1,
but has momentum in r beyond that which would be imparted purely by the field.
As a check that the value E 2 = 1 is consistent with our aims, we can consider
itsphysical meaning another way. First of all, note that where Veff 1, at r =
3 6M/, r is spacelike and t is timelike regardless of the values of M and ;
therefore, it is consistent in any case to say that a particle with E 2 > 1 has nonvanishing
spatial momentum there. Indeed, from Eq. (7.7), we find that t = at
r = 3 6M/ if, and only if, E = 1so the sign of E should in fact be positive for
a particle whose proper time increases with increasing t in the absence of gravity.
Furthermore, note that when = 0, Eq. (7.8) reduces to

dr
d
2
= E2 1 +
2M
.
r
(7.10)
As such, Misner, Thorne and Wheeler describe E as a particles specific energy

at infinity, where the effective potential is trivial [42]. It is relevant to note their
statement (on p. 658), that the conservation of 4-momentum allows and forces one
to take over the [term] E =energy at infinity, valid for orbits that do reach
to infinity, for an orbit that does not reach to infinity. More generally, we should
describe E forarbitrary M and as the energy at vanishing field potential even
when r = 3 6M/ is a negative mathematical abstraction that lies beyond a
singularity at r = 0. In particular, we take E = 1 to be the specific energy of a
test-particle that is at rest with respect to the vanishing of the potential. Its simply
a matter of algebraic consistency.
Thus, we have E = 1 as the specific energy of particles that would come to rest
in the absence of a gravitational field, which are therefore guided purely through
the effective field potential. We therefore use the geodesics with E = 1 to define a
preferred rest frame in the SdS geometries, and say that any particle whose world
line is a geodesic with E = 1 is one that has uniform momentum relative to the
fundamental rest frame.
We can now write the SdS line-element, Eq. (7.4), in the proper frame of a bundle
of these fundamental geodesics, which evolve through t and r all with the same
proper time, , and occupy constant positions in space. Since must be positive
in order to satisfy the requirement, M 2 > 1/9, for r > 0 to be timelikei.e. the
requirement for the SdS line-element to be cosmological rather than a local solution

it is more convenient to work with scale-invariant parameters r r =

3 r,

M M = 3 M, etc., normalising all
dimensional quantities by the cosmic length-scale 3/ (see, e.g., 66 in [4] or

66 in [43] for interesting discussions of this length parameter). This normalisation
ultimately
amounts to striking out the factor
/3 from the r 3 -term in the line-element
t t =
since

3 t,

3 r 2
=

3 M

3r
117

3 ,

3r
r 2Mr 3
,
r
or e.g. writing the flat CDM scale-
factor, Eq. (7.5), as

r ( ) sinh2/3 (3/2) ,
(7.11)
and the corresponding Hubble parameter as

H r /r = coth(3/2),
(7.12)
which exponentially approaches H = 1 on timescales 2/3.

The evolution of each geodesic through scale-invariant t and r is then given,
through Eqs. (7.7) and (7.8) with E = 1, as
t
r
t
=
,
r 2M r 3

( r )2
Eq. (7.14) can be solved using
2
=
2M + r 3
.
r
(7.13)
(7.14)

2
1/2
u +a
du = ln u + u 2 + a after substi-
tuting u 2 = 2M + r 3 . Taking the positive root (so increases with r ), we have,

r ( )
=
r (0)
r ( )
r
2
3/2
3
dr = ln
2M + r + r
,

2M + r 3
3
r (0)
(7.15)
where the lower limit on has been arbitrarily set to 0. Thus, in this frame we can
express r as a function of each observers proper time and an orthogonal (i.e.
synchronous, with constant = 0) spatial coordinate, r (0), which may be arbitrarily
rescaled without altering the description in any significant way.
Then, as long as M is nonzero, a convenient set of coordinates from which to
proceed results from rescaling the spatial coordinate as3
Note that this transformation is not valid when M = 0, which we are anyhow not interested in. An
equivalent transformation in that case is found by setting r (0) e , whence
= e + , and
2 r (, )

2
2
2
2
Eqs. (7.21), (7.24), and (7.27), yield the line-element, ds = d + r d + d + sin2 d 2 .
118
D. Janzen

r (0) (2M)1/3 sinh2/3

3
; M = 0,
2
(7.16)
from which we find, after some rearranging of Eq. (7.15),4

r (, ) = (2M)1/3 sinh2/3

3
[ + ] ,
2
(7.17)
which immediately shows the usefulness of rescaling the r (0) as in Eq. (7.16), since
it allows Eq. (7.15) to be solved explicitly for r (, ). As such, we immediately have
the useful result (cf. Eq. (7.8)),
r
= r =
r
2M + r 3
.
r
(7.18)
The transformation, t (, ), may then be calculated from

t
r
dt
=
=
dr
r
r 2M r 3
r
.
2M + r 3
(7.19)
Then, to solve for t (, ), we can gauge the lower limits of the integrals over t and
r , at = 0, by requiring that their difference, defined by
r
(, )
t (, ) =
r
r 2M r 3
r
dr F( ),
2M + r 3
(7.20)
sets
0 = g = gtt t t + grr r r.
(7.21)
(Thus, will be orthogonal to .) This calculation is straightforward5 :

r
2M + r 3
r
+
F(
)
+
r 2M r 3
r 2M r 3
r
= F( ) 1,
0=
(7.22)
(7.23)

Note that the two identities, e x = sinh(x) + cosh(x) and arsinh(x) = ln x + x 2 + 1 , are
useful here. Eq. (7.17), along with our eventual line-element, Eq. (7.28), was originally found by
Lematre [41], although his solution to Eq. (7.14) (with dimensionality restored),

r = (6M/)1/3 sinh2/3 3 (t t0 )/2 ,
is too large in its argument by a factor of 3.

5 Note that we dont actually have to solve the integral in Eq. (7.20), since only partial derivatives
of t are needed here and below.
119
so that F( ) = .
Now, it is a simple matter to work out the remaining metric components as follows:
our choice of proper reference frame immediately requires
g = gtt ( t)2 + grr ( r )2 = 1,
(7.24)
according to the Lagrangian, Eq. (7.6); and by direct calculation, we find

(7.25)
g = gtt ( t)2 + grr ( r )2

2
r 2M r 3
2M + r 3
2M + r 3
r
=
(7.26)
+
3
3
r
r 2M r
r 2M r
r
= ( r )2 .
(7.27)
But this result is independent of any arbitrary rescaling of ; for if we replaced

= f ( ) in Eq. (7.16), we would then find the metric to transform as g =
g (d /d )2 = ( r )2 , the other components remaining the same.
Therefore, the SdS metric in the proper frame of an observer who is cosmically
at rest, in which the spatial coordinates are required, according to an appropriate
definition of F( ), to be orthogonal to ,6 can generally be written,

2
ds 2 = d 2 + r d 2 + r 2 d 2 + sin2 d 2 .
(7.28)

2
This proves Lematres result from 1949 [41],that slices d = 0 r d 2 =

(dr/d )2 d 2 = dr 2 are Euclidean, with line-element,

d 2 = dr 2 + r 2 d 2 + sin2 d 2 .
(7.29)
However, in the course of our derivation we have also found that Lematres physical
interpretationthat the geometry is Euclidean on the expanding set of particles
which are ejected from the point singularity at the originis wrong.
It is wrong to interpret this solution as describing the evolution of synchronous
space which always extends from r = 0 to r = + along lines of constant
, being truncated at the r = 0 singularity at = from which particles are
continuously ejected as increases. But this is exactly the interpretation one is apt
to make, who is accustomed to thinking of synchronous spacelike hypersurfaces as
space that exists out there, regardless of the space-time geometry or the particular
coordinate system used describe it.
As we noted from the outset, the radial geodesics that we have now described
by the lines = const., along which particles all measure their own proper time
to increase with , describe the world lines of particles that are all fundamentally
6
Note that the radially symmetric part of Eq. (7.4) is already orthogonal to .
120
D. Janzen
at resti.e., at rest with respect to the vanishing of the effective field potential.
Therefore, these particles should not all emerge from the origin at different times,
and then somehow evolve together as a coherent set; but by Weyls principle they
should all emerge from a common origin, and evolve through the field that varies
isotropically in r , together for all time. In that case, space will be homogeneous, since
the constant cosmic time (dr = 0) slices of the metric can be written independent
of spatial coordinates; so every fundamental observer can arbitrarily set its spatial
position as = 0 and therefore its origin in time as = 0.
The spaces of constant cosmic time should therefore be those slices for which
r ( + ) = const.i.e., we set = + as the proper measure of cosmic time
in the fundamental rest frame of the universe defined by this coherent bundle of
geodesics, so that Eq. (7.17) becomes
r ( ) = (2M)1/3 sinh2/3 (3 /2) .
(7.30)
The spacelike slices of constant are at 45 angles in the (, )-plane, and are
therefore definitely not synchronous with respect to the fundamental geodesics. However, given this definition of cosmic time, the redshift of light that was emitted at e
and is observed now, at 0 , should be
1+z =
r (0 )
,
r (e )
(7.31)
where r ( ) has exactly the form of the flat CDM scale-factor (cf. Eq. (7.11)),
which is exactly the form of expansion in our Universe that has been increasingly
constrained over the last fifteen years [1526].
Now, in order to properly theoretically interpret this result for the observed redshift
in our SdS cosmology, it should be considered in relation to FLRW cosmology
and particularly the theorys basic assumptions. As noted in Sect. Implications for
Cosmology, the kinematical assumptions used to constrain the form of the lineelement are: i. a congruence of geodesics, ii. hypersurface orthogonality, iii. homogeneity, and iv. isotropy. Assumptions i. and ii. have a lot to do with how one defines
simultaneity, which I have discussed both in the context of special relativity in
Sect. The Cosmic Present, and now in the context of the SdS cosmology, in which
simultaneous events that occur in the course of cosmic time are not synchronous
even in the fundamental rest frame. As the discussion should indicate, the definition
of simultaneity is somewhat arbitraryand it is an assumption in any caseand
should be made with the physics in mind. Einstein obviously had the physics in mind
when he proposed using an operational definition of simultaneity [33]; but it has
since been realised that even special relativity, given this definition, comes to mean
that time cant pass, etc., as noted in Sect. The Cosmic Present.
Special relativity should therefore be taken as an advance on Newtons bucket
argument, indicating that not only should acceleration be absolute, as Newton showed
(see, e.g., [44] for a recent discussion of Newtons argument), but velocity should
121
be as well, since time obviously passeswhich it cant do, according to special

relativity, if motion isnt always absolute. Usually, however, the opposite is done,
and people who have been unwilling or unable to update the subjective and arbitrary
definitions of simultaneity, etc., from those laid down by Einstein in 1905, have simply concluded that Physical Reality has to be a four-dimensional Block in which time
doesnt pass, and the apparent passage of time is a stubborn illusion; see, e.g., Sect. 5
in [45] for a popular account of this, in addition to Refs. [3438]. The discussion in
Sect. The Cosmic Present shows how to move forward with a realistic, physical,
and most importantly a relativistic description of objective temporal passage, which
can be done only when simultaneity is not equated with synchronicity a priori;
and another useful thought-experiment along those lines, which shows how perfectly
acceptable it is to assume objective temporal passage in spite of relativistic effects,
is presented in my more recent FQXi essay [46].
In contrast to the hardcore relativists who would give up temporal passage in favour
of an operational definition of simultaneity, Einstein was the first relativist to renege
on truly relative simultaneity when he assumed an absolute time in constructing his
cosmological model [28]; and despite immediately being chastised by de Sitter over
this [29], he never did balk in making the same assumption whenever he considered
the cosmological problem [12, 31, 32]as did just about every other cosmologist
who followed, with very few notable exceptions (e.g., de Sitter [3, 29] was one, as
there was no absolute time implicit in his model).
But whenever the assumption of absolute time has been made in cosmology, it
has been made together with special relativitys baggage, as the slices of true simultaneity have been assumed to be synchronous in the fundamental rest frame. Now we
see that, not only is the operational definition wrong in the case of special relativity
(since it comes to require that time does not pass, which is realistically unacceptable),
but here weve considered a general relativistic example in which equating simultaneity and synchronicity makes even less sense in terms of a reasonable physical
interpretation of the mathematical description, since the interpretation is causally
incoherenti.e. Lematres interpretation, that the line-element Eq. (7.28) should
describe an expanding set of particles which are ejected from the point singularity at the origin represents abominable physical insight. The main argument of this
essay was therefore, that while assumption i. of FLRW cosmology is justified from the
point of view that relative temporal passage should be coherent, assumption ii. is not,
and this unjustified special relativistic baggage should be shed by cosmologistsand
really by all relativists, as it leads to further wrong interpretations of the physics.
Assumption iii. hardly requires discussion. It is a mathematical statement of the
cosmological principlethat no observer holds a special place, but the Universe
should look the same from every locationand is therefore as fundamental an
assumption as the principle of relativity. Furthermore, our SdS universe is homogeneous, so there is no problem.
The final assumption, however, is a concern. The isotopy of our Universe is an
empirical factit looks the same to us in every direction, and the evidence is that it
must have done since its beginning. In contrast, the spatial slices of the cosmological
SdS solution are not isotropic: they are a 2-sphere with extrinsic radius of curvature
122
D. Janzen
r , multiplied by another dimension that scales differently with r . Furthermore, by

Eq. (7.19) we know that all these fundamental world lines move uniformly through
this third spatial dimension, t, as r increases.
The SdS cosmology therefore describes a universe that should be conceived as
follows: First of all suppressing one spatial dimension, the universe can be thought
of as a 2-sphere that expands from a point, with all fundamental observers forever
motionless along the surface; then, the third spatial dimension should be thought of
as a line at each point on the sphere, through which fundamental observers travel
uniformly in the course of cosmic time.
Since it is a general relativistic solution, the distinction between curvature along
that third dimension of space and motion through it is not well-defined. However, a
possibility presents itself through an analogy with the local form of the SdS solution.
As with all physically meaningful solutions of the Einstein field equations, this one
begins with a physical concept from which a general line-element is written; then the
line-element gets sent through the field equations and certain restrictions on functions
of the field-variables emerge, allowing us to constrain the general form to something
more specific that satisfies the requisite second-order differential equations. This
is, e.g., also how FLRW cosmology is donei.e., first the RW line-element satisfying four basic physical/geometrical requirements is written down, and then it is
sent through the field equations to determine equations that restrict the form of the
scale-factors evolution, under a further assumption that finite matter densities in the
universe should influence the expansion rate. The local SdS solution, too, is derived
as the vacuum field that forever has spacelike radial symmetry about some central
gravitating body, and the field equations are solved to restrict the form of the metric
coefficients in the assumed coordinate system. But then, as Eddington noted [4],
We reach the same result if we attempt to define symmetry by the propagation of light, so
that the cone ds = 0 is taken as the standard of symmetry. It is clear that if the locus ds = 0
has complete symmetry about an axis (taken as the axis of t) ds 2 must be expressable by
[the radially isotropic line-element with general functions for the metric coefficients].
Therefore, the local SdS metric corresponds to the situation in which light propagates
isotropically, and its path in space-time is described by the null lines of a Lorentzian
metric. Prior to algebraic abstraction (i.e. the assumption of a Lorentzian metric and
a particular coordinate system), the geometrical picture is already set; and it is upon
that basic geometrical set-up that the algebraic properties of the general relativistic
field are imposed.
This construction of the local SdS solution through physical considerations of
light-propagation can be used analogously in constructing a geometrical picture
upon which the cosmological SdS solution can be based; however, a some more
remarks are necessary before coming to that. First of all, as our discussion of the
local SdS and FLRW solutions indicates, in general much of the physics enters into
the mathematical description already in defining the basic geometrical picture and the
corresponding line-element, which broadly sets-up the physical situation of interest.
Only then is the basic physical picture further constrained by requiring that it satisfy
the specific properties imposed by Einsteins field law. In fact, when it comes to
123
the cosmological problem, and we begin as always by assuming what will be our
actual space, and how it will roughly evolve as actual cosmic time passesi.e. by
first assuming prior kinematical definitions of absolute space and time, and then
constructing an appropriate cosmological line-element, which is finally constrained
through the dynamical field lawthere is really a lot of room to make it as we like it.
But now we have a particular line-element in mind (vizthe SdS cosmological
solution), and we can use it in guiding our basic kinematical definitions. In particular, we have the description of a universe that is a two-dimensional sphere that
expands as a well-defined function of the proper time of fundamental observers
who all remain absolutely at rest, multiplied by another dimension through which
those same observers are moving at a rate that varies through the course of cosmic
time. According to the equivalence principle, it may be that the gravitational field is
non-trivial along that particular direction of space, and therefore guides these fundamental observers alongor it could be that this direction is uniform as well, and that
the fundamental observers are moving along it, and therefore describe it relatively
differently.
What is interesting about this latter possibility, is that there would be a fundamental
background metric describing the evolution of this uniform space, and that the metric
used by fundamental observers to describe the evolution of space-time in their proper
frames shouldnt necessarily have to be the same fundamental metric transformed to
an appropriate coordinate system. The metric itself might be defined differently from
the background metric, for other physical reasons. In this case, an affine connection
defining those world lines as fundamental geodesics may not be compatible with the
more basic metric, and could be taken as the covariant derivative of a different one.
The picture starts to resemble teleparallelism much more closely than it does general
relativity; but since the two theories are equivalent, and we have recognised that in
any case the kinematical definitions bust be made firsti.e. since we must set-up
the kinematical definitions in the first place, according to the physical situation we
want to describe, before ensuring that the resulting line-element satisfies the field
equationswell press on in this vein.
Let us suppose a situation where there is actually no gravitational mass at all, but
fundamental inertial observersthe constituent dynamical matter of our system
are really moving uniformly through a universe that fundamentally is isotropic and
homogeneous, and expands through the course of cosmic time. The fundamental
metric for this universe should satisfy even the RW line-elements orthogonality
assumption, although the slices of constant cosmic time would not be synchronous
in the rest frame of the fundamental observers. Since space, in the two-dimensional
slice of the SdS cosmology through which fundamental observers are not moving,
really is spherical, the obvious choice is an expanding 3-sphere, with line-element
ds 2 = dT 2 + R(T )2 d3 2 ,
(7.32)
124
D. Janzen
where the radius R(T ) varies, according to the vacuum field equations, as
R(T ) = C1 e

3T
+ C2 e

3T
; C1 C2 =
3
.
4
(7.33)
In particular, because a teleparallel theory would require a parallelisable manifold,

we note that this is true if

1 3
,
(7.34)
C1 = C2 =
2
and the de Sitter metric is recovered
(cf. Eq. (7.3)). While there may be concern
because in this case R(0) = 3/ > 0 is a minimum of contracting and expanding
space, I will argue below that this may actually be an advantage.
This foliation of de Sitter space is particularly promising for a couple of reasons: i.
the bundle of fundamental geodesics in Eq. (7.3) are world lines of massless particles,
i.e. the ones at r = 0 for all t in Eq. (7.4) with M = 0; and ii. unlike e.g. the 2-sphere
(or spheres of just about every dimension), the 3-sphere is parallelisable, so it is
possible to define an objective direction of motion for the dynamic matter.
Now we are finally prepared to make use, by analogy, of Eddingtons remark on
the derivation of the local SdS metric in terms of light propagation along null lines. In
contrast, we are now beginning from the description of a universe in which particles
that are not moving through space are massless, but we want to write down a different
Lorentzian metric to describe the situation from the perspective of particles that are
all moving through it at a certain rate, who define null lines as the paths of relatively
moving masselss particlesso, we will write down a new metric to use from the
perspective of particles that all move along null lines pointing in one direction of de
Sitter space, describing the relatively moving paths of massless particles that actually
remain motionless in the 3-sphere, as null lines instead. This new line-element can
be written,
ds 2 = A(r, t)dr 2 + B(r, t)dt 2 + r 2 d2 ,
(7.35)
where r points in the timelike direction of the universes increasing radius, and
t describes the dimension of space through which the fundamental particles are
moving. Solving Einsteins field equations proves the Jebsen-Birkhoff theoremthat
A and B are independent of the spacelike variable tand leads to the cosmological
SdS solution, Eq. (7.4), as the abstract description of this physical picture.
Thus, we have come full circle to a statement of the line-element that we started
with. Our analysis began with a proof that in this homogeneous universe, redshifts
should evolve with exactly the form that they do in a flat CDM universe; and in
the last few pages we have aimed at describing a physical situation in which this
line-element would apply in the proper reference frame of dynamical matter, and
the observed large scale structure would be isotropic. And indeed, in this universe,
in which the spatial anisotropy in the line-element is an artifact of the motion of
125
fundamental observers through homogeneous and isotropic expanding space and

their definition of space-times null lines, this would be sofor as long as these
fundamental observers are uniformly distributed in space that really is isotropic and
homogeneous, all snapshots of constant cosmic time (and therefore the development,
looking back in time with increasing radius all the way to the cosmological horizon)
should appear isotropic, since these uniformly distributed observers would always
be at rest relative to one another.
Having succeeded in showing how the SdS metric can be used to describe a homogeneous universe in which the distribution of dynamical matter appears isotropic from
each point, and which would be measured to expand at exactly the rate described by
the flat CDM scale-factori.e. with exactly the form that has been observationally
constrained [1526], but which has created a number of problems because, from a
basic theoretical standpoint, many aspects of the model are not what we expect
we should conclude with a brief discussion of potential implications emerging from
the hypothesis that the SdS cosmology accounts for the fundamental background
structure in our Universe.
The most obvious point to note is that the hypothesis would challenge what Misner,
Thorne, and Wheeler called Einsteins explanation for gravitationthat Space
acts on matter, telling it how to move. In turn, matter reacts back on space, telling it
how to curve [42]. For according to the SdS cosmology, the structure of the absolute
background should remain unaffected by its matter-content, and only the geometry
of space-timethe four-dimensional set of events that develops as things happen in
the course of cosmic timewill depend on the locally-inertial frame of reference in
which it is described.
When I mentioned to Professor Ellis in the essay contest discussions, that I think
the results I have now presented in this appendix may provide some cause to seriously
reconsider the assumption that the expansion rate of the Universe should be influenced
by its matter-contenta fundamental assumption of standard cosmology, based on
Einsteins explanation for gravitation, which Professor Loeb also challenged in
his submission to the contesthis response was, Well its not only of cosmology its
gravitational theory. It describes solar system dynamics, structure formation, black
holes and their interactions, and gravitational waves. The assumption is that the
gravitational dynamics that holds on small scales also holds on large scales. Its
worked so far. And indeed, it has worked so farbut that is not a good reason
to deny consideration of alternate hypotheses. In fact, as noted by Einstein in the
quotation that began this essay, Concepts that have proven useful in the order of
things, easily attain such an authority over us that we forget their Earthly origins
and accept them as unalterable factsThe path of scientific advance is often made
impassable for a long time through such errors. From Einsteins point of view, such
a position as Its worked so far is precisely what becomes the greatest barrier to
scientific advance.
After making this point, Einstein went on to argue that it is really when we
challenge ourselves to rework the basic concepts we have of Nature that fundamental
advances are made, adding [1],
126
D. Janzen
This type of analysis appears to the scholars, whose gaze is directed more at the particulars,
most superfluous, splayed, and at times even ridiculous. The situation changes, however,
when one of the habitually used concepts should be replaced by a sharper one, because the
development of the science in question demanded it. Then, those who are faced with the
fact that their own concepts do not proceed cleanly raise energetic protest and complain of
revolutionary threats to their most sacred possessions.
In this same spirit, we should note that not only our lack of a technical reason
for why the Universe should ever have come to expandi.e. the great explanatory
deficit of modern cosmology described in Sect. The Cosmic Presentbut also both
the cosmological constant problem [47] and the horizon problem [48] are significant
problems under the assumption that cosmic expansion is determined by the Universes matter-content. The former of these two is the problem that the vacuum does
not appear to gravitate: an optimistic estimate, that would account for only the electrons contribution to the vacuum energy, still puts the theoretical value 1030 larger
than the dark energy component that cosmologists have measured experimentally
[49]. The latter problem is that we should not expect the observable part of a general
relativistic big bang universe to be isotropic, since almost all of what can now be seen
would not have been in causal contact prior to structure formationand indeed, the
light reaching us now from antipodal points of our cosmic event horizon has come
from regions that still remain out of causal contact with each other, since they are
only just becoming visible to us, at the half-way point between them.
While there is no accepted solution to the cosmological constant problem, the
horizon problem is supposed to be resolved by the inflation scenario [48]an epoch
of exponential cosmic expansion, proposed to have taken place almost instantly after
the Big Bang, which would have carried off our cosmic horizon much faster than
the speed of light, leaving it at such a great distance that it is only now coming
back into view. Recently, provisional detection of B-mode polarisation in the CMBR
[50] that is consistent with the theory of gravitational waves from inflation [5153]
has been widely lauded as a proof of the theory. However, in order to reconcile
apparent discrepancies with measurements of the CMBRs anisotropy signature from
the Planck satellite, BICEP2 researchers have suggested that ad hoc tweaks of the
CDM model may be necessary [54].
Details involving the emergence of dynamical matter in the SdS cosmology have
not been worked out; however, there is no reason to suspect ab initio that the gravitational waves whose signature has potentially been preserved in the CMBR, would
not also be produced in the scenario described here. More importantly, though, the
SdS cosmology provides a description that precisely agrees with the observed largescale expansion of our Universe, and does so without the need to invent any ad hoc
hypotheses in order to save the appearances that we have found to be so very different from our prior theoretical expectations. The theory simultaneously solves the
expansion problem outlined in Sect. On Cosmic Expansion (since expansion must
proceed at an absolute rate, regardless of the universes dynamical matter-content)
and subverts the major issues associated with the assumption that cosmic expansion
is determined by the Universes matter-content.
127
In the SdS cosmology, the vacuum can very well gravitate locally without affecting
the cosmic expansion rate, and the universe should appear, from every point, to be
isotropic out to the event horizon. Even the flatness problemviz. the problem
that the curvature parameter has been constrained very precisely around zero, when
the expectation, again under the assumption that the Universes structure should be
determined by its matter-content, is that it should have evolved to be very far from
thatwhich inflation is also meant to resolve, is subverted in this picture. For indeed,
the universe described here, despite being closed, would be described by fundamental
observers to expand exactly according to the flat CDM scale-factor, Eq. (7.30).
Additionally, the SdS cosmology provides a fundamentally different perspective
on the so-called coincidence problemviz. that the matter and dark energy density
parameters of the standard FLRW model are nearly equal, when for all our knowledge they could easily be hundreds of orders of magnitude different. Indeed, if we
write down the Friedman equation for a flat CDM universe using Eq. (7.12) with
dimensionality restored,

H = coth2
3
2
3

=
1
(8 + ),
3
(7.36)
we find that the matter-to-dark energy density ratio, 8/, while infinite at
the big bang, should approach zero exponentially quickly. For example, given the
measured value 1035 s2 , we can write
2
= arcoth + 1 11 Gyr arcoth + 1.
3
(7.37)
From here, we find that the dark energy density was 1 % of the matter density at
( = 100) = 1 billion years after the big bang, and that the matter density will be
1 % of the dark energy density at ( = 0.01) = 33 billion years after the big bang.
At first glance, these results may seem to indicate that it is not so remarkable that
should have a value close to 1 at present. However, from an FLRW perspective,
the value of could really have been anythingand if it were only 104 larger
than it is (which is indeed still far less than our theoretical predictions), would
have dropped below 1 % already at = 0.3 billion years, and the Universe should
now, at 13.8 billion years, be nearly devoid of observable galaxies (so we would
have trouble detecting s presence). On the other hand, if were smaller by 104 ,
it would only become detectable after 100 billion years. Therefore, it is indeed a
remarkable coincidence that has the particular order of magnitude that it has,
which has allowed us to conclusively detect its presence in our Universe.
In contrast to current views, within the FLRW paradigm, on the problem that we
know of no good reason why should have the particularly special, very small value
that it haswhich has often led to controversial discussions involving the anthropic
principle and a multiverse settingthe SdS cosmology again does not so much
solve the problem by explaining why the particular value of should be observed,
but really offers a fundamentally different perspective in which the same problem
128
D. Janzen
simply does not exist. For indeed, while the mathematical form of the observed
scale-factor in the SdS cosmology should equal that of a flat CDM universe, the
energy densities described in Friedmanns equations are only effective parameters
within the former framework, and really of no essential importance. And in fact, as
the analysis in the first part of this Appendix indicates (and again, cf. the sections by
Eddington [4] and Dyson
[43]), fundamentally sets the scale in the SdS universe:
from this perspective, 3/, which has an empirical value on the order of 10 billion
years, is the fundamental timescale. It therefore makes little sense to question what
effect different values of would have on the evolution of an SdS universe, since
sets the scale of everything a priori; thus, the observable universe should rescale
with any change in the value of . But for the same reason, it is interesting to note
that the present order of things has arisen on roughly this characteristic timescale.
From this different point-of-view, then, the more relevant question to ponder is: Why
should the structure of the subatomic world be such that when particles did coalesce
to form atoms and stars, those stars evolved to produce the atoms required to form
life-sustaining systems, etc., all on roughly this characteristic timescale?
Despite the SdS cosmologys many attractive features, it may still be objected that
the specific geometrical structure of the SdS model entails a significant assumption
on the fundamental geometry of physical reality, for which there should be a reason;
i.e., the question arises: if the geometry is not determined by the world-matter, then
by what? While a detailed answer to this question has not been worked out (although,
see Sect. 3.3 in [40]), it is relevant to note that the local form of the SdS solution
which is only parametrically different from the cosmological form upon which our
analysis has been based; i.e. M 2 < 1/9 rather than M 2 > 1/9is the space-time
description outside a spherically symmetric, uncharged black hole, which is exactly
the type that is expected to result from the eventual collapse of every massive cluster
of galaxies in our Universeeven if it takes all of cosmic time for that collapse to
finally occur. In fact, there seems to be particular promise in this direction, given
that the singularity at r = 0 in the SdS cosmology is not a real physical singularity,
but the artifact of a derivative metric that must be ill-defined there, since space must
actually always have a finite radius according to the fundamental metric, Eq. (7.32).
This is the potential advantage that was noted above, of the finite minimum radius
of the foliation of de Sitter space defined in Eq. (7.3). And as far as that goes, it
should be noted that the Penrose-Hawking singularity theorem cannot be directly
applied when a positive cosmological constant is present [55], which is indeed
our case. For all these reasons, we might realistically expect a description in which
gravitational collapse leads to universal birth, and thus an explanation of the Big
Bang and the basic cosmic structure weve had to assume.
Along with such new possibilities as an updated description of collapse, and
of gravitation in general, that may be explored in a relativistic context when the
absolute background structure of cosmology is objectively assumed, the SdS cosmology, through its specific requirement that the observed rate of expansion should
be described exactly by the flat CDM scale-factor, has the distinct possibility to
explain why our Universe should have expanded as it evidently hasand therein lies
its greatest advantage.
129
References
1. A. Einstein, Ernst Mach. Phys. Z. 17, 101104 (1916)
2. A. Einstein, Maxwells influence on the evolution of the idea of physical reality, in James Clerk
Maxwell: A Commemoration Volume, ed. by J.J. Thomson (Cambridge University Press, 1931)
3. W. de Sitter, On Einsteins theory of gravitation, and its astronomical consequences. Third
paper. Mon. Not. R. Astron. Soc. 78, 328 (1917)
4. A. Eddington, The Mathematical Theory of Relativity, 2nd edn. (Cambridge University Press,
1923)
5. H. Weyl, Zur allgemeinen relativittstheorie. Phys. Z. 24, 230232 (1923) (English translation:
H. Weyl, Republication of: On the general relativity theory. Gen. Relativ. Gravitat. 35 1661
1666 (2009))
6. G. Lematre, Note on de Sitters universe. J. Math. Phys. 4, 188192 (1925)
7. H.P. Robertson, On relativistic cosmology. Philos. Mag. 5, 835848 (1928)
8. A. Einstein, Notiz zu der arbeit von A. Friedmann ber die krmmung des raumes. Z. Phys.
16, 228228 (1923)
9. H. Weyl, Massentrgheit und kosmos. Naturwissenschaften. 12, 197204 (1924)
10. E. Hubble, A relation between distance and radial velocity among extra-galactic Nebulae. Proc.
Nal. Acad. Sci. 15, 168173 (1929)
11. W. de Sitter, On the distances and radial velocities of extra-galactic nebulae, and the explanation
of the latter by the relativity theory of inertia. Proc. Nal. Acad. Sci. 16, 474488 (1930)
12. A. Einstein, W. de Sitter, On the relation between the expansion and the mean density of the
universe. Proc. Nal. Acad. Sci. 18(3), 213214 (1932)
13. A. Eddington, The Expanding Universe (Cambridge University Press, Cambridge, 1933)
14. A. Eddington, On the instability of Einsteins spherical world. Mon. Not. R. Astron. Soc. 90,
668678 (1930)
15. A.G. Riess et al., Observational evidence from supernovae for an accelerating universe and a
cosmological constant. Astronom. J. 116, 10091038 (1998)
16. S. Perlmutter et al., Measurements of and from 42 high-redshift supernovae. Astrophys.
J. 517, 565586 (1999)
17. A.G. Riess et al., Type Ia supernovae discoveries at z > 1 from the Hubble Space Telescope:
evidence for past deceleration and constraints on dark energy evolution. Astrophys. J. 607,
665687 (2004)
18. A.G. Riess et al., New Hubble Space Telescope discoveries of type Ia supernovae at z 1:
narrowing constraints on the early behavior of dark energy. Astrophys. J. 659, 98121 (2007)
19. W.M. Wood-Vasey et al., Observational constraints on the nature of dark energy: first cosmological results from the ESSENCE supernova survey. Astrophys. J. 666, 694715 (2007)
20. T.M. Davis et al., Scrutinizing exotic cosmological models using ESSENCE supernova data
combined with other cosmological probes. Astrophys. J. 666, 716725 (2007)
21. M. Kowalski et al., Improved cosmological constraints from new, old, and combined supernova
data sets. Astrophys. J. 686, 749778 (2008)
22. M. Hicken et al., Improved dark energy constraints from 100 new CfA supernova type Ia
light curves. Astrophys. J. 700, 10971140 (2009)
23. A.G. Riess et al., A 3% solution: determination of the Hubble constant with the Hubble Space
Telescope and Wide Field Camera 3. Astroph. J. 730(119), 118 (2011)
24. N. Suzuki et al., The Hubble Space Telescope cluster supernova survey. V. Improving the
dark-energy constraints above z > 1 and building an early-type-hosted supernova sample.
Astrophys. J. 746(85), 124 (2012)
25. D. Hinshaw et al., Nine-year Wilkinson microwave anisotropy probe (WMAP) observations:
cosmological parameter results. Astrophys. J. Suppl. Ser. 208(19), 125 (2013)
26. Planck Collaboration (P. A. R. Ade et al.), Planck 2013 results. XVI. Cosmological parameters
(2013), ArXiv:1303.5076 [astro-ph.CO]
27. H. Bondi, Cosmology (Cambridge University Press, Cambridge, 1960)
130
D. Janzen
28. A. Einstein, Kosmologische betrachtungen zur allgemeinen relativittstheorie. Sitzungsber. K.

Preuss. Akad. Wiss. pp. 142152 (1917) (English translation in H.A. Lorentz et al., (eds.) The
principle of relativity, pp. 175188. Dover Publications (1952))
29. W. de Sitter, On the relativity of inertia. Remarks concerning Einsteins latest hypothesis. Proc.
R. Acad. Amst. 19, 12171225 (1917)
30. A.S. Eddington, Space, Time, and Gravitation (Cambridge University Press, Cambridge, 1920)
31. A. Einstein, Zum kosmologischen problem der allgemeinen relativittstheorie. Sitzungsber.
Preuss. Akad. Wiss. 235237 (1931)
32. A. Einstein, The Meaning of Relativity, 2nd edn. (Princeton University Press, New Jersey,
1945)
33. A. Einstein, Zur Elektrodynamik bewegter Krper. Ann. der Phys. 17 (1905) (English translation in H.A. Lorentz et al., (eds.) The principle of relativity, pp. 3565. Dover Publications
(1952))
34. H. Putnam, Time and physical geometry. J. Philos. 64, 240247 (1967)
35. H. Minkowski, Space and time, in The Principle of Relativity, ed. by H.A. Lorentz, et al. (Dover
Publications, New York, 1952), pp. 7391
36. A. Flsing, E. Osers, trans. Albert Einstein: A Biography (Viking, New York, 1997)
37. H. Weyl, Philosophy of Mathematics and Natural Science (Princeton University Press, New
Jersey, 1949)
38. R. Geroch, General Relativity From A to B (University of Chicago Press, Chicago, 1978)
39. S.E. Rugh, H. Zinkernagel, Weyls principle, cosmic time and quantum fundamentalism (2010),
ArXiv:1006.5848 [gr-qc]
40. D. Janzen, A solution to the cosmological problem of relativity theory. Dissertation, University
of Saskatchewan 2012, http://hdl.handle.net/10388/ETD-2012-03-384
41. G. Lematre, Cosmological Application of Relativity. Rev. Mod. Phys. 21, 357366 (1949)
42. C.W. Misner, K.S. Thorne, J.A. Wheeler, Gravitation(W. H. Freeman and Company, 1973)
43. F.J. Dyson, Missed opportunities. Bull. Am. Math. Soc. 78(5), 635652 (1972)
44. T. Maudlin, Philosophy of Physics: Space and Time (Princeton University Press, New Jersey,
2012)
45. B. Greene, The Fabric of the Cosmos (Vintage Books, New York, 2004)
46. D. Janzen, Time is the denominator of existence, and bits come to be in it. Submitted to FQXis
2013 Essay Contest, It From Bit or Bit From It? (2013)
47. S. Weinberg, The cosmological constant problem. Rev. Mod. Phys. 61(1), 123 (1989)
48. A.H. Guth, Inflationary universe: A possible solution to the horizon and flatness problems.
Phys. Rev. D 23(2), 347356 (1981)
49. C.P. Burgess, The cosmological constant problem: why its hard to get dark energy from microphysics (2013), ArXiv: 1309.4133 [hep-th]
50. BICEP2 Collaboration (Ade, P. A. R., et al.): Planck 2013 results. XVI. Cosmological parameters (2013), ArXiv:1303.5076 [astro-ph.CO]
51. L.P. Grishchuck, Amplification of gravitational waves in an isotropic universe. Zh. Eksp. Theor.
Fiz. 67, 825838 (1974)
52. A.A. Starobinsky, Spectrum of relict gravitational radiation and the early state of the universe.
JETP Lett. 30, 682685 (1979)
53. V.A. Rubakov, M.V. Sazhin, A.V. Veryaskin, Graviton creation in the inflationary universe and
the grand unification scale. Phys. Lett. 115B(3), 189192 (1982)
54. T. Commissariat, Neil Turok urges caution on BICEP2 results (2014), http://physicsworld.
com/cws/article/news/2014/mar/18/neil-turok-urges-caution-on-bicep2-results. Accessed 18
March 2014
55. S.W. Hawking, R. Penrose, The singularities of gravitational collapse and cosmology. Proc. R.
Soc. Londo. A, Math. Phys. Sci. 314 (1519) 529548 (1970)
Chapter 8
Not on but of
Olaf Dreyer
Abstract In physics we encounter particles in one of two ways. Either as fundamental

constituents of the theory or as emergent excitations. These two ways differ by how
the particle relates to the background. It either sits on the background, or it is an
excitation of the background. We argue that by choosing the former to construct
our fundamental theories we have made a costly mistake. Instead we should think
of particles as excitations of a background. We show that this point of view sheds
new light on the cosmological constant problem and even leads to observable consequences by giving a natural explanation for the appearance of MOND-like behaviour.
In this context it also becomes clear why there are numerical coincidences between
the MOND acceleration parameter a0 , the cosmological constant and the Hubble
parameter H0 .
Which of Our Basic Physical Assumptions Are Wrong?

In theoretical physics we encounter particles in two different ways. We either
encounter them as fundamental constituents of a theory or as emergent entities. A
good example of the first kind is a scalar field. Its dynamics is given by the Lagrangian

d4x
1
()2 m2 2 +
2

.
(8.1)
An example of an emergent particle is a spin-wave in a spin-chain. It given by

|k =
N

n=1

nk
|n
exp 2 i
N
(8.2)
If |0 is the ground state of the spin-chain then |n is the state where spin n is flipped.
O. Dreyer (B)
Dipartimento di Fisica, Universit di Roma La Sapienza and Sez. Roma1 INFN,
P.le A. Moro 2, 00185 Rome, Italy
131
132
O. Dreyer
Of all the differences between these two concepts of particles we want to stress
one in particular: the relation these particles have to their respective backgrounds.
The scalar field is formulated as a field on spacetime. The above Lagrangian does
not include the metric. If we include the metric we obtain

1 ab
g a b m2 2 + .
(8.3)
d 4 x g
2
This way the scalar field knows about a non-trivial spacetime. Einstein told us
that this is not a one-way street. The presence of a scalar field in turn changes the
background. The proper equations that describe this interaction between matter and
background are obtained by adding the curvature tensor to the above Lagrangian.
This interaction does not change the basic fact about the scalar field that it is sitting
on spacetime. It is distinct from the spacetime that it sits on.
This is in contrast to the emergent particle. There is no clean separation possible
between the particle and the background, i.e. the ground state in this case. The spinwave above is an excitation of the background not an excitation on the background.
We can now state what basic assumption we think needs changing. Currently we
built our fundamental theories assuming the matter-on-background model. We will
argue here that this assumption is wrong and that instead all matter is of the emergent
type and that the excitation-of-background model applies.
Our basic assumption that is wrong: Matter does not sit on a background but is an
excitation of the background.
In the following we will do two things. First we will show that the assumption
that matter sits on a background creates problems and how our new assumption
avoids these problems. Then we will argue that the new assumption has observable
consequences by showing how MOND like behavior arises naturally.
The Cosmological Constant

The basic assumption that matter sits on a background directly leads to one of the
thorniest problems of theoretical physics: the cosmological constant problem. All
matter is described by quantum fields which can be thought of as a collection of
quantum harmonic oscillators. One feature of the spectrum of the harmonic oscillator
is that it has a non-vanishing ground state energy of /2. Because the quantum field
sits on spacetime the ground state energy of all the harmonic oscillators making up
the field should contribute to the curvature of space. The problem with this reasoning
is that it leads to a prediction that is many orders of magnitude off. In fact if one
assumes that there is some large frequency limit for the quantum field then the
energy density coming from this ground state energy is proportional to the fourth
power of this frequency:
4

(8.4)
8 Not on but of
133
If one chooses to be the Planck frequency then one obtains a value that is 123
orders of magnitude larger than the observed value of the cosmological constant.
This has been called the worst prediction of theoretical physics and constitutes one
part of the cosmological constant problem (see [1, 2] for more details).
Let us emphasize again how crucial the basic assumption of matter-on-background
is for the formulation of the cosmological constant problem. Because we have separated matter from the background we have to consider the contributions coming from
the ground state energy. We have to do this even when the matter is in its ground
state, i.e. when no particles are present. This is to be contrasted with the excitationof-background model. If there are no waves present there is also no ground state
energy for the excitations to consider. The cosmological constant problem can not
be formulated in this model.
There are two objections to this reasoning that we have to deal with. The first
objection is that although the above reasoning is correct it is also not that interesting
because there is no gravity in the picture yet. We will deal with this objection in
the next section. The other objection concerns the ontological status of the vacuum
fluctuations. Isnt it true that we have observed the Casimir effect? Since the Casimir
effect is based on vacuum fluctuations the cosmological constant problem is not
really a problem that is rooted in any fundamental assumption but in observational
facts. For its formulation no theoretical foundation needs to be mentioned. We will
deal with this objection now.
There are two complementary views of the Casimir effect and the reality of vacuum
fluctuations. In the more well known explanation of the Casimir force between two
conducting plates the presence of vacuum fluctuations is assumed. The Casimir force
then arises because only a discrete set of frequencies is admissible between the plates
whereas no such restriction exists on the outside. The Casimir effect between two
parallel plates has been observed and this has led to claims that we have indeed seen
vacuum fluctuations. This claim is not quite true because there is another way to think
about the Casimir effect. In this view vacuum fluctuations play no role. Instead, the
Casimir effect results from forces between the currents and charges in the plates. No
vacuum diagrams contribute to the total force [3]. If the emergent matter is described
by the correct theory, say quantum electrodynamics, we will find the Casimir effect
even if there are no vacuum fluctuations.
We see that the cosmological constant problem can be seen as a consequence of
us viewing matter as sitting on a background. If we drop this assumption we can not
even formulate the cosmological constant problem.
Gravity
The above argument for viewing matter as an excitation of a background is only useful
if we can include gravity in the picture. In [4] we have argued that this can be achieved
by regarding the ground state itself as the variable quantity that is responsible for
gravity. In the simplest case the vacuum is described by a scalar quantity . If we
134
O. Dreyer
assume that the energy of the vacuum is given by

E=
1
8

d 3 x ( )2 ,
(8.5)
then we can calculate the force between two objects mi , i = 1, 2. If we introduce the
gravitational mass of an object by
1
m=
4

d ,
(8.6)
where m is the boundary of the object m, then the force between them is given by
F=
m1 m2
.
r2
(8.7)
Usually we express Newtons law of gravitation not in terms of the gravitational

masses mi , i = 1, 2, but in terms of the inertial masses mi , i = 1, 2. In [4] we have
argued that the inertial mass of an object is given by
m=
2m
m,
3a
(8.8)
where a is the radius of the object. In terms of the inertial masses mi Newtons law
then takes the usual form
m1 m2
(8.9)
F=G 2 ,
r
where G has to be calculated from the fundamental theory:

G=
3a
2m
2
(8.10)
In this picture of gravity the metric does not play a fundamental role. Gravity appears
because the ground state depends on the matter.
MOND as a Consequence
The picture of gravity that we have given in the last section is valid only for zero
temperature. If the temperature is not zero we need to take the effects of entropy into
account and instead of looking at the energy we have to look at the free energy
F = E TS.
(8.11)
8 Not on but of
135
We thus have to determine the dependence of the entropy S on the temperature T

and the ground state . The entropy should not depend on directly because every
corresponds to the same ground state. The entropy should only be dependent on
changes of . If we are only interested in small values of T we find
S = T ( )2 ,
(8.12)
for some constant . The total free energy is thus

F = E TS
(8.13)
= E(1 8 T ).
2
(8.14)
We see that a non-zero temperature does not change the form of the force but just its
strength. The new gravitational constant is given by
GT =
1
GT =0 .
1 T2
(8.15)
The situation changes in an interesting way if there is a large maximal length scale
Lmax present in the problem. The contributions to the entropy of the form ()2
come from excitations with a wavelength of the order
L = ||1 .
(8.16)
If this wavelength L is larger than Lmax than these excitations should not exist and
thus not contribute to the entropy. Instead of a simple ()2 term we should thus
have a term of the form
C(Lmax | |) ( )2 ,
(8.17)
where the function C is such that it suppresses contributions from excitations with
wavelengths larger than Lmax . For wavelengths much smaller than the maximal wavelength we want to recover the usual contributions to the entropy. Thus, if Lmax | |
is much bigger than unity we want C to be one:
C(x) = 1,
for x 1.
(8.18)
For x 1 we assume that the function C possesses a series expansion of the form
C(x) = x + x 2 + . . . .
(8.19)
For small values of Lmax | | we thus find that the dependence of the entropy on
is of the form

2
(8.20)
T d 4 x Lmax ||3 .
136
O. Dreyer
It is here that we make contact with the Lagrangian formulation of Milgroms odd
new dynamics (or MOND, see [5, 6] for more details). In [7] Bekenstein and Milgrom
have shown that a Lagrangian of the form

( )2
1
2
a F
d x +
8 G 0
a02
3
(8.21)
gives rise to MOND like dynamics if the function F is chosen such that
(x) = F (x 2 ).
(8.22)
Here (x) is the function that determines the transition from the classical Newtonian
regime to the MOND regime. It satisfies
(x) =
1
x
x1
x1
From this it follows that the function F satisfies
x1
x2
2
F(x ) =
2/3 x 3 x 1
(8.23)
(8.24)
The behavior of the Lagrangian is then

a02 F
| |2
a02
| |2
| |3
2
3a0
a01 | | 1
a01 | | 1
(8.25)
This is exactly the behavior of the free energy that we have just derived if we make
the identification
2
= T 2 Lmax .
3a0
(8.26)
There are currently two candidates for a maximal length scale Lmax . These are the
Hubble scale
(8.27)
LH = cH0

and the cosmic horizon scale

L =
1
.
(8.28)
It is a remarkable fact of the universe that we live in that both of these length scales
satisfy the relationship that we derived in (8.26) if we further assume that the constant
T 2 is of order one. We have thus established a connection between the acceleration
parameter a0 , the cosmological constant , and the Hubble parameter H0 . In standard
cosmology these coincidences remain complete mysteries.
8 Not on but of
137
Discussion
Particles are either fundamental or they are emergent. If they are fundamental they
are sitting on a background; if they are emergent they are excitations of a background.
Rather than being a purely philosophical issue we have argued that this distinction is
important and that the assumption that particles are fundamental is wrong. Assuming
instead that particles are emergent leads to the resolution of theoretical problems as
well as having observational consequences. We have argued that the cosmological
constant problem as it is usually formulated can not even be stated if we think of
particles as excitations of a background. Also, we have shown that this picture gives
a straight forward way of understanding the appearance of MOND like behavior in
gravity. The argument also makes clear why there are numerical relations between the
MOND parameter a0 , the cosmological constant , and the Hubble parameter H0 .
Our derivation of MOND is inspired by recent work [817] that uses Verlindes
derivation of Newtons law of gravity [18] as a starting point. Our derivation differs
from these in that it does not rely on holography in any way. Our formulae for the
entropy are all completely three-dimensional.
Are there other possible consequences of this point of view that might be observable? In [19] we have argued that in fact there are. The backgrounds that the particles
are excitations of appear in nature as the result of phase transitions. Was our cosmological background also a result of such a phase transition? If so, what would be the
consequence? One feature that distinguishes phase transitions is the disappearance
of scales. This manifests itself in the appearance of power laws that describe the system. Absence of scale and power laws have made an appearance in cosmology lately.
The mechanism of inflation produces a scale-free spectrum of metric perturbations
in the early universe that has been observed recently. In [19] we argue that this is
no coincidence. In fact we argue that it is precisely such a spectrum that would be
expected from a background-producing phase transition in the early universe. Furthermore the precise tilt of the spectrum is related to the anomalous scaling of the
order parameter and thus to the fundamental discreteness of nature.
We thus see that the point of view advocated here is not just of philosophical
interest. We shed light on the cosmological constant problem, find a connection with
MOND phenomenology, and provide a simple alternative to inflation.
Acknowledgments I would like to thank Mamazim Kassankogno, and Stevi Roussel Tankio
Djiokap for helpful discussion and the Foundational Questions Institute, FQXi, for financial support
and for creating an environment where it is alright to play with unorthodox ideas.
References
1.
2.
3.
4.
S. Weinberg, The cosmological constant problem. Rev. Mod. Phys. 61(1), 123 (1989)
S. Carroll, The cosmological constant. Living Rev. Relativ. 4, 80 (2001)
R.L. Jaffe, Casimir effect and the quantum vacuum. Phys. Rev. D 72(2), 5 (2005)
O. Dreyer, Internal relativity (2012). arXiv:1203.2641
138
O. Dreyer
5. M. Milgrom, A modification of the Newtonian dynamics as a possible alternative to the hidden

mass hypothesis. Astrophys. J. 270, 365370 (1983)
6. B. Famaey, S. McGaugh, Modified Newtonian dynamics (MOND): observational phenomenology and relativistic extensions (2011). Arxiv preprint: arXiv:1112.3960v2
7. J. Bekenstein, M. Milgrom, Does the missing mass problem signal the breakdown of Newtonian
gravity? Astrophys. J. 286, 714 (1984)
8. C.M. Ho et al., Quantum gravity and dark matter. Gen. Relativ. Gravit. 43(10), 25672573
(2011)
9. C.M. Ho et al., Cold dark matter with MOND scaling. Phys. Lett. B 693(5), 567570 (2010)
10. P.V. Pikhitsa, MOND reveals the thermodynamics of gravity (2010). Arxiv preprint:
arXiv:1010.0318
11. F.R. Klinkhamer, Entropic-gravity derivation of MOND. Mod. Phys. Lett. A 27, 6 (2012)
12. F.R. Klinkhamer, M. Kopp, Entropic gravity, minimum temperature, and modified Newtonian
dynamics (2011). Arxiv preprint: arXiv:1104.2022v6
13. X. Li, Z. Chang, Debye entropic force and modified Newtonian dynamics. Commun. Theor.
Phys. 55, 733736 (2011)
14. J.A. Neto, Nonhomogeneous cooling, entropic gravity and MOND theory. Int. J. Theor. Phys.
50, 35523559 (2011)
15. V.V. Kiselev, S.A. Timofeev, The holographic screen at low temperatures (2010). Arxiv preprint:
arXiv:1009.1301v2
16. E. Pazy, N. Argaman, Quantum particle statistics on the holographic screen leads to modified
Newtonian dynamics. Phys. Rev. D 85(10), 104021 (2012)
17. L. Modesto, A. Randono, Entropic corrections to Newtons law (2010). Arxiv preprint:
arXiv:1003.1998v1
18. E. Verlinde, On the origin of gravity and the laws of Newton. JHEP 1104, 029 (2011)
19. O. Dreyer, The world is discrete (2013). arXiv:1307.6169
Chapter 9
Patterns in the Fabric of Nature

Steven Weinstein
Abstract From classical mechanics to quantum field theory, the physical facts at one
point in space are held to be independent of those at other points in space. I propose
that we can usefully challenge this orthodoxy in order to explain otherwise puzzling
correlations at both cosmological and microscopic scales.
Nature uses only the longest threads to weave her patterns, so that each small piece of her
fabric reveals the organization of the entire tapestry. (The Character of Physical Law, Richard
Feynman)
Introduction
Despite radical differences in their conceptions of space, time, and the nature of
matter, all of the physical theories we presently usenon-relativistic and relativistic,
classical and quantumshare one assumption: the features of the world at distinct
points in space are understood to be independent. Particles may exist anywhere,
independent of the location or velocity of other particles. Classical fields may take
on any value at a given point, constrained only by local constraints like Gausss law.
Quantum field theories incorporate the same independence in their demand that field
operators at distinct points in space commute with one another.
The independence of physical properties at distinct points is a theoretical
assumption, albeit one that is grounded in our everyday experience. We appear to
be able to manipulate the contents of a given region of space unrestricted by the
contents of other regions. We can arrange the desk in our office without concern for
the location of the couch at home in our living room.
S. Weinstein (B)
Perimeter Institute for Theoretical Physics, 31 Caroline St N,
Waterloo, ON N2L 2Y5, Canada
e-mail: [email protected]; [email protected]
S. Weinstein
U. Waterloo Department of Philosophy, 200 University Ave West,
Waterloo, ON N2L 3G1, Canada
139
140
S. Weinstein
Yet there are realms of physical theory, more remote from everyday experience
and physical manipulation yet accessible to observation, in which there appear to be
striking correlations between the values of physical properties at different points in
space. Quantum theory predicts (and experiment confirms) the existence of strongly
correlated measurement outcomes apparently inexplicable by classical means. I refer,
of course, to the measurements of the properties of pairs of particles originally envisioned by Einstein, Podolsky and Rosen (EPR) [7], measurements that suggested
to EPR the incompleteness of the theory. Bell [2] showed that no theory satisfying two seemingly natural conditions could possibly account for these correlations.
The condition known variously as Bell locality, strong locality or factorizability has
been endlessly analyzed. It is violated by quantum mechanics, as well as alternative
theories such as the deBroglie-Bohm theory. The other condition, statistical independence (a.k.a. measurement independence or setting Independence), has only rarely
been called into question. A theory that violates statistical independence is one in
which contains non-local degrees of freedom, which is to say that the properties
of a physical system at one point may be constrained in a lawlike manner by the
properties at other points which are not in causal contact with the first point.
On a completely different scale, the electromagnetic radiation that pervaded the
early universethe remnants of which form the cosmic microwave background
appears to have been extraordinarily homogeneous. It is strikingly uniform, yet the
theories that describe the early universeclassical electrodynamics (for the radiation) and general relativity (for the expanding spacetime the radiation fills)do not
stipulate any sort of restrictions or correlations that would go anywhere near explaining this. To the extent that they have been explained at all, it has been through the
postulation of an as-yet unobserved field known as the inflaton field.
What I want to do here is raise the possibility that there is a more fundamental
theory possessing nonlocal constraints that underlies our current theories. Such a
theory might account for the mysterious nonlocal effects currently described, but not
explained, by quantum mechanics, and might additionally reduce the extent to which
cosmological models depend on finely tuned initial data to explain the large scale
correlations we observe. The assumption that spatially separated physical systems
are entirely uncorrelated is a parochial assumption borne of our experience with the
everyday objects described by classical mechanics. Why not suppose that at certain
scales or certain epochs, this independence emerges from what is otherwise a highly
structured, nonlocally correlated microphysics?
Nonlocal Constraints
All physical theories in current use assume that the properties of physical systems at
different points in space are independent. Correlations may emerge dynamically
many liquids crystallize and develop a preferred orientation when cooled, for
examplebut the fundamental theories permit any configuration as an initial condition.
9 Patterns in the Fabric of Nature
141
For example, consider the straightforward and simple theory of the free massless
scalar field (
x ). A scalar field is simply an assignment of a single number (a scalar
rather than a vector) to every point in space and time. The evolution of the field is
given by the well-known wave equation
x , t)
2 (
= c2 2 (
x , t) ,
t 2
in conjunction with initial data (
x ) and (
x )/t giving the value of the field and
its rate of change at some initial time. This initial data can be specified arbitrarilyit
is unconstrained.
A more realistic field theory is the classical electrodynamics of Maxwell, which
does feature constraints. In Maxwells theory, we have a pair of coevolving fields, the
electric field E and the magnetic field B . The fields are described by vectors at each
point rather than scalars. The significant difference between the electromagnetic field
and the free scalar field is that the electric and magnetic fields may not be specified
x) = 0
arbitrarily. They are subject to constraints E (
x ) = 4(
x ) and B(
which hold at every point x in space. The divergence of the electric field at any
given point must be equal to a multiple of the charge density at that point, and the
divergence of the magnetic field must be zero. The divergence is a measure of the
outflow of the field in the neighborhood of a point, and the two constraints tell us
respectively that any such outflow of the electric field is due to the presence of a
charge at that point acting as a source, while the magnetic field can have no sources
(there are no magnetic charges). These constraints are local in that they provide a
constraint on values of the field at each point that does not depend on values of the
field or the charge distribution at other points.
What would a nonlocal constraint look like? Heres a candidate: E (

x) =
4(
x (1, 1, 1)). This says that the divergence of the electric field at one point is
equal to a constant times the charge density at a point which is one unit away in the
x, y and z directions. But this constraint is hardly worthy of the name, since it only
holds at a single time: unlike the constraint E(
x ) = 4(
x ), it is not preserved
by the equations of motion (Maxwells equations for the field and the Lorentz force
law for the charge distribution). I.e., it will not continue to hold as the field evolves.
Since it is not preserved, it does not hold at arbitrary moments of time, hence it is
not a true regularity or law.
Lets return to simple classical particle mechanics for an example of a true nonlocal constraint, one that is conserved in time. The particles are characterized by their
positions and their momenta. The constraint we will impose is that the total momentum (the sum of the momenta of each of the particles) is zero. This is a constraint
because we cannot specify the momentum freely for each particle; if we know the
momentum of all but one of the particles, the momentum of the other particle is fixed.
It is nonlocal, because the momentum of that particle is a function of the momenta
of particles some distance away. Unlike the first nonlocal constraint we considered,
it is conserved, since total momentum is a conserved quantity in particle mechanics.
142
S. Weinstein
Fig. 9.1 Timelike

compactification
But it is not a particularly interesting constraint, because all but one of the momenta
may be freely specified. Whereas the two constraints in electromagnetism reduce the
number of degrees of freedom from 6 to 4 at each point in space (so that there are
only two-thirds the number of degrees of freedom), this constraint only reduces the
total number of degrees of freedom by one.
A more interesting nonlocal constraint may be obtained by considering once more
the wave equation, this time in one space dimension (for simplicity). Suppose that
the spacetime on which the field takes values is compactified in the time direction,
so that the entirety forms a cylinder (see Fig. 9.1).
The solutions must clearly be periodic, and this amounts to imposing a nonlocal
constraint. More specifically, whereas in the ordinary initial value problem, initial
data may be any smooth functions (x, 0) and t (x, 0) (where t stands for /t),
we now require that (x, 0) = (x, T ) and t (x, 0) = t (x, T ), where T is the
circumference of the cylinder. This is just to say that the time evolution from 0 to
T must return us to the same starting point. What are the constraints, then, on this
initial data? They are essentially those data that can be written as sums of sine or
T
(for any integer value of n).1
cosine waves with wavelength 2n
The restriction to a discrete (though infinite) set of plane waves means that initial
data do not have compact support; they are periodic. However, for sufficiently small
1 Solutions to the wave equation can be written as sums of plane waves, with Fourier space repikt + G(k)e
ikt . Since these plane waves must have period T (in the
resentation (k,
t) = F(k)e
preferred frame dictated by the cylinder), we have a constraint k = 2n
T (where n is a positive or
negative integer), so that initial data are no longer arbitrary smooth functions of k
(k,
0) = F(k)
+ G(k)
t (k, 0) = ik( F(k)

G(k))
but are rather constrained by the requirement k =
2n
T .
Thus the initial data are the functions
2n
1 2n
(x, 0) =
(
, 0)ei T x dk
T
T n=
2n
1 2n
t (
t (x, 0) =
, 0)ei T x dk
T
T n=
i.e., they consist of arbitrary sums of plane waves with wave number k =
of n.
2n
T , for any integer value
143
x, the local physics is indistinguishable from the local physics in ordinary spacetime. Only at distance scales on the order of T does the compact nature of the time
direction become evident in the repetition of the spatial structure. Thus we have here
an example of a nonlocal constraint which can give the appearance of unconstrained
local degrees of freedom.
Now, this spacetime obviously has closed timelike curves, and it is interesting
to note that under such conditions, classical computers are as powerful as quantum
computers [1]. Thus there is some reason to think that a nonlocal constraint might
allow one to mimic other quantum phenomena using classical physics. In any event,
we will now proceed to a discussion of the way in which the presence of nonlocal
constraints opens the door to a little-explored loophole in Bells theorem, in that their
presence undermines the statistical independence assumption required for the proof
of the theorem.
Bells Theorem
Einstein believed quantum theory to be an incomplete description of the world, and he
and his collaborators Podolsky and Rosen attempted to show this in their 1935 paper
[7]. The argument involves a pair of particles specially prepared in an entangled state
of position and momentum.2 Quantum mechanics makes no definite predictions for
the position and momentum of each particle, but does make unequivocal predictions
for the position or momentum of one, given (respectively) the position or momentum
of the other. EPR argued on this basis that quantum mechanics must be incomplete,
since the measured properties of one particle can be predicted with certainty following measurement of the properties of the other, even when these measurements are
spatially separated and thus causally independent.3
In 1964, John Bell proved a result based on David Bohms streamlined version of
the EPR experiment [2, 4]. Instead of positions and momenta, Bohm focuses on the
spins of a pair of particles (in this case fermions). Prepared in what has come to be
known as a Bell state,
1
= (|+x A |x B |x A |+x B ),
2
(9.1)
quantum mechanics predicts that a measurement of the component of spin of particle

A in any direction (e.g., the z direction) is as likely to yield +1 as 1 (in units of /2),
and so the average value A is 0. However, quantum mechanics also indicates that an
outcome of +1 for a measurement of the spin of A in the z direction is guaranteed to
2
The state used by EPR is an eigenstate of the operators representing the sum of the momenta and
the difference of the positions of the two particles.
3 The argument of the EPR paper is notoriously convoluted, but I follow [9] in regarding this as
capturing Einsteins understanding of the core argument.
144
S. Weinstein
yield an outcome of 1 for B for a measurement of the spin of B in the z direction,

etc. This is directly analogous to the correlations between position and momentum
measurements in the original EPR experiment.
In and of themselves, these phenomena offer no barrier to a hidden-variable theory,
since it is straightforward to explain such correlations by appealing to a common
causethe sourceand postulating that the particles emanate from this source in
(anti)correlated pairs. However, one must also account for the way that the anticorrelation drops off as the angle between the components of spin for the two particles
increases (e.g., as A rotates from x toward z while B remains oriented along the x
direction). It was Bells great insight to note that the quantum theory implies that the
anticorrelation is held onto more tightly than could be accounted for by any local
theorythat is, any theory satisfying the seemingly natural condition known in the
literature variously as strong locality, Bell locality, or factorizability. Bell showed
that the predictions of a local theory must satisfy a certain inequality, and that this
inequality is violated by quantum theory for appropriate choices of the components
of spin to be measured. Bells result was widely understood to provide a barrier
to the sort of completion of quantum mechanics considered by Einstein. That is,
Einsteins hope for a more fundamental theory underlying quantum theory would
have to violate strong locality, of which more below.
However, there is a further assumption known as the statistical independence
assumption (also known as measurement independence) that is necessary for Bells
result. This assumption is quite closely related to the assumption of local degrees
of freedom, or the absence of nonlocal constraints. Without the assumption, Bells
result does not go through, and the possibility re-emerges of a local completion of
quantum theory after all.4
Rather than repeat the derivation of Bells result, let me just focus on the meaning
of the two crucial assumptions of strong locality and statistical independence. The
physical situation we are attempting to describe has the following form:
A source (represented by the ellipse) emits a pair of particles, or in some other way
causes detectors A and B to simultaneously (in some reference frame) register one of
two outcomes. The detectors can be set in one of two different ways, corresponding,
in Bohms version of the EPR experiment, to a measurement of one of two different
components of spin.
Let us now suppose that we have a theory that describes possible states of the
particles and which gives rise to either probabilistic or deterministic predictions as
to the results of various measurements one might make on the particles. The state of
the particle will be represented by a parameter , describing either a discrete set of
) and B(b,
) correspond
states 1 , 2 . . . or a continuous set. The expressions A(a,
4
A more detailed discussion of Bells derivation and the role of the Statistical Independence
assumption can be found in [18].
145
to the expected (average) values of measurements of properties a and b at detectors

A and B (respectively) in a given state . (The appeal to average values allows for
stochastic theories, in which a given might give rise to any number of different
outcomes, with various probabilities.)
In general, one might suppose that A also depends on either the detector setting
, b, B)), and one might suppose the
b or the particular outcome B (i.e., A = A(a,
That it does not, that the expectation value A in a given state does not
same for B.
depend on what one chooses to measure at B, or on the value of the distant outcome
B (and vice-versa) is Bells strong locality assumption. Given this assumption, one
can write the expression E(a, b, ) for the expected product of the outcomes of
measurements of properties a and b in a given state as
) B(b,
E(a, b, ) = A(a,
).
(9.2)
This strong locality is also known as factorizability, deriving as it does from the
fact that the joint probability of a pair of outcomes can be factorized into the product
of the marginal probabilities of each outcome. We can thus represent the analysis of
the experimental arrangement in this way, where the expression for E(a, b, ) in the
center encodes the assumption of strong locality:
Now the further assumption required for Bells result is that the probability of a
given state is independent of the detector settings. In other words, Bell assumes
that the theory will be one in which
P(|a, b) = P().
(9.3)
This is statistical independence. For example, we might suppose that the theory tells
us that one of three states 1 , 2 , 3 will be generated by our particle preparation procedure. The statistical independence condition tells us that the likelihood of obtaining
any one of these states is independent of how the detectors will be set at the time
of detection. In other words, knowledge of the future settings of the detectors (their
settings at the time the particles arrive) does not provide any further information as
to which of the three states was emitted.5
The assumption of statistical independence has been called into question only
infrequently, but when it has, the critique has often been motivated by an appeal
to the plausibility of Lorentz-invariant backward causation (also known as retrocausation) whereby the change of detector settings gives rise to effects that propagate
along or within the backward lightcone and thereby give rise to nontrivial initial
5
Actually, a slightly weaker condition than SI is sufficient to derive the CHSH inequality. See [8]
and the discussion thereof in Sect. 3.3.1 of [14].
146
S. Weinstein
Fig. 9.2 EPR: Spacetime

diagram
correlations in the particle properties encoded in (e.g., [6, 13, 16]). In my [18] I offer
a critique of this way of thinking. Here instead I would like to offer a rather different
sort of motivation for thinking that statistical independence might be violated, coming
as promised from the possibility of nonlocal constraints.
Depicted in Fig. 9.2 is a run of the EPR-Bohm experiment in which the setting of
A is changed from a1 to a2 while the particles (or whatever it is that emanates from
the source) are in flight. What we have here is a pair of particles traveling toward
detectors A and B, with detector A switching from setting a1 to a2 while the particles
are in flight, and detector B simply set to b1 .
Lets again suppose that the particles are in one of three states 1 , 2 , 3 .
According to classical, relativistic physics, the detector settings a2 and b1 are determined by the goings-on in their past lightcones, which include the particle preparation
event but also far more. Suppose that setting a2 is compatible with a variety of initial
data at the time of preparation, and the same for b1 . Let a2 be the presumably large
subset of microscopic states (in the past lightcone of the detection event) consistent
with a final detector setting of a2 , and let b1 be those states compatible with b1 .
Though the particle preparation is contained in the past lightcones of the detection
events, let us suppose that the state of the particles, 1 , 2 , or 3 , does not play a
dynamical role in determining the setting of either detector. The question at hand is
whether there is any reason to think that, nevertheless, the state of the particles is
correlated with the detector setting, which is to say whether the theory constrains
the state of the particles on the basis of a2 and b1 .
Now, if the underlying theory is one in which local degrees of freedom are
independent, there is no reason to think that knowledge of a2 and b1 should
tell us anything about which of 1 , 2 , 3 are realized. On the other hand, if there
are nonlocal constraints, then it may well be otherwise. Suppose that a2 is compatible with 1 and 2 but incompatible with 3 . In other words, suppose that there
are no initial microstates that generate a2 which are consistent with the particle pair
starting in state 3 . Then we already have a violation of the statistical independence
condition, without even bothering yet to consider correlations with the other detector
B.
147
Of course, there may be, and typically are, many things going on in the past
lightcone of a detection event at the time the particle pair is produced. Most of these
will at least have the appearance of being irrelevant to the final setting of the detector.
There is certainly no guarantee that a nonlocal constraint will generate the kind of
correlations between detector settings and specially prepared particles that we are
talking about. The precise nature of the nonlocal constraint or constraints that could
explain quantum correlations is a decidedly open question.
Superdeterminism, Conspiracy and Free Will

The idea that the rejection of statistical independence involves preexisting and
persisting correlations between subsystems has been broached before, under the
headings conspiracy theory, hyperdeterminism, and superdeterminism. From
here on, let us adopt superdeterministic as a generic term for this way of thinking
about theories that violate this condition. Bell [3], Shimony [15], Lewis [11] and others have suggested that superdeterministic theories imply some sort of conspiracy on
the part of nature. This is frequently accompanied by the charge that the existence
of such correlations is a threat to free will. Let me briefly address these worries
before returning to the big picture.
The idea that postulating a correlation between detector settings and particle
properties involves a conspiracy on the part of nature appears to derive from the
idea that it amounts to postulating that the initial conditions of nature have been set
up by some cosmic conspirator in anticipation of our measurements. It seems that
the conspiracy theorist is supposing that violations of statistical independence are
not lawlike, but rather are ad hoc. But nonlocal constraints are lawlike, since (by
definition), we require them to be consistent with the dynamical evolution given by
the laws of motion. If they exist, they exist at every moment of time. This is no more
a conspiracy than Gausss law is a conspiracy.
Another worry about giving up statistical independence and postulating generic
nonlocal, spacelike correlations has to do with a purported threat to our free will.
This particular concern has been the subject of renewed debate in the last couple of years, prompted in part by an argument of Conway and Kochen [5]. The
core of the worry is that if detector settings are correlated with particle properties, this must mean that we cannot freely choose the detector settings. However,
as t Hooft [17] points out, this worry appears to be based on a conception of free
will which is incompatible with ordinary determinism, never mind superdeterminism.
Hume [10] long ago argued that such a conception of free will is highly problematic,
in that it is essential to the idea that we freely exercise our will that our thoughts are
instrumental in bringing about, which is to say determining, our actions.
148
S. Weinstein
The Cosmos
So much for the possible role of nonlocal constraints in underpinning quantum
phenomena. The other point of interest is early universe cosmology. Our universe
appears to have emanated from a big bang event around 14 billion years ago, and to
have been highly homogeneous for quite some time thereafter. The cosmic microwave
background radiation is a fossil remnant of the time, around 400,000 years into the
universes existence, when radiation effectively decoupled from matter, and this radiation appears to be quite evenly distributed across the sky, with slight inhomogeneities
which presumably seeded later star and galaxy formation.
The task of explaining the homogeneity of the early matter distribution is known
as the horizon problem. This, along with the flatness problem and monopole problem, were for some time only explained by fine-tuning, which is to say that they were
not really explained at all. Later, inflationary models entered the picture, and these
provide a mechanism for generating inhomogeneity in a more generic fashion. However, these models are still speculativethere is no direct evidence for an inflaton
fieldand moreover inflation itself requires rather special initial conditions [12].
The existence of a nonlocal constraint on the matter distribution and on the state of
the gravitational field might address one or more of these problems without recourse
to inflation. Certainly, a detailed description of the very early universe requires few
variables, since the universe looks essentially the same from place to place with
respect to both matter distribution (high temperature, homogeneous) and spatial
structure (flat). A reduction in the number of variables is what we would expect from
a constrained system, and any constraint demanding that the matter distribution is
identical from place to place is indeed nonlocal. On the face of it, this constraint
is not preserved under dynamical evolution because of the action of gravity. One
might speculate, though, that the constraint holds between matter and gravitational
degrees of freedom, and that the early universe is simply a demonstration of one
way to satisfy it. The interplay of gravity and matter mix up the degrees of freedom
as time goes on, and the current remnants of these macroscopic correlations are the
quantum correlations discussed above.
Conclusion
The idea of using nonlocal constraints to account for the large-scale matter distribution
in the universe and the large-scale spacetime structure of the universe is interesting
but highly speculative, and the idea that these same constraints might account for
quantum correlations as well is even more speculative. The most conservative strategy of exploration would be to ignore cosmological scenarios and instead focus on
the persistent and experimentally repeatable correlations in the quantum realm. But I
think it is worth considering a connection between the two, if for no other reason than
the fact that it has proven difficult to construct a testable and sensible quantum theory
149
of gravity, suggesting that the relation between gravitation and quantum phenomena
might be different from anything heretofore explored.
A more conservative approach focusing just on quantum phenomena might ponder the way in which the ordinarily superfluous gauge degrees of freedom of modern
gauge theories might serve as nonlocal hidden variables. The vector potential in electrodynamics, for example, ordinarily plays no direct physical role: only derivatives
of the vector potential, which give rise to the electric and magnetic fields, correspond to physical degrees of freedom in classical and quantum electrodynamics.
The Aharonov-Bohm effect shows that the vector potential does play an essential
role in the quantum theory, but the effect is still gauge-invariant. One might nevertheless conjecture that there is an underlying theory in which the potential does play a
physical role, one in which the physics is not invariant under gauge transformations.
It may be impossible for us to directly observe the vector potential, and the uncertainties associated with quantum theory may arise from our ignorance as to its actual
(and nonlocally constrained) value. From this perspective, quantum theory would be
an effective theory which arises from modding out over the gauge transformations,
with the so-called gauge degrees of freedom being subject to a nonlocal constraint
and accounting for the correlations we observe in EPR-type experiments.
I would conclude by reminding the reader that the sort of nonlocality under
discussion in no way violates either the letter or the spirit of relativity. No influences
travel faster than light. The idea is simply that there are correlations between spatially separate degrees of freedom, and thus that the fabric of nature is a more richly
structured tapestry than we have heretofore believed.
References
1. S. Aaronson, J. Watrous, Closed timelike curves make quantum and classical computing equivalent. arXiv:0808.2669v1
2. J.S. Bell, On the Einstein-Podolsky-Rosen paradox. Physics 1, 195200 (1964)
3. J.S. Bell, La nouvelle cuisine, in John S. Bell on the Foundations of Quantum Mechanics, ed.
by M. Bell, K. Gottfried, M. Veltman (World Scientific, Singapore, 2001), pp. 216234
4. D. Bohm, Quantum Theory (Prentice-Hall, Englewood Cliffs, 1951)
5. J. Conway, S. Kochen, The free will theorem. Found. Phys. 36, 14411473 (2006)
6. O. De Costa Beauregard, S-matrix, Feynman zigzag and Einstein correlation. Phys. Lett. A67,
171174 (1978)
7. A. Einstein, B. Podolsky, N. Rosen, Can quantum-mechanical description of reality be considered complete? Phys. Rev. 47, 777780 (1935)
8. A. Fahmi, Non-locality and classical communication of the hidden variable theories (2005).
arxiv:quant-ph/0511009
9. A. Fine, The Shaky Game: Einstein Realism and the Quantum Theory (University of Chicago
Press, Chicago, 1986)
10. D. Hume, A Treatise of Human Nature (Oxford University Press, Oxford, 2000). (original
publication)
11. P. Lewis, Conspiracy theories of quantum mechanics. Br. J. Philos. Sci. 57, 359381 (2006)
12. R. Penrose, Difficulties with inflationary cosmology, in Proceedings of the 14th Texas Symposium on Relativistic Astrophysics (New York Academy of Sciences 1989), pp. 249264
150
S. Weinstein
13. H. Price, Times Arrow and Archimedes Point: New Directions for the Physics of Time (Oxford
University Press, Oxford, 1996)
14. M.P. Seevinck, Parts and Wholes: An Inquiry into Quantum and Classical Correlations. PhD
thesis, Utrecht University (2008). arxiv:0811.1027
15. A. Shimony, An exchange on local beables, in Search for a Naturalistic World View, vol. II
(Cambridge University Press, Cambridge, 1993), pp. 163170
16. R.I. Sutherland, Bells theorem and backwards in time causality. Int. J. Theor. Phys. 22, 377384
(1983)
17. G. t Hooft. On the free will postulate in quantum mechanics (2007). arXiv:quant-ph/0701097
18. S. Weinstein, Nonlocality without nonlocality. Found. Phys. 39, 921936, (2009).
arXiv.org:0812.0349
Chapter 10
Is Quantum Linear Superposition

an Exact Principle of Nature?
Angelo Bassi, Tejinder Singh and Hendrik Ulbricht
Abstract The principle of linear superposition is a hallmark of quantum theory.

It has been confirmed experimentally for photons, electrons, neutrons, atoms, for
molecules having masses up to ten thousand amu, and also in collective states such
as SQUIDs and Bose-Einstein condensates. However, the principle does not seem to
hold for positions of large objects! Why for instance, a table is never found to be in
two places at the same time? One possible explanation for the absence of macroscopic
superpositions is that quantum theory is an approximation to a stochastic nonlinear
theory. This hypothesis may have its fundamental origins in gravitational physics, and
is being put to test by modern ongoing experiments on matter wave interferometry.
The Absence of Macroscopic Superpositions

In the year 1927, American physicists Clinton Davisson and Lester Germer performed
an experiment at Bell Labs, in which they scattered a beam of electrons off the surface
of a nickel plate. In doing so, they accidentally discovered that the scattered electrons
exhibited a diffraction pattern analogous to what is seen in the Bragg diffraction of
X-rays from a crystal. This experiment established the wave nature of electrons, and
confirmed de Broglies hypothesis of wave-particle duality. An electron can be in
This essay received the Fourth Prize in the FQXi Essay Contest, 2012.
A. Bassi (B)
Department of Physics, University of Trieste, Strada Costiera 11,
34151 Trieste, Italy
T. Singh
Tata Institute of Fundamental Research, Homi Bhabha Road,
Mumbai 400005, India
H. Ulbricht
School of Physics and Astronomy, University of Southampton,
Southampton SO17 1BJ, UK
151
152
A. Bassi et al.
more than one position at the same time, and these different position states obey the
principle of quantum linear superposition: the actual state of the electron is a linear
sum of all the different position states.
The principle of linear superposition is the central tenet of quantum theory, an
extremely successful theory for all observed microscopic phenomena. Along with
the uncertainty principle, it provides the basis for a mathematical formulation of
quantum theory, in which the dynamical evolution of a quantum system is described
by the Schrdinger equation for the wave function of the system.
The experimental verification of linear superposition for electrons heralded a
quest for a direct test of this principle for larger, composite particles and objects.
Conceptually, the idea of the necessary experimental set up is encapsulated in the
celebrated double slit interference experiment. A beam of identical particles is fired
at a screen having two tiny slits separated by a distance of the order of the de Broglie
wavelength of the particles, and an interference pattern is observed on a photographic
plate at the far end of the screen. Such an experiment was successfully carried out for
Helium ions, neutrons, atoms, and small molecules, establishing their wave nature
and the validity of linear superposition.
When one considers doing such an interference experiment for bigger objects
such as a beam of large molecules the technological challenges become enormous.
The opening of a slit should be larger than the physical size of the molecule (say
a few hundred nanometres) and the separation between slits should be smaller than
the coherence length (few micrometres) [1]. The first experiment in this class was
performed in 1999 in Vienna with C60 molecules (fullerene, having nearly 700 nucleons per molecule) and the observed interference, as predicted by quantum theory,
dramatically vindicated the validity of linear superposition for such large systems [2].
Today, a little over a decade later, experimentalists have succeeded in pushing this
limit to molecules with about 10,000 nucleons, and are aiming to push it higher to
particles with a million nucleons in the next few years. This is an extraordinary feat,
considering the great difficulties involved in maneuvering such large particles as they
travel from the source, through the slits, and on to the detector where the interference
pattern forms.
But will the principle of linear superposition continue to hold for larger and larger
objects? The answer is indeed yes, according to the modern outlook towards quantum
theory. The theory does not say that superposition should hold only in the microscopic
worldin fact, a molecule with ten thousand nucleons, for which the principle has
been confirmed, isnt exactly microscopic!
However when we look at the day to day world around us linear superposition
does not seem to hold! A table for instance is never found to be here and there
at the same time. In other words, superposition of position states does not seem to
hold for macroscopic objects. In fact already at the level of a dust grain, which we
can easily see with the bare eye, and which has some 1018 nucleons, the principle
breaks down. What could be going on in the experimentally untested desert between
104 nucleons, where linear superposition is valid, and 1018 nucleons, where it is
not valid?
10 Is Quantum Linear Superposition
153
The observed absence of macroscopic superpositions is the source of the quantum

measurement problem, as illustrated by the double slit experiment for a beam of electrons. The interference pattern which forms when an electron passes through both
the slits is destroyed when a detector (measuring apparatus) is placed close to and
behind one of the slits. The detector clicks (does not click) with a probability proportional to the square of the complex amplitude for the electron to pass through the
upper slit (lower slit). This is the so-called Born probability rule. The presence of
the detector breaks the superposition of the two states electron going through upper
slit and electron going through lower slit and the interference pattern is lost. One
says that the quantum state of the electron has collapsed, either to electron going
through upper slit only, or to electron going through lower slit only. The collapse
of the superposition of the two electron states (up and down), and the emergence of
probabilities contradict the Schrdinger equation, since this equation is deterministic
and linear. It predicts that a superposition should be preserved during the measurement process, that there should be no probabilities, and that position superpositions
should be observed for macroscopic objects as well, such as a measuring apparatus.
Why are superpositions of different position states for macroscopic objects not
observed, in contrast with what quantum theory predicts? Possible explanations
include reinterpretations of quantum theory such as the many worlds interpretation or consistent histories, and mathematical reformulations of the theory such as
Bohmian mechanics. By introducing one or more additional assumptions, these reinterpretations and reformulations ensure that macroscopic superpositions will not be
observed, while making the same experimental predictions as quantum theory.
Quantum Theory Is Approximate

Here we explore another possibility, which has been investigated extensively during
the last three decades or so [35]. What if quantum linear superposition is an approximate principle of nature? This means the following: consider an object (say an atom,
a molecule, or a table) which consists of N nucleons and is in a superposition of
two quantum states of an observable one is trying to measure (for example spin
or position). According to quantum theory the life time of such a superposition is
infinite! However, no experiment to date rules out the possibility that the lifetime
of such a superposition is finite, and is a monotonically decreasing function of N .
The dependence of on N should be such that for microscopic systems such as
atoms the superposition life time is astronomically large (longer than the age of the
Universe). Furthermore, for large objects such as a table or a macroscopic apparatus
the superposition life time should be so small that linear superposition is simply not
observed on the scale of current experiments.
Somewhere in between the microworld and the macroworld, i.e. between objects
of 104 and 1018 nucleons, the superposition lifetime would be neither too large, nor
too small, but just right enough to be detectable in the laboratory. In principle, this
could be achieved as follows. Suppose one has prepared in a controlled manner a
154
A. Bassi et al.
beam of very large identical molecules which are such that the superposition life time
between two different position states is . Let such a beam be passed through two slits
in a double slit experiment, thereby creating a superposition of two different position
states. Let the distance of the photographic plate from the slits be large enough that
the time of travel of a molecule from the slits to the plate far exceeds . Then the
superposition will exponentially decay before the molecule reaches the plate, and no
interference pattern will be seen. And this will happen even though no detector has
been placed behind either of the slits!
The requirements on a mathematically consistent generalization of quantum
theory in which superposition is an approximate principle are extremely stringent.
To begin with the theory should be nonlinear: superposition of two allowed quantum
states of the system should not be a stable allowed state. And yet for microsystems this
theory should be in excellent agreement with standard quantum theory so that results
of all known experiments are reproduced. The nonlinear process responsible for the
breakdown of superposition should be stochastic (i.e. random) in nature because the
outcome of a particular quantum measurement cannot be predicted. All the same,
the nonlinear process should be such that a collection of outcomes of repeated measurements obey the Born probability rule. Another reason the nonlinear effect should
be stochastic is that this helps avoid faster-than-light signalling. A deterministic (i.e.
non-stochastic) nonlinear quantum mechanics is known to allow for superluminal
communication. The nonlinear mechanism should be amplifying in nature, being
negligible for microscopic systems, but becoming more and more important for
large many particle systems, so that breakdown of superposition becomes more and
more effective.
It is a remarkable achievement that such a theory could be developed, and was
given shape in the eighties by physicists Ghirardi, Pearle, Rimini and Weber [6, 7].
This theory, which has come to be known as Continuous Spontaneous Localization
(CSL) has been extensively investigated in the following two decades, with regard
to its properties and solutions, constraints from laboratory and astrophysical observations, and direct tests of its predictions in the laboratory [5]. It is perhaps the only
well studied generalization of quantum theory whose experimental predictions differ
markedly from those of quantum theory in the macroworld and which hence provides
an excellent benchmark against which the accuracy of quantum theory can be tested.
In its original form (known as the GRW model), the CSL model is based on the
following two principles:
1. Given a system of n distinguishable particles, each particle experiences a sudden
spontaneous localization (i.e. collapse of position state) with a mean rate , to a
spatial region of extent rC .
2. In the time interval between two successive spontaneous localizations, the system
evolves according to the Schrdinger equation.
Thus two new fundamental constants of nature have been introduced, and these are
assumed to take definite numerical values, in order to successfully reproduce the
observed features of the microscopic and the macroscopic world. The constant ,
assumed to be of the order 1016 s1 , determines the rate of spontaneous localization
155
for a single particle. Amplification is achieved by showing that for a composite object
of n particles, the collapse rate is (n)1 s. The second constant rC is a length scale
assumed to be about 105 cm, and indicates that a widely spread quantum state
collapses to a size of about rC during localization.
In its modern version, the CSL model consists of coupling a randomly fluctuating
classical field with the particle number density operator of a quantum system, so
as to produce collapse towards spatially localized states. The collapse process is
continuous and dynamics is described by a single nonlinear stochastic differential
equation which contains both aspects: standard Schrdinger evolution as well the
nonlinear stochastic terms which result in collapse and breakdown of superposition.
The fundamental constants and rC continue to be retained. Today, experiments are
being devised to test CSL and are approaching the ballpark range where the preferred
values of and rC will be confirmed or ruled out.
The CSL dynamical equation can be used to show that macroscopic superpositions last for extremely short time intervals, thus explaining why such position
superpositions are not observed. Similarly the superposition lifetime for microstates
is extremely large. The same CSL equation elegantly solves the measurement problem. Suppose a microscopic quantum system is prepared in a superposition of two
quantum states, say A and B, of an observable which is to be measured. The CSL
equation for this system is essentially the same as the Schrdinger equation, the nonlinear effect being negligible, and evolution proceeds in the standard deterministic
linear fashion, preserving superposition of the states A and B. When the quantum
system comes into interaction with the measuring apparatus, which let us say is characterized by a pointer, one must now write down the CSL equation for the combined
system consisting of the quantum system and the measuring apparatus, and now the
nonlinear stochastic terms come into play. The state A would cause the pointer to be
at a position P1 and the state B would cause the pointer to be at a different position
P2 . The superposition of A and B would cause the pointer to be in superposition
of the macroscopically different states P1 and P2 . This however is not permitted by
CSL, since the pointer is macroscopic: the superposition very quickly decays to the
pointer state P1 or P2 . This is precisely what is interpreted as the collapse of the
superposed quantum state to either A or B. Furthermore, the CSL equation shows
that the outcome of an individual measurement is random, but repeated measurements are shown to result in outcomes P1 or P2 with relative frequency as given by
the Born probability rule [8].
Testing the Idea with Experiments

The nonlinear stochastic modification of the Schrdinger equation and the introduction of the constants and rC modify the predictions of quantum theory for various
standard results from known experiments and astrophysics. This allows bounds to be
put on these constants. These include bounds coming from decay of supercurrents in
SQUIDs, spontaneous X-ray emission from Germanium, absence of proton decay,
156
A. Bassi et al.
heating of the intergalactic medium, dissociation of cosmic hydrogen, heating of

interstellar dust grains, and temperature distortions in the cosmic microwave background. A novelty of the CSL mechanism is that the process of spontaneous localization produces a very tiny increase in the energy of the localized particle, thus violating
energy conservation. Some of the bounds on the fundamental constants come from
the lack of observation of any such violation. Using other arguments based on latent
image formation in photography [9] and formation of image in the human eye [10], it
has been suggested that the theoretical value of should be as high as 108 s1 . We
thus already see a fascinating debate taking place between theory and experiment:
the principle of quantum linear superposition is being confronted by experiment in
the true sense of the word.
However the most direct tests of CSL and linear superposition will come from
interference experiments for large objects. Unlike quantum theory, CSL predicts
that under certain suitable conditions an interference pattern will not be seen. Thus
an accurate experiment carried out under these conditions will definitely establish
whether quantum linear superposition is an exact or approximate principle. If interference is seen with a large molecule, it sets an upper bound on the value of . This
experimental field, known as matter wave interferometry, has made great strides in
recent years, and is one of the most important sources for testing proposed alternatives
to quantum theory, such as CSL [1, 1114].
The experiments involve overcoming technical challenges to prepare intense
beams of molecules in gas phase, to preserve spatial and temporal coherence of
the beam, and to detect the particles efficiently. Beam splitters for molecules are typically highly ordered periodic diffraction gratings of nanowires made from metal
or semiconductors, or they are standing light fields realized using the so-called
Kapitza-Dirac effect. The 1999 Vienna experiment with C60 fullerene used far-field
Fraunhofer molecular diffraction with a grating constant of 100 nm. For larger molecules, where it becomes imperative to effectively increase the beam intensity, more
promising results have been achieved through near field interference using the TalbotLau interferometer (TLI).
A TLI operating in the near-field (i.e. the spatial period of the gratings and the
interference pattern are on the same scale) was specifically invented to deal with
beams of low intensity and low collimation in interference experiments [15]. A three
grating TLI operates with weakly collimated beams: the first grating prepares beam
coherence and imprints a spatial structure on the beam. The second grating is thus
simultaneously illuminated by some 104 individual coherent molecular beams and
creates a self-image on a third grating, on which the interference pattern forms.
Effectively, the number of molecules contributing to the final interference pattern
is multiplied by the number of illuminating slits of the first grating, and all the
coherent beams from the 104 source slits are incoherently summed to contribute
to the same interference pattern. More recently, a modified version, known as the
Kapitza-Dirac TLI [16] has been employed, in which the secondthe diffraction
grating is replaced by an optical phase grating (Fig. 10.1). Molecules are diffracted
at periodic optical potentials due to the Kapitza-Dirac effect. The KDTLI has been
used to demonstrate interference with a molecule known as perfluoro-alkylated C60 ,
157
Fig. 10.1 Different

configurations of the
Talbot-Lau interferometer
are shown. a Three material
grating. b KapitzaDirac-Talbot-Lau
interferometer. c Optical
time-domain ionizing matter
interferometer (OTIMA).
Figure courtesy: Kinjalk
Lochan. Figure source [5]
having a mass 7,000 amu [17]. This is the largest molecule on which a successful
matter wave experiment has been carried out so far, and sets an upper bound of
105 s1 on the CSL parameter . This is only three orders of magnitude away from
the theoretical value 108 for predicted by Adler, and this latter value could be
confronted by an experiment with molecules having a mass of about 5,00,000 amu.
However, for molecules above 105 amu, their electromagnetic interactions with
the material gratings disable the interference pattern, and new technologies must
be sought to efficiently control and manipulate the center of mass motion of heavy
particles. Experiments are performed in ultra-high vacuum conditions to prevent
decoherence by collisions. Beams should be slow, so that the de Broglie wavelength
is not too low, they should be highly collimated and should have a high phase space
density. These features can be achieved through promising cooling techniques currently under development. Another important aspect is whether to use neutral or
charged particles. All large particle interference experiments thus far have been performed with neutralsthey have the advantage that they suffer lesser decohering
effects from interaction with the environment. On the other hand charged particles
are easier to manipulate and control, especially while preparing coherent particle
beams of heavy molecules [5].
158
A. Bassi et al.
A clever new proposal combines the best of both worlds: manipulation of charged
particles during the preparation of the beam, and interference after neutralization.
This novel three light grating TLI aims towards the interference of particles up to
109 amu and is known as the optical time-domain matter-wave (OTIMA) interferometer [18]. Charged particles will be provided by a mass filtered cluster aggregration
source. The charged clusters are neutralized at the first grating using light-matter
effects, diffracted at the second grating, and ionized again for detection at the third
grating.
An alternative approach to testing quantum superposition and CSL is optomechanics, which involves coupling micromechanical devices to light fields with which
they interact [19, 20]. A tiny mechanical object such as a micromirror is cooled to
extremely low temperatures and prepared to be, say, in a quantum superposition of
the vibrational ground state and first excited state. This superposed state is coupled
to a sensitive optical interferometer: if the superposed state decays in a certain time,
as predicted by CSL, the optical response of the interferometer will be different, as
compared to when the superposition does not decay (Fig. 10.2). Optomechanics aims
to test superposition for very massive particles in the range 106 1015 amu, but the
vibrational amplitude is very small compared to the desired amplitude, which is the
fundamental length scale rC of CSL. This makes it a challenge for optomechanics
to reach the expected regime where new physics is expected [2124].
However, very promising progress can be expected by combining techniques from
optomechanics and matter wave interferometry. Massive particles trapped in optical
Fig. 10.2 Optomechanics:

a Prototype of
optomechanically cooled
cantilever. Quantum optical
detection techniques enable
the sensitive read out of
vibrations as they couple to
light fields. b Mechanical
resonator interference in a
double slit (MERID). The
centre of mass motion of a
single optically trapped
nanoparticle is first cooled
and then superimposed by an
optical double potential. The
interference pattern evolves
in free fall after switching off
the trapping field. Figure
courtesy: Kinjalk Lochan.
Figure source [5]
159
traps are analogues of optomechanical systems, with the significant difference being
that by suitable choice of the trapping potentials, superposition of position states can
be created with a separation comparable to rC . After the superposed state has been
created the trapping potential is swiched off; the particles are allowed to fall freely,
and their spatial density distribution can be studied to verify the existence of an
interference pattern. Such an experiment can be carried out for very massive objects
(such as polystyrene beads of about 30 nm diameter and mass 106 amu). Known as
MERID (mechanical resonator interference in a double slit), it is a promising future
technique for testing the CSL model [25]. It can be expected that within the next two
decades matter wave interferometry and optomechanics experiments will reach the
ballpark of 106 109 amu where predictions of CSL will differ from quantum theory
sufficiently enough for them to be discriminated in the laboratory.
Why is Quantum Theory Approximate?

Continuous Spontaneous Localization has been proposed as a phenomenological
modification of quantum theory which is successful in explaining the observed
absence of macroscopic superpositions, and the Born rule. It is beyond doubt though
that underlying CSL there ought to be deep physical principles which compel us to
accept such a radical modification of quantum theory. Fascinating progress has been
taking place towards unravelling these underlying principles, mainly along three different directions, all of which point to an involvement of gravity, and a revision of
our understanding of spacetime structure at the deepest level.
It has been suggested independently by Karolyhazy et al. [26], Diosi [27] and
Penrose [28] that gravity is responsible for the absence of macroscopic superpositions. While their arguments differ somewhat at the starting point they all come to
approximately the same conclusion that the absence of superpositions will become
apparent around 106 109 amu, a range that agrees well with the prediction of CSL!
The key idea is that every object obeys the uncertainty principle and hence there is
an intrinsic minimal fluctuation in the spacetime geometry produced by it. When
the quantum state describing this object propagates in this fluctuating spacetime, it
loses spatial coherence beyond a certain length scale after a critical time. Such a
loss of spatial coherence is an indicator of breakdown of superposition of different
position states. This length and time scale is shown to be astronomically large for
microscopic particles, and very small for macroscopic objects, thus demonstrating
macroscopic localization. There is tantalizing evidence in the literature [29] that the
stochastic mechanism of CSL is provided by spacetime fluctuations, and that the
fundamental constants and rC derive from gravity. An ongoing optomechanics
experiment is specifically devoted to testing the role of gravity in causing breakdown
of superposition [20].
A second remarkable line of development, known as Trace Dynamics, has come
from the work of Stephen Adler and collaborators [30] who suggest that it is
unsatisfactory to arrive at a quantum theory by quantizing its very own limit,
160
A. Bassi et al.
namely classical dynamics. Instead, quantum theory is derived as an equilibrium

statistical thermodynamics of an underlying unitarily invariant classical theory of
matrix dynamics. It is fascinating that the consideration of Brownian motion fluctuations around the equilibrium theory provides a nonlinear stochastic modification of
the Schrdinger equation of the kind proposed in CSL. This is another clue that the
absence of macroscopic superpositions may have to do with quantum theory being
an approximation to a deeper theory.
Thirdly, the absence of macroscopic superpositions is possibly related to another
deep rooted but little appreciated incompleteness of quantum theory. In order to
describe dynamical evolution the theory depends on an external classical time, which
is part of a classical spacetime geometry. However, such a geometry is produced by
classical bodies, which are again a limiting case of quantum objects! This is yet
another sense in which the theory depends on its classical limit. There hence ought
to exist a reformulation of quantum theory which does not depend on a classical
time [31]. Such a reformulation has been developed by borrowing ideas from Trace
Dynamics and there is evidence that there exist stochastic fluctuations around such
a reformulated theory, which have a CSL type structure, and are responsible for the
emergence of a classical spacetime and breakdown of superposition in the macroscopic world [3234].
The CSL model, as it is known today, is nonrelativistic in character, which means
that it is a stochastic generalization of the nonrelativisitic Schrdinger equation. It is
intriguing that the model has thus far resisted attempts at a relativisitic generalization. On the other hand it is known that the collapse of the wave function during a
quantum measurement is instantaneous and non-local. This has been confirmed by
experimental verification of Bells inequalities. Thus it is certainly true that there is
a need for reconciliation between CSL induced localization, and the causal structure
of spacetime dictated by special relativity. In a remarkable recent paper [35] it has
been shown that if one does not assume a predefined global causal order, there are
multipartite quantum correlations which cannot be understood in terms of definite
causal order and spacetime may emerge from a more fundamental structure in a
quantum to classical transition.
For nearly a century the absence of macroscopic superpositions, in stark contradiction with what quantum theory predicts, has confounded physicists, and led
Schrdinger to formulate his famous cat paradox. The quantum measurement problem, a direct consequence of this contradiction, has been debated endlessly, and very
many solutions proposed, by physicists as well as philosophers. However up until
recent times the debate has remained largely theoretical, for no experiment has ever
challenged the phenomenal successes of quantum theory. Times have changed now.
There is a phenomenological model which proposes that quantum linear superposition is an approximate principle; there are serious underlying theoretical reasons
which suggest why this should be so, and most importantly, experiments and technology have now reached a stage where this new idea is being directly tested in
the laboratory. Perhaps after all it will be shown that the assumption that linear
superposition is exact is a wrong assumption. We will then have nothing short of a
161
revolution, which will have been thrust on us by experiments which disagree with
quantum mechanics, thus forcing a radical rethink of how we comprehend quantum
theory, and the structure of spacetime.
Acknowledgments This work is supported by a grant from the John Templeton Foundation.
Technical Endnotes (For Details See [5])

The Physics of Continuous Spontaneous Localization
The essential physics of the CSL model can be described by a simpler model, known
as QMUPL (Quantum Mechanics with Universal Position Localization), whose
dynamics is given by the following stochastic nonlinear Schrdinger equation

dt = H dt + (q qt )dWt (q qt )2 dt t ,

2
(10.1)
where q is the position operator of the particle, qt t |q|t is the quantum
expectation, and Wt is a standard Wiener process which encodes the stochastic effect.
Evidently, the stochastic term is nonlinear and also nonunitary. The collapse constant
sets the strength of the collapse mechanics, and it is chosen proportional to the mass
m of the particle according to the formula = mm0 0 , where m 0 is the nucleons mass
and 0 measures the collapse strength. If we take 0 102 m2 s1 the strength
of the collapse model corresponds to the CSL model in the appropriate limit.
The above dynamical equation can be used to prove position localization; consider
for simplicity a free particle (H = p 2 /2m) in the gaussian state (analysis can be
generalized to other cases):

t (x) = exp at (x x t )2 + ik t x + t .
(10.2)
By substituting this in the stochastic equation it can be proved that the spreads in
position and momentum
1
q (t)
2

1
;
atR
p (t)
(atR )2 + (atI )2
,
atR
(10.3)
do not increase indefinitely but reach asymptotic values given by

q () =

Kg
15
10
m,
m
m

p () =

Kg m
m
m
19
10
,
2
Kg
s
(10.4)
162
A. Bassi et al.
such that: q () p () = / 2 which corresponds to almostthe minimum

allowed by Heisenbergs uncertainty relations. Here, = 2 0 /m 0
105 s1 .
Evidently, the spread in position does not increase indefinitely, but stabilizes to
a finite value, which is a compromise between the Schrdingers dynamics, which
spreads the wave function out in space, and the collapse dynamics, which shrinks it in
space. For microscopic systems, this value is still relatively large (q ()1 m for an
electron, and 1 mm for a C60 molecule containing some 1,000 nucleons), such as to
guarantee that in all standard experimentsin particular, diffraction experiments
one observes interference effects. For macroscopic objects instead, the spread is
very small (q ()3 1014 m, for a 1 g object), so small that for all practical
purposes the wave function behaves like a point-like system. This is how localization
models are able to accommodate both the wavy nature of quantum systems and
the particle nature of classical objects, within one single dynamical framework.
The same stochastic differential equation solves the quantum measurement problem and explains the Born probability rule without any additional assumptions. For
illustration, consider a two state microscopic quantum system S described by the
initial state
(10.5)
c+ |+ + c |
interacting with a measuring apparatus A described by the position of a pointer
which is initially in a ready state 0 and which measures some observable O, say
spin, associated with the initial quantum state of S. As we have seen above, the
pointer being macroscopic (for definiteness assume its mass to be 1 g), is localized
in a gaussian state G , so that the initial composite state of the system and apparatus
is given by

(10.6)
0 = c+ |+ + c |+ G .
According to standard quantum theory, the interaction leads to the following type of
evolution:

c+ |+ + c | G
c+ |+ + + c | ,
(10.7)
where + and are the final pointer states corresponding to the system being in the
collapsed state |+ or | respectively. While quantum theory explains the transition
from the entangled state (10.7) to one of the collapsed alternatives by invoking a new
interpretation or reformulation, the same is achieved dynamically by the stochastic
nonlinear theory given by (10.1).
It can be proved from (10.1) that the initial state (10.6) evolves, at late times, to
t =
|+ + + t |

.
1 + 2t
(10.8)
163
The evolution of the stochastic quantity t is determined dynamically by the stochastic

equation: it either goes to t 1, with a probability |c+ |2 , or to t 1, with a
probability |c |2 . In the former case, one can say with great accuracy that the state
vector has collapsed to the definite outcome |+ + with a probability |c+ |2 .
Similarly, in the latter case one concludes that the state vector has collapsed to |
with a probability |c |2 . This is how collapse during a quantum measurement
is explained dynamically, and random outcomes over repeated measurements are
shown to occur in accordance with the Born probability rule. The time-scale over
which t reaches its asymptotic value and the collapse occurs can also be computed
dynamically. In the present example, for a pointer mass of 1 g, the collapse time turns
out to be about 104 s.
Lastly, we can understand how the modified stochastic dynamics causes the
outcome of a diffraction experiment in matter wave-interferometry to be different
from that in quantum theory. Starting from the fundamental Eq. (10.1) it can be
shown that the statistical operator t = E[|t t |] for a system of N identical
particles evolves as
2
t (x, y) = 0 (x, y)eN (xy) t/2 .
(10.9)
Experiments look for a decay in the density matrix by increasing the number of the
particles N in an object, by increasing the slit separation |x y|, and by increasing
the time of travel t from the grating to the collecting surface. The detection of an
interference pattern sets an upper bound on . The absence of an interference pattern
would confirm the theory and determine a specific value for (provided all sources
of noise such as decoherence are ruled out.)
A detailed review of the CSL model and its experimental tests and possible
underlying theories can be found in [5].
References
1. K. Hornberger, S. Gerlich, P. Haslinger, S. Nimmrichter, M. Arndt, Rev. Mod. Phys. 84, 157
(2011)
2. M. Arndt, O. Nairz, J. Vos-Andreae, C. Keller, G. Van der Zouw, A. Zeilinger, Nature 401, 680
(1999)
3. S.L. Adler, A. Bassi, Science 325, 275 (2009)
4. A. Bassi, G.C. Ghirardi, Phys. Rep. 379, 257 (2003)
5. A. Bassi, K. Lochan, S. Satin, T.P. Singh, H. Ulbricht, Rev. Mod. Phys. 85, 471 (2013)
6. G.C. Ghirardi, A. Rimini, T. Weber, Phys. Rev. D 34, 470 (1986)
7. G.C. Ghirardi, P. Pearle, A. Rimini, Phys. Rev. A 42, 78 (1990)
8. A. Bassi, D.G.M. Salvetti, J. Phys. A 40, 9859 (2007)
9. S.L. Adler, J. Phys. A 40, 2935 (2007)
10. A. Bassi, D.-A. Deckert, L. Ferialdi, Europhys. Lett. 92, 50006 (2010)
11. W. Feldmann, R. Tumulka, J. Phys. A: Math. Theor. 45, 065304 (2012)
12. S. Nimmrichter, K. Hornberger, P. Haslinger, M. Arndt, Phys. Rev. A 83, 043621 (2011)
13. O. Romero-Isart, Phys. Rev. A 84, 052121 (2011)
14. M. Arndt, A. Ekers, W. von Klitzing, H. Ulbricht, New J. Phys. 14 (2011)
15. J. Clauser, M. Reinsch, Appl. Phys. B: Lasers Opt. 54, 380 (1992)
164
A. Bassi et al.
16. K. Hornberger, S. Gerlich, H. Ulbricht, L. Hackermller, S. Nimmrichter, I. Goldt, O. Boltalina,

M. Arndt, New J. Phys. 11, 043032 (2009)
17. S. Gerlich, S. Eibenberger, M. Tomandl, S. Nimmrichter, K. Hornberger, P.J. Fagan, J. Txen,
M. Mayor, M. Arndt, Nat. Commun. 2, 263 (2011)
18. S. Nimmrichter, P. Haslinger, K. Hornberger, M. Arndt, New J. Phys. 13, 075002 (2011)
19. S. Bose, K. Jacobs, P. Knight, Phys. Rev. A 56, 4175 (1997)
20. W. Marshall, C. Simon, R. Penrose, D. Bouwmeester, Phys. Rev. Lett. 91, 130401 (2003)
21. T. Kippenberg, K. Vahala, Science 321, 1172 (2008)
22. M. Aspelmeyer, S. Groeblacher, K. Hammerer, N. Kiesel, J. Opt. Soc. Am. B 27, A189 (2010)
23. J. Chan, T. Alegre, A. Safavi-Naeini, J. Hill, A. Krause, S. Groeblacher, M. Aspelmeyer,
O. Painter, Nature 478, 89 (2011)
24. A. OConnell, M. Hofheinz, M. Ansmann, R. Bialczak, M. Lenander, E. Lucero, M. Neeley,
D. Sank, H. Wang, M. Weides et al., Nature 464, 697 (2010)
25. O. Romero-Isart, A. Panzer, F. Blaser, R. Kaltenbaek, N. Kiesel, M. Aspelmeyer, J. Cirac, Phys.
Rev. Lett. 107, 020405 (2011)
26. F. Karolyhazy, A. Frenkel, B. Lukcs, in Quantum Concepts in Space and Time, ed. by
R. Penrose, C.J. Isham (Clarendon, Oxford, 1986)
27. L. Disi, Phys. Lett. A 120, 377 (1987)
28. R. Penrose, Gen. Relativ. Gravit. 28, 581 (1996)
29. L. Disi, Phys. Rev. A 40, 1165 (1989)
30. S.L. Adler, Quantum Theory as an Emergent Phenomenon (Cambridge University Press,
Cambridge, 2004) pp. xii + 225
31. T.P. Singh, J. Phys. Conf. Ser. 174, 012024 (2009)
32. T. P. Singh, (2011). arXiv:1106.0911
33. K. Lochan, T.P. Singh, Phys. Lett. A 375, 3747 (2011)
34. K. Lochan, S. Satin, T.P. Singh, Found. Phys. 42, 1556 (2012)
35. O. Oreshkov, F. Costa, C. Brukner, Nat. Commun. 3, 1092 (2012)
Chapter 11
Quantum-Informational Principles
for Physics
Giacomo Mauro DAriano
Abstract It is time to to take a pause of reflection on the general foundations of

physics, for re-examining the logical solidity of the most basic principles, as the
relativity and the gravity-acceleration equivalence. The validity at the Planck scale
of such principles is under dispute. A constructive criticism engages us in seeking new
general principles, which reduce to the old ones only in the already explored domain
of energies. At the very basis of physics there are epistemological and operational
rules for the same formulability of the physical law and for the computability of its
theoretical predictions. Such rules give rise to new solid principles, leading us to a
quantum-information theoretic formulation, that hinges on the logical identification
of the experimental protocol with the quantum algorithm.
The information-theoretic program for physics foundations has already been
advocated in the past by several authors [1]. Recently the program succeeded in
deriving the full structure of quantum theory from informational principles [25],
and we will very briefly examine them here, as exemplars of good principles. The
problem is now to extend the informational program to relativistic quantum field
theory, the most fundamental theoretical structure of physics. The plan here proposed is to ground quantum field theory on two new principles pertaining only the
formulability and computability of the physical law: (1) the Deutsch-Church-Turing
principle, and (2) the topological homogeneity of interactions. As we will see, in
conjunction with the principles of quantum theory, these two new principles entail a
quantum cellular automata extension of quantum field theory.
The quantum automaton extends field theory in the sense that it includes localized
states and measurements, for whose description quantum field theory is largely inadequate. The quantum automaton doesnt suffer any formal violation of causality, e.g.
The following dissertation is a minimally updated version of the original essay presented at the
FQXi Essay Contest 2012 Questioning the foundations. A short summary of the follow-ups
and recent research results is given in the Postscriptum.
G.M. DAriano (B)
Dipartimento di Fisica dellUniversit degli Studi di Pavia, via Bassi 6, 27100 Pavia, Italy
G.M. DAriano
Istituto nazionale di Fisica Nucleare, Gruppo IV, Sezione di Pavia, Italy
Giacomo Mauro DAriano 2015
165
166
G.M. DAriano
superluminal tails of the probability distributions. It is not afflicted by any kind of

divergence, being exactly computable by principle. Relativistic covariance and other
symmetries are violated, but are recovered at the usual scale of energy.
The generality of the new principles does not deplete them of physical content. On
the contrary, the Dirac automatonthe most elementary theory of this kinddespite
its simplicity leads us to unexpected interesting predictions, e.g. it anticipates a
maximum mass for the Dirac particle just as a consequence of the unitarity of quantum
evolution, without invoking black-hole general-relativity arguments. It also opens
totally unexpected routes for redefining mechanical notions. As regards gravity, the
theory seems to suggest the route of the emergent thermodynamic force of JacobsonVerlinde [6, 7], here, specifically, as a purely quantum-digital effect of the Dirac
automaton.
Good and Bad Principles

Which kinds of principles are good and which are bad? We can limit ourself to four
main different types of principles: (1) dogmatic, (2) empirical, (3) simplifying (or
conventional), and (4) epistemological.
The dogma. This is definitely the worst case. Do we have dogmas in physics? We
have few subtle ones. It is not a blasphemy to regard the non existence of an absolute
reference frame as a dogma. What about the reference frame of the background
radiation? We indeed always invoke the frame of fixed stars for establishing if a
frame is inertial. The denial of the existence of an absolute frame is a relic of the
anthropocentrism repudiation that followed the Keplerian revolution. We will come
back to this dogma later.
The empirical principle. A principle is empirical if it has no logical motivation
other than its empirical evidence. A typical example is the Einsteins equivalence
principle: the identity between inertial and gravitational mass is an observed fact.
But do we have a good reason for it? The principle implies that the trajectory of a
mass in a gravitational field is independent on the mass, and this leads us to reinterpret
gravity as a property of spacethe starting point of general relativity theory, which
is then a re-interpretation of the principle, not a logical motivation. Another relevant
example of empirical principle is the invariance of the speed of light with the reference
framequite an odd one, isnt it? This lead Einstein to his first formulation of
special relativity. The principle was later recognized by Einstein himself to be only
an instance of the more general Galilei principle (the invariance of the physical
law with the reference frame) upon including the laws of electromagnetism: this
was definitely a great logical improvement. The empirical ones are good temporary
practical principles when we relinquish further explanation.
The simplifying principle. A simplifying principle is an unfalsifiable conventional assumption that abridges the formulation of the physical law. An example of such kind of principle is the assumption of homogeneity of time (it is
impossible to compare two different time-intervals in temporal sequence). But a
11 Quantum-Informational Principles for Physics
167
purported non-homogeneity of time would introduce only an unnecessary functional

time-parametrization in the physical law. Another example is the assumption that the
speed of light be isotropic in space. Reichenbach [8] correctly argued that in order
to determine simultaneity of distant events we need to know the light speed, but in
order to measure the light speed we need to establish simultaneity of two different
events for synchronizing clocks, and this leaded to a logical loop. What we can do?
Using a single clock we can only determine the two-way average speed of light on a
closed path. Reichenbach wrote indeed unconventional Lorentz transformations for
non-isotropic light speed, with the only result of introducing an additional anisotropy
parameter that is utterly irrelevant in practice. In conclusion: the simplifying principles are good ones, but we must keep in mind their conventional nature.
The epistemological principle. This is the most solid kind of principle: a principle that cannot be violated, even in-principle, because its violation would involve
contradicting the scientific method itself. Somebody would argue that claiming principles only of this kind would be equivalent to claiming an ultimate theory. This is
true. But should this be a good reason for not seeking principles of this kind, and for
evaluating their ultimate logical consequences? Clearly, to be a principle for physics
it cannot involve only pure logic: it must also incorporate the basic axiomatic of the
physical experiment. G. Ludwig has been a great advocate of such kind of principles
[9]. Einstein himself formulated special relativity in terms of precise protocols for
synchronizing clocks in order to establish coordinate systems. In the recent literature
operational axiomatic frameworks of this kind have emerged for quantum theory,
later converging to a unified framework [2, 1012]. The basic notionstests, events
and systemsmake the framework equivalent to a category theory for physics [13].
At the same time, it is also the skeleton axiomatization of a general information
theory, and, as mentioned, it ultimately leads to the informational axiomatization of
quantum theory [25]. A remarkable fact about the operational approach is that it
logically identifies the experimental protocol with the computer algorithm, providing
a stronger logical connection between theory and experiment.
The Relativity Principle

The relativity principle of Galilei and Einstein seems to possess a definite epistemological character, since it establishes the independence of the physical law from the
reference system, apparently a necessary requirement for the law formulation and
experimentation. The principle instead is based on the no-absolute dogma, and
nothing forbids defining the law within an absolute frame as long as we are able to
translate it to any other frame (which is what we actually do when we invoke the
fixed stars frame). This viewpoint may look as a sacrilege, it is the only logical
possibility for violations of Lorentz covariance.
168
G.M. DAriano
The Causality Principle

Causality has been always a taboo in physics. It is a principle underlying all modern
physics, and has been central to debates on the foundations of relativity and quantum
mechanics for over a century. Despite this, there is still a philosophical train of thought
arguing that the causality notion should be removed from physics. B. Russell was
one of the major advocates of this opinion [14].
On the other hand, causality is such a natural assumption that is often overlooked
as an axiom (see e.g. the first quantum axiomatization work of Hardy [10]). Instead,
it is the first of the informational axioms of quantum theory [2], also referred to as
no signaling from the future. In simple words it says: in a cascade of measurements
on the same system, the outcome probability of a measurement does not depend on
the choice of the measurement performed at the output. The principle also implies
no-signaling without interactionshortly no-signaling, and also commonly known
as Einstein causality. I should make now clear that, being causality an axiom of
quantum theory, any information purportedly originated in the future, as a time travel,
would logically constitute a falsification of the theory. For example, it would mean
to require nonlinearities in state evolution, or other variations of the theory.
As we will see later, in the present informational context special relativity emerges
as an approximate principle due to the joint implication of three principles: (1) the
causality principle, (2) the Deutsch-Church-Turing principle, and (3) the principle
of topological homogeneity of interactions.
The problem of physical causation is a huge topic in philosophy, and a thorough
discussion would take a thick volume. For the philosopher disbeliever I just want to
add that the reconciliation with the Humean position (that causality is just a human
way of looking at phenomena) passes through the probabilistic nature of the causal
link stated in the axiom, which involve the comparison between two probabilities: the
Humean viewpoint corresponds to the Bayesian interpretation of probability [15].
If causality cannot be proved, it can be falsified, as for any other scientific theory.
How? By considering any binary test that is granted to be deterministic, namely to
have zero probability for one outcome: if operating at the output of the test we can
make this same outcome to happen, then we can logically claim a signaling from the
future, given for granted the apparatus and its preparation.
Causal reasoning has always been a basic methodology in physics and in science
generally, but the romantic dream of a time travel keeps a sentiment against it alive.
Informational Principles for Quantum Theory

In addition to causality, there are five other informational principles that are needed
for deriving quantum theory [2]: (ii) local tomography, (iii) perfect distinguishability,
(iv) atomicity of composition, (v) ideal compressibility, and (vi) purification. All six
principles apart from (vi) hold for both classical and quantum information: only the
purification one singles out quantum theory.
169
The information-theoretical framework hinges around the notion of event, which

can occur probabilistically and has inputs and outputs systems. A complete collection of such events with overall unit probability is what is called testphysically a
measurement instrument. The systems are just the usual physical systems. Informationally, tests and events represent subroutines, whereas the systems are registers on
which information is read and written. Axiom (ii) (stating that joint states of multiple
systems can be discriminated by measurements on single systems) has become popular [16], since it reconciles the holism of quantum theory with the reductionism of the
experimental approach [17]. Axiom (iii) is crucial for hypothesis falsification, and
reconciles probabilism with logic. Axiom (iv) establishes that maximal knowledge
of two transformations implies maximal knowledge of their composition, a requirement that seems obvious indeed. The compression axiom (v) is the one that leads to
the notion of sub-systems (e.g. the qubit is a subsystem of the qutrit). It entails the
possibility of addressing separately the unknown from the perfectly known. Finally,
the purification postulate (vi) informally speaking is the principle of conservation
of information. In simple words it says that irreversibility and mixing can be always
regarded as the result of discarding an environment, otherwise everything is describable in terms of pure states and reversible transformations. Another informal way
of stating the principle is that ignorance about a part is always compatible with the
maximal knowledge about the whole.
The six principles for quantum theory have nothing of mechanical nature: what I
call quantum theory is just the theory of systems, i.e. the mathematical framework
of Hilbert spaces, algebra of observables, unitary transformations. It has no bearing on
the mechanics, namely particles, dynamics, quantization rules: for these the name
quantum mechanics would be more appropriate. Quantum mechanics, however,
is just a small portion of the more general quantum field theory, which itself is a
theory of systems: the quantum fields. The only mechanical elements remaining in
quantum field theory are the so-called quantization rules (or the path-integral) that
one may want to avoid in order to make the theory completely autonomous from the
classical theory, whereas, reversely, it should be classical mechanics to be derived as
an approximation of quantum field theory via a classicalization rule. But, how can
we formulate a field theory that is quantum ab initio? We need more informational
principles, in addition to the six ones of quantum theory. Those principles, which
will substitute the relativity principles, are: the Deutsch-Church-Turing principle,
and the principle of topological homogeneity.
Substitutes for the Relativity Principle

The Deutsch-Church-Turing principle. Rephrasing Deutsch [18]: Every physical
process describable in finite terms must be perfectly simulated by a quantum computer made with a finite number of qubits and a finite number of gates. In the logic
of specularity between experimental protocols and algorithms (both include also
outcomes), I would say: Every finite experimental protocol is perfectly simulated
170
G.M. DAriano
by a finite quantum algorithm. It is immediate to see that the principle implies two
sub-principles: (a) the density of information is finite, and (b) interactions are local.
The kind of information that we are considering here is quantum, whence the assertion that the density of information is finite means that the dimension of the Hilbert
space for any bounded portion of reality is finite. This means that e.g. there are no
Bosons, and the bosonic particle is only an asymptotic approximate notion. Richard
Feynman himself is reported to like the idea of finite information density, because
he felt that There might be something wrong with the old concept of continuous
functions. How could there possibly be an infinite amount of information in any
finite volume? [1]. The finite dimension of the Hilbert space also implies locality of
interactions, namely that the number of quantum systems connected to each gate is
finite.
Topological homogeneity of interactions. The principle states that the quantum
algorithm describing a physical law is a periodic quantum network. In the informational paradigm the physical law is represented by a finite set of connected quantum
gates, corresponding to a finite protocol, theoretically specular of a finite quantum
algorithm. Thus locality of interactions is required in order to define a physical law
in terms of a finite protocol under the local control of the experimenter, whereas
homogeneity represents the universality of the law, which is assumed to hold everywhere and ever. It follows that algorithmically the physical law is represented by a
quantum unitary cellular automaton [19]. The space-period and the time-period
of the automaton correspond to the minimum space and time units lP and tP the
Planck distance and the Planck time, respectively. At some very small scalethe
Planck scalethe world is discrete.
The Quantum Cellular Automaton

Causality together with the Deutsch-Church-Turing principle imply that information
propagates at finite speed, the maximum speed being the speed of light
c = l P /t P the causal speed of the automaton. The two principles together thus
imply that the state of any finite set of systems can be evaluated exactly as the evolution for of finite number of time-steps of a larger but still finite number of systems in
the past causal cone, regardless the quantum network being unbounded. We take as
vacuum state any state that is locally invariant under the automaton evolution. The
localized states are then those that differ from the vacuum, only for a finite number of systems. The future causal cone of these state-supporting systems is then the
place where only we need to evaluate the evolution, again with no need of boundary conditions. We do not have any divergence, nor ultraviolet (no continuum), nor
infrared (no calculation for infinite extension): the Deutsch-Church-Turing principle
excludes tout court the continuum and the infinite dimension.
Recovering the old quantum field theory. The old field theory is re-covered as
an approximation via an analytical asymptotic evaluation of the automaton evolution in the relativistic limit of small wave vectors and for delocalized states, which
171
correspond to the customary quantum particles. In this way one can both derive the
Dirac equation in the relativistic regime, but also describe the physics of very large
Planckian masses and in the ultra-relativistic regime of huge momenta [20].
Emerging physics. It must be stressed that the homogeneity of interactions is
a purely topological property, not a metrical one: to be near for systems means
just to be interacting, and the length of the graph links has no physical meaning.
Space-time metric emerges from the pure topology by event counting, and the Planck
length lP and time tP conceptually are only digital-analog conversion factors. Also
the particle mass m of the Dirac automaton is a pure number 0 m 1, and the
Planck mass mP is the conversion factor to customary kilograms.
Universal automata constants. The three quantities lP , tP , mP , are the irreducible
universal constants of the automata theory, and the adimensional mass is the only
free parameter of the Dirac automaton. The Planck constant can be rewritten in terms
of the automata universal constants as = mp lp 2 t p 1 .
Inertial mass. As I already explained in my previous FQXi essay [21, 22], the
inertial mass is reinterpreted as the slowing down of the information flow via the
coupling between the modes flowing along the directions in the network at maximal
speed c (for d > 1 space-dimensions is a coupling between different chiralities [23]).
Particle speed and Planck mass as bound on mass. The speed of a zero-mass
particle depends on the wave-length, and approaches zero at Planckian wavelengths
anisotropically in space (see Fig. 11.1). For massive particles the speed of light in
the Dirac equation decreases also versus the mass for very large Planckian masses,
the automaton evolution becoming stationary at the Planck mass [24], since for
larger masses the evolution would be non unitary. It follows that the particle mass is
mounded by the Planck mass, at which it behaves essentially as a mini black hole.
It is remarkable how these conclusions are reached without using general relativity,
just a result of quantum theory.
Energy and momenta are finite in the digital world. The maximum momentum
is the De Broglie relation /lP . We can have only one particle and one antiparticle
per Planck cell, and the bound on how much energy can be crammed into a unit of
Fig. 11.1 Group velocity vg
(normalized to c) for a
zero-mass particle
automaton versus the
adimensional momentum
(k x , k y ) (from Ref. [23]). The
speed is approximately
isotropic for low momentum
(relativistic regime), and
becomes anisotropic for very
large momenta (ultrarelativistic regime)
172
G.M. DAriano
space is determined by the maximum energy per particle, which cannot be more that
t p 1 = 6.14663109 J, a huge energy! This is the energy for achieving 2 ops [25]
of the automaton during the Planck time, as given by the Margulus-Levitin theorem
[26] (each step of the automaton is obtained with two rows of quantum gates).
A Quantum-Digital Space-Time
The quantum nature of the automaton is crucial for the emergence of space-time.
There are two main points against using a classical automaton.
First point against a classical automaton. With a classical automaton one cannot
have isotropic space emerging from an homogeneous classical causal network, due
to the Weyl Tile argument [27]: we count the same number of tiles in a square
lattice both along the diagonal and in the direction of the square sides: where the 2
comes from? Indeed, the maximal speed of information
in bcc-lattice automaton, as
in the Dirac case, would be faster by a factor 2 or 3 along diagonals than along
lattice axes, ending up with an anisotropic space for any homogeneous lattice [28],
(the problem is not cured by the continuum limit). Instead, in a quantum network
isotropy is recovered through quantum superpositions of different paths (see e.g.
Fig. 11.2c), and we have again isotropy of max-speed in the relativistic regime of
small momenta (Fig. 11.1), whereas anisotropy would be in principle visible only in
the ultra-relativistic regime of huge momenta (Figs. 11.1,11.2b) or for ultra-localized
states (Fig. 11.2d). In a similar manner the quantum nature of the network provides
the mechanism for restoration of all continuum symmetries in the relativistic regime.
The digital version of Lorentz transformations for a classical homogeneous causal
network can be found in Ref. [29]: the usual Lorentz covariance cannot be restored
from them. Recovering Lorentz covariance from a classical causal network (i.e.
100
100
50
50
0 0
0 0
50
50
100
100
Fig. 11.2 How particles would look in a digital world made by a quantum automaton: the Dirac
automaton for d = 2 space dimensions. The height of the plot is proportional to the absolute
amplitude of finding a particle with up-spin. Colors represents the spin state. The two figures depict
the evolved state after 60 steps of an initial state centered in the center of the plane. Left spin-up
localized state. Right Gaussian spin-up particle state, with x 2 = 2 y 2 = 8lp. (Theory in Ref. [23])
173
describing a causal ordering partial relation) conflicts with the homogeneity principle,
and needs a random topology, as in the causal set program of Sorkin.
Second point against a classical automaton. The second reason against classical
automata is that quantum superposition of localized states provides a mechanism for
directing information in space, in a continuum of directions, by superimposing localized states at neighboring locations with constant relative phase between them, thus
giving momentum to the information flow. Such mechanism is not possible in a
classical causal network with finite information density. It is the interplay between
quantum coherence and nonlocality that plays the crucial role of keeping information going along a desired direction with minimal spreading, a task that cannot be
accomplished by a classical automaton.
Emergence of classical mechanics. The Hamiltonian for the classical field theory corresponding to the quantum automaton can be reversely derived from the unitary operator of the automaton [21, 22]. Customary quantum particles are Gaussian
coherent superposition of single-system states with constant relative phase between
neighboring systems, corresponding to the particle momentum: the classical trajectory is the typical path along the quantum network, namely the path with maximum
probability of the Gaussian packet.
Where Is Gravity?
The big question is now where gravity comes from. I still dont have a definite
answer, but I believe that the equivalence principle must be rooted in the automaton
mechanism: the gravitational force must emerge at the level of the Dirac free theory,
which itself defines the inertial mass. This does not occur in customary quantum
field theory, but may happen in the quantum automaton theory, in terms of a tiny
thermodynamic effect that can occur even for few particles: a purely quantumdigital effect. Indeed, the digital nature of the quantum automaton seems to make
it the natural scenario for the generalized holographic principle at the basis of the
Jacobson-Verlinde idea of gravity as entropic force [6, 7]. The hypothesis of gravity as
a quantum-digital effect is very fascinating: it would mean we are indeed experiencing
the quantum-digital nature of the world, in everyday experience: through gravity!
Postscriptum
All predictions contained in this Essay has been later derived, and are now available
in technical papers. The reader should look at Ref. [23]. Other results can be found
in Ref. [20, 30, 31].
The main result is contained in manuscript [23], entitled Derivation of the
Dirac equation from informational principles. There it is proved the remarkable
result that from the only general assumptions of locality, homogeneity, isotropy,
174
G.M. DAriano
linearity and unitarity of the interaction network, only two quantum cellular automata
follow that have minimum dimension two, corresponding to a Fermi field. The two
automata are connected by CPT, manifesting the breaking of Lorentz covariance.
Both automata converge to the Weyl equation in the relativistic limit of small wavevectors, where Lorentz covariance is restored. Instead, in the ultra-relativistic limit
of large wave-vectors (i.e. at the Planck scale), in addition to the speed of light one
has extra invariants in terms of energy, momentum, and length scales. The resulting
distorted Lorentz covariance belongs to the class of the Doubly Special Relativity of
Amelino-Camelia/Smolin/Magueijo. Such theory predicts the phenomenon of relative locality, namely that also coincidence in space, not only in time, depends on
the reference frame. In terms of energy and momentum covariance is given by the
group of transformations that leave the automaton dispersion relations unchanged.
Via Fourier transform one recovers a space-time of quantum nature, with points in
superposition. All the above results about distorted Lorentz covariance are derived
in the new Ref. [32].
The Weyl QCA is the elementary building block for both the Dirac and the
Maxwell field. The latter is recovered in the form of the de Broglie neutrino theory of
the photon. The Fermionic fundamental nature of light follows from the minimality
of the field dimension, which leads to the Boson as an emergent notion [33].
The discrete framework of the theory allows to avoid all problems that plague
quantum field theory arising from the continuum, including the outstanding problem
of localization. Most relevant, the theory is quantum ab initio, with no need of
quantization rules.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Feynman and Computation (Westview Press, Boulder, 2002)

G. Chiribella, G.M. DAriano, P. Perinotti, Phys. Rev. A 84, 012311 (2011)
C. Brukner, Physics 4, 55 (2011)
L. Hardy, (2011). arXiv:1104.2066
L. Masanes, M.P. Mller, New J. Phys. 13, 063001 (2011)
T. Jacobson, Phys. Rev. Lett. 75, 1260 (1995)
E. Verlinde, J. High Energy Phys. 4, 029 (2011)
H. Reichenbach, The Philosophy of Space and Time (Dover, New York, 1958)
G. Ludwig, G. Thurler, A New Foundation of Physical Theories (Springer, New York, 2006)
L. Hardy, (2001). arXiv:quant-ph/0101012
H. Barnum, J. Barrett, L. Orloff Clark, M. Leifer, R. Spekkens, N. Stepanik, A. Wilce, R. Wilke,
New J. Phys. 12, 3024 (2010)
B. Dakic, C. Brukner, in Deep Beauty: Understanding the Quantum World Through Mathematical Innovation, ed. by H. Halvorson (Cambridge University Press, 2011)
B. Coecke, Contemp. Phys. 51, 59 (2010)
B. Russell, Proc. Aristot. Soc. 13, 1 (1912)
C.A. Fuchs, (2010). arXiv:1003.5209
L. Hardy, W.K. Wootters, Found. Phys. 42, 454 (2012)
G.M. DAriano, in Philosophy of Quantum Information and Entanglement, ed. by A. Bokulich,
G. Jaeger (Cambridge University Press, Cambridge, 2010)

18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
175
D. Deutsch, R. Soc. Lond. 400, 97 (1985)

B. Schumacher, R.F. Werner, (2004). arXiv:quant-ph/0405174
A. Bisio, G.M. DAriano, A. Tosini, (2012). arXiv:1212.2839
G.M. DAriano, FQXi Essay Contest Is Reality Digital or Analog? (2011)
G.M. DAriano, Il Nuovo Saggiatore 28, 13 (2012)
G.M. DAriano, P. Perinotti, Phys. Rev. A 90, 062106 (2014)
G.M. DAriano, Phys. Lett. A 376, 678 (2012). arXiv:1012.0756
S. Lloyd, (2012). arXiv:1206.6559
N. Margolus, L.B. Levitin, Phys. D 120, 188 (1998)
H. Weyl, Philosophy of Mathematics and Natural Sciences (Princeton University Press, Princeton, 1949)
T. Fritz, Discret. Math. 313, 1289 (2013)
G.M. DAriano, A. Tosini, Stud. Hist. Philos. Mod. Phys. 44
G.M. DAriano, F. Manessi, P. Perinotti, A. Tosini, (2014). arXiv:1403.2674
A. Bisio, G.M. DAriano, A. Tosini, Phys. Rev. A 88, 032301 (2013)
A. Bibeau-Delisle, A. Bisio, G.M. DAriano, P. Perinotti, A. Tosini, (2013). arXiv:1310.6760
A. Bisio, G.M. DAriano, P. Perinotti, arXiv:1407.6928
Chapter 12
The Universe Is Not a Computer

Ken Wharton
Abstract When we want to predict the future, we compute it from what we know
about the present. Specifically, we take a mathematical representation of observed
reality, plug it into some dynamical equations, and then map the time-evolved result
back to real-world predictions. But while this computational process can tell us what
we want to know, we have taken this procedure too literally, implicitly assuming that
the universe must compute itself in the same manner. Physical theories that do not
follow this computational framework are deemed illogical, right from the start. But
this anthropocentric assumption has steered our physical models into an impossible
corner, primarily because of quantum phenomena. Meanwhile, we have not been
exploring other models in which the universe is not so limited. In fact, some of these
alternate models already have a well-established importance, but are thought to be
mathematical tricks without physical significance. This essay argues that only by
dropping our assumption that the universe is a computer can we fully develop such
models, explain quantum phenomena, and understand the workings of our universe.
(This essay was awarded third prize in the 2012 FQXi essay contest; a new afterword
compares and contrasts this essay with Robert Spekkens first prize entry.)
Introduction: The Newtonian Schema

Isaac Newton taught us some powerful and useful mathematics, dubbed it the System
of the World, and ever since weve assumed that the universe actually runs according
to Newtons overall scheme. Even though the details have changed, we still basically
hold that the universe is a computational mechanism that takes some initial state
as an input and generates future states as an output. Or as Seth Lloyd says, Its a
scientific fact that the universe is a big computer [1].
Such a view is so pervasive that only recently has anyone bothered to give it a name:
Lee Smolin now calls this style of mathematics the Newtonian Schema [2]. Despite
K. Wharton (B)
Department of Physics and Astronomy, San Jos State University,
San Jos, CA 95192-0106, USA
177
178
K. Wharton
the classical-sounding title, this viewpoint is thought to encompass all of modern

physics, including quantum theory. This assumption that we live in a Newtonian
Schema Universe (NSU) is so strong that many physicists cant even articulate what
other type of universe might be conceptually possible.
When examined critically, the NSU assumption is exactly the sort of anthropocentric argument that physicists usually shy away from. Its basically the assumption
that the way we humans solve physics problems must be the way the universe actually operates. In the Newtonian Schema, we first map our knowledge of the physical
world onto some mathematical state, then use dynamical laws to transform that state
into a new state, and finally map the resulting (computed) state back onto the physical
world. This is useful mathematics, because it allows us humans to predict what we
dont know (the future), from what we do know (the past). But is it a good template for
guiding our most fundamental physical theories? Is the universe effectively a quantum computer? This essay argues no on both counts; we have erred by assuming
the universe must operate as some corporeal image of our calculations.
This is not to say there arent good arguments for the NSU. But it is the leastquestioned (and most fundamental) assumptions that have the greatest potential to
lead us astray. When quantum experiments have thrown us non-classical curveballs,
we have instinctively tried to find a different NSU to make sense of them. Thanks
to this deep bias, its possible that we have missed the bigger picture: the mounting
evidence that the fundamental rules that govern our universe cannot be expressed
in terms of the Newtonian Schema. Its evidence that weve so far found a way to
fold back into an NSU, but at a terrible costand without debate or recognition that
weve already developed the core framework of a promising alternative.
The next section will detail the problems that arise when one tries to fit quantum
phenomena into an NSU. The following sections will then outline the alternative to
the NSU and show how it naturally resolves these same problems. The conclusion
is that the best framework for our most fundamental theories is not the Newtonian
Schema, but a different approach that has been developed over hundreds of years,
with ever-growing importance to all branches of physics. It seems astounding that we
have not recognized this alternate mathematics as a valid Schema in its own right, but
no alternative makes sense if weve already accepted Lloyds fact that the universe
is a (quantum) computer. Only by recognizing that the NSU is indeed an assumption
can we undertake an objective search for the best description of our universe.
Challenges from the Quantum

Until the 20th century, the evidence against the NSU was circumstantial at best. One
minor issue was that (fundamental) classical laws can equally well be run forward
and backwardsay, to retrodict the historical locations of planets. So theres nothing in the laws to imply that the universe is a forward-running computer program,
calculating the future from some special initial input. Instead, every moment is just
as special as every other moment.
12 The Universe Is Not a Computer
179
Of course, the same is true for a deterministic and reversible computer algorithm
from the data at any time-step, one can deduce the data at all other time-steps.
Combined with a special feature of the Big Bang (its status as an ordered, lowentropy boundary condition), this concern mostly vanishes.1
But quantum phenomena raise three major challenges to the NSU. Standard
quantum theory deals with each of them in basically the same wayby assuming the NSU must be correct, and using suspiciously anthropocentric reasoning to
recast the universe in an image of our quantum calculations.
First, we have Heisenbergs uncertainty principle (HUP). In the classical context
of Heisenbergs original paper [3], this means we can never know the initial state
of the universe with enough precision to compute the future. This would not alone
have challenged the NSUa universal computer could potentially use the full initial
state, even if we did not know it. But it weakens the above argument about how the
Big Bang is special, because not even the Big Bang can beat the HUPas confirmed
by telltale structure in the cosmological microwave background. The special lowentropy order in the universes initial state is accompanied by random, non-special,
disorder.
But conventional quantum theory rejects the above reading of the HUP. In spirit
with the NSU, the unknown quantities are no longer even thought to exist. Note
the implication: if we humans cant possibly know something, then the universe
shouldnt know it either. The Big Bang is restored as the universes special input,
and the NSU is saved. But this step leads to new problemsnamely, we cant use
classical laws anymore, because we dont have enough initial data to solve them. To
maintain an NSU, were forced to drop down from classical second-order differential
equations to a simpler first-order differential equation (the Schrdinger equation).
This leads to the second major challengethe Schrdinger equation yields the
wrong output. Or more accurately, the future that it computes is not what we actually observe. Instead, it merely allows us to (further) compute the probabilities of
different possible outcomes. This is a huge blow to the NSU. Recall the three steps
for the Newtonian Schema: (1) Map the physical world onto a mathematical state,
(2) Mathematically evolve that state into a new state, and (3) Map the new state
back onto the physical world. If one insists on a universe that computes itself via
the Schrdinger equation, the only way to salvage the NSU is to have step 3 be a
probabilistic map. (Even though the inverse of that map, step 1, somehow remains
deterministic.)
Once again, since we are restricted from knowing the exact outcome, conventional
quantum theory puts the same restrictions on the NSU itself. In step 3, the story
goes, not even the universe knows which particular outcome will occur. And yet one
particular outcome does occur, at least when one looks. Even worse, the measurement
process blurs together steps 2 and 3, affecting the state of the universe itself in
a manner manifestly inconsistent with the Schrdinger equation. The question of
Although it does raise questions, such as why the laws happen to be time-symmetric, if the
boundary conditions are so time-asymmetric.
180
K. Wharton
exactly where (and how) the universe stops using the Schrdinger equation is the
infamous measurement problem of quantum theory. It becomes harder to think of
the universe as computing itself if the dynamical laws are not objectively defined.
So its perhaps unsurprising that many physicists imagine an NSU that ignores step
3 altogether; the universe is simply the computation of the ever-evolving Schrdinger
equation, the mismatch with reality notwithstanding. The only consistent way to deal
with this mismatch is to take the Everettian view that our entire experience is just
some small, subjective sliver of an ultimate objective realitya reality that we do
not experience [4].
Which brings us to the third challenge to the NSU: the dimensionality of the
quantum state itself. The phenomenon of quantum entanglementwhere the behaviors of distant particles are correlated in strikingly non-classical waysseems to
require a quantum state that does not fit into the spacetime we experience. The quantum state of a N-particle system formally lives in a configuration space of 3 N
dimensions. If the universe is the self-computation of such a state, then we live in
a universe of enormous dimensionality. Any consistent, NSU view of quantum theory (not merely the Everettians) must maintain that Einsteins carefully-constructed
spacetime is fundamentally incorrect. Instead, one must hold that Schrdinger accidentally stumbled onto the correct mathematical structure of the entire universe.
Of course, configuration space was not an invention of Schrdingers; it continues
to be used in statistical mechanics and other fields where one does not know the exact
state of the system in question. Poker probabilities, for example, are computed in
such a space. Only after the cards are turned face up does this configuration space of
possibilities collapse into one actual reality.
In the case of cards, though, its clear that the underlying reality was there all
alongconfiguration space is used because the players lack information. In the
case of a theory that underlies everything, thats not an option. Either the quantum state neglects some important hidden variables, or else reality is actually a
huge-dimensional space. Conventional thinking denies any hidden variables, and
therefore gives up on ordinary spacetime. Again, note the anthropocentrism: we use
configuration spaces to calculate entangled correlations, so the universe must be a
configuration space.2
The NSU becomes almost impossible to maintain in the face of all these
challenges. Treating the universe as a computer requires us to dramatically alter our
dynamical equations, expand reality to an uncountable number of invisible dimensions, and finesse a profound mismatch between the output of the equations and
what we actually observe.
Of course, no one is particularly happy with this state of affairs, and there are
many research programs that attempt to solve each of these problems. But almost
none of these programs are willing to throw out the deep NSU assumption that
may be at ultimate fault. This is all the more surprising given that there is a wellestablished alternative to the Newtonian Schema; a highly regarded mathematical
2
Like a poker player that denies any reality deeper than her own knowledge, imagining the facedown cards literally shifting identities as she gains more information.
181
framework that is in many ways superior. The barrier is that practically no one takes
this mathematics literally, as an option for how the universe might work. The next
sections will outline this alternative and reconsider the above challenges.
The Lagrangian Schema

While a first-year college physics course is almost entirely dominated by the
Newtonian Schema, some professors will include a brief mention of Fermats Principle of least time. Its a breathtakingly simple and powerful idea (and even pre-dates
Newtons Principia)it just doesnt happen to fit in with a typical engineeringphysics curriculum.
Fermats Principle is easy to state: Between any two points, light rays take the
quickest path. So, when a beam of light passes through different materials from point
X to point Y, the path taken will be the fastest possible path, as compared to all other
paths that go from X to Y. In this view, the reason light bends at an air/water interface
is not because of any algorithm-like chain of cause-and-effect, but rather because its
globally more efficient.
However elegant this story, its not aligned with the Newtonian Schema. Instead
of initial inputs (say, position and angle), Fermats principle requires logical inputs
that are both initial and final (the positions of X and Y). The initial angle is no longer
an input, its a logical output. And instead of states that evolve in time, Fermats
principle is a comparison of entire pathspaths that cannot evolve in time, as they
already cover the entire timespan in question.
This method of solving physics problems is not limited to light rays. In the 18th
century, Maupertuis, Euler, and Lagrange found ways to cast the rest of classical
physics in terms of a more general minimization3 principle. In general, the global
quantity to be minimized is not the time, but the action. Like Fermats Principle,
this so-called Lagrangian Mechanics lies firmly outside the Newtonian Schema.
And as such, it comprises an alternate way to do physicsfully deserving of the title
Lagrangian Schema.
Like the Newtonian Schema, the Lagrangian Schema is a mathematical technique
for solving physics problems. In both schemas, one first makes a mathematical representation of physical reality, mapping events onto parameters. On this count, the
Lagrangian Schema is much more forgiving; one can generally choose any convenient parameterization without changing the subsequent rules. And instead of a
state, the key mathematical object is a scalar called the Lagrangian (or in the case of
continuous classical fields, the Lagrangian density, L), a function of those parameters
and their local derivatives.
There are two steps needed to extract physics from L. First, one partially constrains
L on the boundary of some spacetime region (e.g., fixing X and Y in Fermats
Principle). For continuous fields, one fixes continuous field parameters. But only
3
Actually, extremization.
182
K. Wharton
the boundary parameters are fixed; the intermediate parameters and the boundary
derivatives all have many possible values at this stage.
The second step is to choose one of these possibilities (or assign them probabilistic
weights). This is done by summing the Lagrangian (densities) everywhere inside the
boundary to yield a single number, the action S. The classical solution is then found
by minimizing the action; the quantum story is different, but its still a rule that
involves S.
To summarize the Lagrangian Schema, one sets up a (reversible) two-way map
between physical events and mathematical parameters, partially constrains those
parameters on some spacetime boundary at both the beginning and the end, and then
uses a global rule to find the values of the unconstrained parameters. These calculated
parameters can then be mapped back to physical reality.
Newton Versus Lagrange

There are two fairly-widespread attitudes when it comes to the Lagrangian Schema.
The first is that the above mathematics is just thatmathematicswith no physical
significance. Yes, it may be beautiful, it may be powerful, but its not how our
universe really works. Its just a useful trick weve discovered. The second attitude,
often held along with the first, is that action minimization is provably equivalent to
the usual Newtonian Schema, so theres no point in trying to physically interpret the
Lagrangian Schema in the first place.
To some extent, these two attitudes are at odds with each other. If the two schemas
are equivalent, then a physical interpretation of one should map to the other. Still,
the arguments for schema-equivalence need to be more carefully dismantled. This
is easiest in the quantum domain, but its instructive to first consider a classical case,
such as Fermats Principle.
A typical argument for schema-equivalence is to use Fermats principle to derive
Snells law of refraction, the corresponding Newtonian-style law. In general, one can
show that action minimization always implies such dynamic laws. (In this context,
the laws are generally known as the Euler-Lagrange equations.) But a dynamical
law is not the whole Newtonian Schemaits merely step 2 of a three-step process.
And the input and output steps differ: Snells law takes different inputs than Fermats
Principle and yields an output (the final ray position) that was already constrained
in the action minimization. Deriving Newtonian results from a Lagrangian premise
therefore requires a bit of circular logic.
Another way to frame the issue is to take a known embodiment of the Newtonian
Schemaa computer algorithmand set it to work solving Lagrangian-style problems with initial and final constraints. The only robust algorithms for solving such
problems are iterative,4 with the computer testing multiple histories, running back
and forth in time. And this sort of algorithm doesnt sound like a universe that com4
As in the Gerchberg-Saxton algorithm [5].
183
putes itselfthe most obvious problem being the disconnect between algorithmic
time and actual time, not to mention the infinite iterations needed to get an exact
answer.
Still, conflating these two schemas in the classical domain where they have some
modest connection is missing the point: These are still two different ways to solve
problems. And when new problems come around, different schemas suggest different
approaches. Tackling every new problem in an NSU will therefore miss promising
alternatives. This is of particular concern in quantum theory, where the connection
between the two schemas gets even weaker. Notably, in the Feynman path integral
(FPI), the classical action is no longer minimized when calculating probabilities, so
its no longer valid to derive the Euler-Lagrange equations using classical arguments.5
So what should we make of the Lagrangian Schema formulations of quantum
theory? (Namely, the FPI and its relativistic extension, Lagrangian quantum field
theory, LQFT.) Feynman never found a physical interpretation of the FPI that didnt
involve negative probabilities, and LQFT is basically ignored when it comes to
interpretational questions. Instead, most physicists just show these approaches yield
the same results as the more-typical Newtonian Schema formulations, and turn to the
latter for interpretational questions. But this is making the same mistake, ignoring
the differences in the inputs and outputs of these two schemas. Its time to consider
another approach: looking to the Lagrangian Schema not as equivalent mathematics,
but as a different framework that can provide new insights.
Quantum Challenges in a Lagrangian Light

Previously we outlined outlined three challenges from quantum theory, and the high
cost of answering them in the NSU framework. But what do these challenges imply
for an LSU? How would the founders of quantum theory have met these challenges
if they thought the universe ran according to the mathematics of the Lagrangian
Schemanot as a computer, but rather as a global four-dimensional problem that
was solved all at once? Surprisingly, the quantum evidence hardly perturbs the
LSU view at all.
The first challenge was the uncertainty principle, but the classical LSU had this
built in from the start, because it never relied on complete initial data in the first
place. Indeed, for classical particles and fields, theres a perfect match between the
initial data one uses to constrain the Lagrangian and the amount of classical data one
is permitted under the HUP. In Fermats principle, if you know the initial light ray
position, the HUP says you cant know the initial angle.
Its only when one combines the quantum wave equations with the probabilistic Born rule that
FPI probabilities are recovered; see the discussion of Eq. (1) in [6].
184
K. Wharton
Curiously, this perfect match is only one-way. The HUP allows more ways to
specify the initial data than is seemingly permitted by the Lagrangian Schema. For
example, the HUP says that one can know the initial position or the angle of the light
ray, but Fermats principle only works with constrained initial positions.
But this is not a problem so much as a suggested research direction, evident only
to a Lagrangian mindset. Perhaps the HUP is telling us that weve been too restricted
in the way weve fixed the initial and final boundaries on classical Lagrangians. The
natural question becomes: What would happen if we required action-minimization
for any HUP-compatible set of initial and final constraints? For classical fields, the
answer turns out to be that such constraints must be roughly quantized, matching
equations that look like quantum theory [7].
Because the LSU doesnt need complete initial data to solve problems, theres
nothing wrong with the second-order differential equations of classical physics
(including general relativity, or GR). With this change, one can revive Heisenbergs
original interpretation of the HUP, yielding a natural set of initially-unknown hidden variables (such as the ray angles in Fermats Principle). In simple cases [8], at
least, these hidden variables can not only explain the probabilistic nature of the outcomes, but can actually be computed (in hindsight, after the final boundary becomes
known). Furthermore, theres no longer a compelling reason to drop to the first-order
Hamiltonian equations, the standard Newtonian Schema version of quantum theory.
And since its this leap from Lagrangian to Hamiltonian that introduces many of the
deepest problems for quantum gravity (the problem of time, etc.), there are good
reasons to avoid it if at all possible.
The Lagrangian Schema also provides a nice perspective on the second challenge:
the failure of Newtonian-style equations to yield specific, real-world outcomes (without further probabilistic manipulations). Recall this was the most brutal challenge
to the NSU itself, raising the still-unresolved measurement problem and breaking
the symmetry between the past and future. But the LSU doesnt utilize dynamical
equations, so it dodges this problem as well. The temporal outcome is not determined by an equation, its imposed as an input constraint on L. And because of
the time-symmetric way in which the constraints are imposed, theres no longer
any mathematical difference between the past and future; both constraints directly
map to the real world, without further manipulation. In fact, the Lagrangian procedure of fixing the future boundary looks remarkably like quantum measurements,
providing a new perspective on the measurement problem [9].
A common complaint at this point is that the above feature is a bug, in that it
somehow makes the Lagrangian Schema unable to make predictions. After all, what
we usually want to know is the outcome B given the input A, or at least the conditional
probability P(Bi |A) (the probability of some possible outcome Bi given A). But if
one particular outcome (say, B1 ) is itself an external constraint imposed on L, a
logical input rather than an output, then we cant solve the problem without knowing
the temporal outcome. Furthermore, since in this case B1 is 100 % certain, the other
possibilities (B2 , B3 , etc.) can never happen, contrary to quantum theory.
185
But like the NSU, this complaint conflates our useful calculations with objective
reality. In truth, any particular observed event does indeed have a single outcome,
with after-the-fact 100 % certainty. If we dont yet know that outcome, we can still
imagine fixing different outcome constraints Bi , and using L to compute an expected
joint probability P(A, Bi ) for each possibility. Its then a simple matter to normalize
subject to some particular initial condition A and generate the conditional probabilities P(Bi |A). These probabilities live in our heads until the actual outcome appears
and show us what has been the case all along, at which point we update our incomplete knowledge. This is basic Bayesian probability (see the above poker example),
and many have noted that it is a more natural interpretation of the standard quantum
collapse [10, 11].
Finally, consider the challenge of quantum entanglement. The problem with the
NSU mindset is that it demands an input state that can compute all possible outputs,
even if we dont know what type of measurement will eventually be made. In
N-particle systems, the number of possible future measurements goes up exponentially with N. Keeping track of *all* possible future measurements requires a state
that lives in an enormous configuration space. It simply doesnt fit in the universe
we observe, or in Einsteins GR.
But as weve seen, the NSU conflates the information we humans need to solve a
problem and the data that must actually correspond to reality. In any particular case,
a vast portion of this traditional quantum state turns out to be needlessit never
gets mapped to reality and is erased by the so-called collapse. Thats because
all possible measurements dont occur; only the actual measurement occurs. Once
the future measurement choice is known, the joint probabilities take on familiar
forms, with descriptions that have exact mathematical analogies to cases that do fit
in spacetime [6, 12].
Which brings us to the key point: If one wants to fit quantum theory into the
spacetime of GR, one must use the Lagrangian Schema, solving the problem all at
once. Only then can the solution take into account the actual future measurement
which, recall, is imposed as a boundary constraint on L. So an LSU-minded physicist,
when encountering entanglement, would have no reason to add new dimensions.
The spooky link between entangled particles would merely be joint correlations
enforced by virtue of both particles contributing to the same global action [12].
When viewed from a Lagrangian Schema mindset, the transition from classical to
quantum phenomena is not only less jarring, but is arguably a natural extension. Sure,
some things have to changeperhaps extending the principle of action minimization
[7]but theyre changes that only make sense in an LSU, with no NSU translation.
Classical physics provided a few cases where the two Schemas seemed to almost
overlap, perhaps lulling us into a feeling that these two approaches must always
overlap. But the fact that quantum phenomena are so incomprehensible in an NSU,
and more natural in an LSU, should make us consider whether weve been using a
deeply flawed assumption all along.
186
K. Wharton
Conclusions: Our Lagrangian Universe

The best reasons for taking the Lagrangian Schema seriously lie in quantum theory,
but there are other reasons as well. Its the cleanest formulation of general relativity,
with the automatic parameter-independence that GR requires, and bypasses problematic questions such as how much initial data one needs to solve the Newtonian-style
version. The LSU blends time and space together just like GR, while the NSU has
to grapple with a dynamic evolution that seems to single out time as special. The
standard model of particle physics is not a set of dynamic equations, but is instead
a Lagrangian density, with deep and important symmetries that are only evident in
such a framework. Even NSU-based cosmological mysteries, such as why causallydisconnected regions of the universe are so similar, no longer seem as problematic
when viewed in an LSU light.
But from the computational perspective of the NSU, any description of an LSU
seems baffling and unphysical. When trying to make sense of the LSU, a NSU-minded
physicist might ask a number of seemingly-tough questions. Which past events cause
the future boundary constraint? How do objects in the universe know what future
boundary theyre supposed to meet? Doesnt Bells Theorem [13] prove that quantum
correlations cant be caused by past hidden variables? A close look reveals these
questions are already biasedthey all implicitly assume that we live in an NSU. But
without the mentality that the past causes the future by some algorithmic process,
the above questions are no longer well-posed.
Constructing a complete theory built upon the Lagrangian Schema is a vast project,
one that has barely even begun. The necessary first step, though, is to recognize that
the NSU is an assumption, not a statement of fact. Even then, it will be difficult to put
such a deep bias behind us completely, to distinguish our most successful calculations
from our most fundamental physical models. But it also wasnt easy to fight other
anthropocentric tendencies, and yet the Earth isnt the center of the universe, our sun
is just one of many, there is no preferred frame of reference. Now theres one last
anthropocentric attitude that needs to go, the idea that the computations we perform
are the same computations performed by the universe, the idea that the universe is
as in the dark about the future as we are ourselves.
Laying this attitude to one side, at least temporarily, opens up a beautiful theoretical vista. We can examine models that have no Newtonian Schema representation, and yet nicely incorporate quantum phenomena into our best understanding of
spacetime. We can treat the universe as a global, four-dimensional boundary-value
problem, where each subset of the universe can be solved in exactly the same manner, with exactly the same rules. Stories can be told about what happens between
quantum measurements, and those very measurements can be enfolded in a bigger
region, to simultaneously tell a bigger story. And most importantly, such models
will suggest further models, with alterations that only make sense in a Lagrangian
frameworkperhaps a local constraint like L = 0, or treating the Euler-Lagrange
equations as just an approximation to a fundamentally underdetermined problem.
187
It is these models, the balance of the evidence suggests, that have a chance of
representing how our universe really works. Not as we humans solve problems, not
as a computer, but as something far grander.
Afterword: On Spekkens Winning Essay

The Universe is not a Computer was awarded third prize in the 2012 FQXi Essay
Contest, ranked behind other excellent essaysmost notably the first prize winner [14], written by the Perimeter Institute physicist Robert Spekkens. Like the above
essay, Spekkens zeroed in on how most physical theories are framed using a statesplus-dynamics Newtonian Schema, although with a different focus and conclusion.
With these essays now being presented in the same volume, this afterword is an
opportunity to compare and constrast these two viewpoints.
Spekkens essay begins by noting that physical theories generally are divided into
dynamics (laws that implement time evolution) and kinematics (the space of
physical states permitted by a theory). These, of course, are the key components of
any theory that falls under the Newtonian Schema, described above. After noting
how very different theories are framed in precisely this manner, Spekkens essay
states that physicists typically agree that any proposal must be described in these
terms.
Both of our essays are in general agreement that it is this framing of physical
theories, in terms of dynamics + kinematics, that contains a widespread mistaken
assumptionbut we are in disagreement as to the precise nature of the mistake.
Spekkens essay makes the excellent point that seemingly-different theories (which
postulate different kinematics and dynamics) can in fact be empirically indistinguishable when these two components are taken together. I agree with Spekkens that two
such theories should not be viewed as competitive explanations but rather as essentially identical. (One possible lesson for theorists might be that they are proposing
too many different theories, and can focus on a bare few.)
On the other hand, the above essay argues that theorists have been far too conservative in postulating different theories, in that they almost exclusively are couched in
the Newtonian Schema. One can make a case that framing theories in terms of kinematics + dynamics is more an instinctive habit than a well-thought-out agreement,
and that new Lagrangian-schema approaches are needed. The mistake, in this view,
is that the kinematics + dynamics framework is too restrictive, not too permissive.
These different conclusions are not mutually exclusive. Take Spekkens example
of how classical physics can be couched in terms of forces and Newtons laws (on
one hand) and Hamiltonian dynamics (on the other). These two theories are indeed
empirically indistinguishable, and should not be thought of as essentially different.
But they also both fall under the Newtonian Schema. It is notable that classical
Lagrangian mechanics does not specify any dynamics, and therefore lies in a different
category of theory altogether (a category unaddresed in Spekkens essay). In this
188
K. Wharton
sense, our essays are both making the case that Newtonian Schema theories are more
similar than they might appear, and the above essay is making the additional case
that Lagrangian Schema theories are different and under-explored.
One counterpoint to this claim might be to note that Lagrangian Schema theories
can also be expressed in terms of dynamics + kinematics; namely, there are no
dynamical laws, and the allowed kinematical states are merely four-dimensional
histories that obey certain restrictions. In other words, Lagrangian Schema theories
are all kinematics, no dynamics.
Might Spekkens claim for empirical indistinguishability perhaps be extended
to apply to Lagrangian Schema theories as well, showing them to all be essentially equivalent to a class of Newtonian Schema counterparts? After all, classical
Lagrangian mechanics is empirically equivalent to Newtonian mechanics (if leaving
aside the input/output differences discussed above), and the probabilities generated
by the Feynman path integral are empirically equivalent to the probabilities generated
by the combination of the Schrdinger equation and the Born Rule [6]. Combined
with the many inter-Newtonian-Schema examples in Spekkens essay, this may make
it seem like such an argument might be successfully developed.
But the essential differences between three-dimensional states governed by
dynamics and four-dimensional histories with no dynamics is far more dramatic
than these examples imply. Indeed, counter-examples have recently been published
[15, 16] demonstrating simple Lagrangian Schema toy models with no dynamical
counterpart whatsoever. And far from being some unimportant curiosity, it is this
precise style of model that most naturally maps to the very quantum phenomena that
defy Newtonian Schema explanations.
For example, consider the discussion concerning kinematical- and dynamicallocality in Spekkens essay. There, the point was that since fully-local Newtonian
Schema accounts run afoul of the Bell inequalities, trying to rescue kinematical locality was essentially impossible: Any such theory would necessarily have dynamical
nonlocality, and would therefore always be empirically indistinguishable from a theory with kinematical nonlocality. But in the case of the Lagrangian Schema, there
is no dynamics, local, nonlocal, or otherwise. The promise of rescuing kinematical
locality (as discussed in Sect. Quantum Challenges in a Lagrangian Light) is now
far more than just an empty redefinition of termsindeed, it is one of the primary
motivations for pursuing Lagrangian Schema explanations in the first place.
So despite my general agreement with almost everything in Spekkens winning
essay, that essay is still framed in the Newtonian Schema mindset that is arguably
a deep and mistaken assumption in its own right. The claim in Spekkens abstract
that A change to the kinematics of a theory... can be compensated by a change
to its dynamics without empirical consequence is not always true when there are
no dynamics in the original theory (as per the counter-examples in [15, 16]). Still,
since it does appear that this claim is true for Newtonian Schema theories, Spekkens
essay will hopefully help to focus the debate where it is needed: not between empirically indistinguishable Newtonian Schema explanations of quantum phenomena, but
rather between dynamical and all at once explanatory accounts of our universe.
189
References
1. S. Lloyd, http://www.edge.org/3rd_culture/lloyd06/lloyd06_index.html. Accessed 16 Aug
2012
2. L. Smolin, Phys. World, p. 21 (2009)
3. W. Heisenberg, Zeitschrift fur Physik 43, 172 (1927)
4. H. Everett, Rev. Mod. Phys. 29, 454 (1957)
5. R.W. Gerchberg, W.O. Saxton, Optik 35, 237 (1972)
6. K.B. Wharton, D.J. Miller, H. Price, Symmetry 3, 524 (2011)
7. K.B. Wharton, arXiv:0906.5409
8. K.B. Wharton, Found. Phys. 40, 313 (2010)
9. K. Wharton, arXiv:1106.1254
10. R.W. Spekkens, Phys. Rev. A 75, 32110 (2007)
11. N. Harrigan, R.W. Spekkens, Found. Phys. 40, 125 (2010)
12. P.W. Evans, H. Price, K.B. Wharton, Brit. J. Found. Sci. (2012). doi:10.1093/bjps/axr052
13. J.S. Bell, Rev. Mod. Phys. 38, 447 (1966)
14. R.W. Spekkens, arXiv:1209.0023
15. H. Price, Stud. Hist. Philos. Mod. Phys. 39, 752 (2008)
16. K. Wharton, Information 5, 190 (2014)
Chapter 13
Against Spacetime
Giovanni Amelino-Camelia
Abstract The notion of location physics really needs is exclusively the one of
detection at a given detector and the time for each such detection is most primitively
assessed as the readout of some specific material clock. The redundant abstraction
of a macroscopic spacetime organizing all our particle detections is unproblematic
and extremely useful in the classical-mechanics regime. But I here observe that in
some of the contexts where quantum mechanics is most significant, such as quantum
tunneling through a barrier, the spacetime abstraction proves to be cumbersome.
And I argue that in quantum-gravity research we might limit our opportunities for
discovery if we insist on the availability of a spacetime picture.
A 19th-century Alice could have asked how do we know we all share the same
time? To my knowledge nobody asked the question then. And if some Alice did
ask, Bobs reaction surely would have been going something like what a stupid
question! of course we all share the same time!
We now know that this question is meaningful and actually the answer is no: we
established experimentally that observers in relative motion do not share the same
time.
A similar example of apparently stupid 19th-century question can be structured
around double-slit experiments. And sometimes the question we are not asking is
about the meaningfulness of a notion we are taking for granted: particularly in the
second half of the 19th century we were very busy attempting to establish the properties of the ether, but we then figured out that there is no place in physics for any
property of the ether.
To me the most defining mission of science is to establish which of the questions
we are not even asking (implicitly assuming their answer is evidently yes) is
actually a meaningful question whose answer is no. And these notes are about my
latest speculation of this sort: in modern physics, and particularly in quantum-gravity
research, we are very busy trying to establish the fundamental properties of spacetime,
and I start to wonder whether, after all, there might be no properties of spacetime to
worry about.
G. Amelino-Camelia (B)
Dipartimento di Fisica, Sapienza Universit di Roma and INFN Sez. Roma1,
P.le A. Moro 2, 00185 Rome, Italy
191
192
G. Amelino-Camelia
I am aware of the fact that this speculation is very audacious. Spacetime appears
toplay a robust reliable role in our conceptualization of physics. This robustness is
something we appreciate from very early on in life, since our resident devices (eyes,
ears, nose..., of course particle detectors) inform us of the reliability of our spacetime
inferences. And this reliability so far extends all the way to the most powerful devices
we invented, including our best particle colliders (see Fig. 13.1). But in some sense
we never actually see spacetime, we never detect spacetime, and over the last
century we gradually came to appreciate that there would be no spacetime without
particles. And if we contemplate the possibility of reducing our description of Nature
to its most primitive building blocks, the most minimalistic description of physics
that is afforded to us, then it becomes evident that the notion of location physics
really needs is exclusively the one of detection at a given detector and that the time
for each such detection is most primitively assessed as the readout of some specific
material clock at the detector. At least in this sense, the abstraction/inference of a
macroscopic spacetime organizing all such timed detections is redundant. We could
(and perhaps should) build all of our description of physics, including the so-called
spacetime observables, using as primitive/most-fundamental notions the ones of
material detectors and clocks.
Of course, the spacetime abstraction is unrenounceably convenient for organizing and streamlining our description of observations done in the classical-mechanics
regime. But I here highlight some aspects of quantum mechanics, such as tunneling
through a barrier, for which the spacetime abstraction proves to be cumbersome. And
Fig. 13.1 Image courtesy of the ATLAS collaboration; an aspect of Event 71902630, Run 204769
at ATLAS-LHC, a 4-muon event relevant for the Higgs search. The most primitive ingredients for
such analyses are of course timed particle detections, each one at a specific physical detector among
the very many that compose ATLAS. But key for the success of such experiments is the fact that it
happens to be the case that these observed localizations at detectors all combine to allow us to infer
a location for the interaction vertex. It is indeed crucial for the analysis of this event to establish
whether the four muon tracks (thick red lines) do originate from the same vertex, and this is not
a trivial task, especially because of pile up. [Each time two bunches of protons cross at ATLAS
several proton-proton collisions occur (shown in figure as dots where particle tracks meet) within
a region whose size is of only a few centimeters.]
13 Against Spacetime
193
I argue that, as we try to get ready for going even beyond quantum mechanics, in the
context of quantum-gravity research, we must contemplate even more virulent departures from the spacetime paradigm. So my thesis is that as we get to know Nature
better our insistence on the consistency of the abstraction of a macroscopic spacetime
picture may be gradually transforming from a source of valuable inspiration into an
obstruction to knowledge.
Reliability of the Spacetime Abstraction

in Classical Mechanics
Emission-Detection Setups and Wasteful Redundancies
It is useful for my purposes to start with some comments on emission-detectiontype measurements in a background Minkowski spacetime. I consider observers
Alice and Bob, which are (emitters/) detectors, equipped with clocks. Alice and Bob
establish that they are in relative rest at a distance L with synchronized clocks, by
using standard Einsteinian procedures relying on the time of flight between them
of some reference particles (in practice they would use low-energy photons). Once
this is done, Alice might write the following equation to describe the trajectory in
spacetime of a certain specific particle:
x = vt
(13.1)
Of the infinitely many potential truths codified in this equation only two facts are
established experimentally: the particle is emitted at Alice at time t = 0 of Alices
clock and the particle is detected at Bob at time t = L/v of Bobs clock.
The redundancy of the spacetime abstraction for emission-detection-type measurements is particularly clear in relativistic theories, where it plays a role in the
relationship between active and passive relativistic transformations. There the redundancy is in characterizations such as the particle was at a distance of 10 m from Alice,
where Bob is, which evidently can be fully replaced by the particle was at Bob.
Bubble Chambers and Convenient Redundancies

If all we had were emission-detection-type measurements the abstraction of a
spacetime would have probably never been made, since in those measurements it
is wastefully redundant. But our spacetime measurements are not all of emissiondetection type. Many of them are such that several particles are detected all attributable to a single spacetime event, and in such cases the spacetime abstraction is very
valuable. I already showed a striking example of this sort in Fig. 13.1. And I should
194
G. Amelino-Camelia
stress how this is important in astrophysics: when a star bursts we detect bunches
of particles (mainly photons), and the discussion of the physics content of these
measurements could in principle be limited to those timed particle detections. But it
is evidently advantageous to recognize that in these instances the sequence of particle
detections can be organized to infer a localized explosion far away in spacetime.
And the spacetime abstraction acquires an added element of tangibility when we
perform sequences of measurements of localization of the same particle (or body). As
representative of this vast class of measurements let me consider the case of bubblechamber measurements. Of course also in a bubble-chamber setup the primitive
measurements are timed particle detections, typically photographic-camera detections of photons. That collection of photons however proves to be reliably describable
in terms of inferences for the positions of certain bubbles, and in turn the collection
of positions of bubbles allows us to build an inference of trajectory for a charged
microscopic particle. Evidently here too the spacetime abstraction is redundant, but
the advantages of adopting the spacetime-trajectory inference in such cases are undeniable (see Fig. 13.2).
Fig. 13.2 A cartoonist impression of the type of information we gather with bubble-chamber
experiments (the choice of specific example is based on an image shown in The Particle Odyssey: A
Journey to the Heart of the Matter by F. Close, M. Marten and C. Sutton). The color-coded particle
trajectories show an antiproton (grey) colliding with a proton at rest (not visible), and thereby
producing four + (positive pions; red) and four (negative pions; green). One of the + decays
into a muon (yellow) and a neutrino (not visible). A magnetic field in the chamber causes the
trajectories of positively-charged particles to bend in direction opposite to the one for negativelycharged particles. All this information is of course also coded in the primitive measurements
here relevant, which are photographic-camera detections of photons from the bubbles. But the
description in terms of that huge number of photon detections at the photographic camera is far less
advantageous than the streamlined description in terms of spacetime trajectories of a few charged
particles
195
Spacetime and the Ether

I have so far only highlighted some aspects of the redundancy of our spacetime
inferences.
What is then spacetime?
Does spacetime exist?
I shall leave these questions to the appetites of philosophers. Physics can confine
their discussion to two simple observations:
(i) the fact that we can add reference to a spacetime without adding any new
item to our list of elementary/primitive measurement procedures (still only timed
particle detections) is a complete proof of the redundancy of spacetime in science;
(ii) but, while awareness of its redundancy may at some point be valuable, the
abstracted notion of spacetime is tangibly useful, and as long as this is the
situation there would be no reason for us to change the way we use this notion.
It will be clear from these notes that I am speculating about the possibility that the
status of spacetime in our current theories might resemble the status of the ether at the
beginning of the 20th century. This was nicely summarized by Poincar [1]: Whether
the ether exists or not matters littlelet us leave that to the metaphysicians; what is
essential for us is that everything happens as if it existed, and that this hypothesis is
found to be suitable for the explanation of phenomena. After all have we any other
reason for believing in the existence of material objects? That too is only a convenient
hypothesis; only it will never cease to be so, while some day, no doubt, the ether will
be thrown aside as useless.
Spacetime in Quantum Mechanics

Is our insistence on the availability of consistent spacetime inferences always useful,
or at least harmless?
Interesting challenges for the spacetime abstraction are already found within
quantum mechanics, even though it may appear that one can list very many quantummechanics applications where the spacetime abstraction proves to be reliable. In most
cases quantum mechanics is tested in experimental contexts where (because of the
physical scales involved) most of the analysis is still conducted within classical
mechanics, with only a few agents in the measurement procedure needing quantummechanical treatment. It is therefore legitimate to suspect that much of the success
of our spacetime inferences in standard applications of quantum mechanics might
be simply inherited from classical mechanics. And there are hints that this might
be the case. I have a long-term project attempting to substantiate rather broadly the
limitations that the spacetime abstraction faces within quantum mechanics [2]. For
the purposes of these notes it is sufficient for me to comment, in this section, on two
of these examples of open challenges for our understanding of quantum mechanics
which may suggest that the spacetime abstraction is already turning from useful to
cumbersome.
196
G. Amelino-Camelia
Troubles with Special-Relativistic Position Observable

Let me start by looking back at Sect. Emission-Detection Setups and Wasteful
Redundancies, where I commented on the redundancy of the spacetime abstraction for certain emission-detection-type measurements. Now let us imagine that the
particle sent out by Alice reaches Bob in a case where quantum mechanics is very
significant. For example we could consider a suitable double slit between Alice
and Bob. Famously the spacetime description must be creatively adapted to such contexts. At least within the Galilean-relativistic version of quantum mechanics there
still is a legitimate place for the spacetime abstraction: we cannot infer anymore a
definite spacetime trajectory for a particle but we have a position observable X (t)
which encodes in probabilistic manner the information on the spacetime trajectory
of the particle. However, in spite of a stubborn effort by the community lasting for
some 80 years, we have not found a suitable position observable generalizing to the
special-relativistic theory this luxury of the Galilean limit. Several attempts have
been made primarily inspired by early work by Newton and Wigner [3], but the
outcome remains largely unsatisfactory [4, 5].
Our understanding of the origin of this challenge has improved significantly over
the last decade through work [69] establishing a covariant formulation of quantum
mechanics. There one sees (as here summarized in Appendix A) that both the spatial
localization X and the time localization T are not good (Dirac) observables of the
theory, because they do not commute with the Hamiltonian constraint. It is increasingly clear that the observable aspects of special-relativistic quantum mechanics are
all codified in properties of the asymptotic states, with the properties of the prepared (incoming) asymptotic state being linked to the properties of the measured
(outgoing) asymptotic state by an S-matrix. And evidently this is a good match for
the conceptual perspective I already introduced in Sect. Emission-Detection Setups
and Wasteful Redundancies, where I observed that even within classical mechanics
all that physics really needs can be coded in properties of the prepared system (at
Alice) and properties of the measured system (at Bob). Matters being as this we
do need material detectors and material clocks but we do not need spacetime.
Troubles with Quantum Tunneling

The observations I reported so far may appear to suggest that the spacetime abstraction is robust in the Galilean limit of quantum mechanics, but looses some of its
reliabiilty in the special-relativistic case. But actually already in the Galilean limit of
quantum mechanics the spacetime abstraction runs occasionally into troubles, especially when quantum-mechanical effects are dominant. For the purposes of these
notes (I shall discuss more examples in Ref. [2]) let me clarify what I mean by this
characterization by focusing on the specific example of quantum tunneling through
a barrier.
197
In the classical limit a particle encountering a potential barrier higher than its
kinetic energy simply cannot manage its way to the other side of the barrier. Quantummechanical effects provide such a particle with a small probability of ending up on
the other side of the barrier. This is well known and well understood. But how should
one describe the position of the particle when it is formally inside the barrier?
And especially how much time does it take a particle to quantum-tunnel through a
barrier? These are tough questions, whose importance has been clear since the early
days of quantum mechanics [10, 11], and remain the main focus of a very active area
of both theory and experimental research [1215].
For speeds much smaller than the speed of light we express the speed v of a
particle in terms of its kinetic energy K and its mass m, and the kinetic energy
is in turn obtained subtracting to the total nonrelativistic energy E the potential
energy U :

2K
2(E U )
=
.
v=
m
m
Since in quantum tunneling E U < 0 this recipe for the speed (and therefore the
corresponding derivation of the travel time) becomes meaningless. We are dealing
with a pure quantum-mechanical effect, the best of cases for exploring the role of
the spacetime abstraction within quantum mechanics.
And (also see Appendix B) what we find experimentally in trying to determine
the tunneling time does challenge the spacetime abstraction. We have at this point
growingly robust evidence of the fact that the results for the tunneling time depend
on the type of clock used. About a dozen different ways (clocks) for determining the
travel time through the barrier are being used, all of which would agree if used as
timing devices in classical-limit contexts, but their determinations of tunneling times
differ [1215].
A useful organizing notion for at least some of these tunneling-time measurements is the Feynamn-path time (see Ref. [12] and references therein) obtained
by averaging the durations of all relevant Feynman paths with the weighting factor
ex p(i S/) (where S here of course denotes the action). But some actual timing procedures (clocks) turn out to agree with the real part of the Feynamn-path time, others
agree with its imaginary part, and others agree with the modulus of the Feynamn-path
time [12]. Consistently with the thesis of these notes there appears to be no time of
spacetime but only time of a specific clock [12].
Relevance for Quantum Gravity?

We often learn physics going upstream. A good example is the program of quantization of theories: from Natures perspective theories start off being quantum and
happen to be amenable to description in terms of classical mechanics only in some
peculiar limiting cases, but our condition is such that we experience more easily
those limiting cases rather than the full quantum manifestation of the laws.
198
G. Amelino-Camelia
I have here argued that also the spacetime abstraction might be a result of our
going upstream. There is no way to introduce spacetime operatively without clocks
and detectors. And yet it is standard to develop theories by introducing the spacetime
picture as first ingredient, then introducing a long list of formal properties of fields
(or particles) in that spacetime, and only in the end we worry about actually having
detectors and clocks in our theory. This worked so far. But there is no guarantee it
will continue to work.
In quantum-gravity research there is a long-standing effort of understanding how
spacetime should be described when both Plancks constant and Newtons constant
G N are nonnegligible. We cannot claim much success addressing this issue.
We could perhaps try attacking the problem from the completely different perspective I am here advocating: we could look for candidate theories of the exchange
of signals among (physical, material) emitters/detectors, now allowing for such theories an interplay between and G N , and without insisting on the availability of a
spacetime abstraction suitable for organizing exactly all such exchanges of signals.
Detectors First and Black-Hole Holography

Perhaps the most natural opportunity for finding first applications of this new
perspective could be the context of studies concerning the holographic description of
black holes. Such holographic descriptions appear to be puzzling if one conceptualizes the physics of black holes as contained in the spacetime region determined by
the black hole. And it should be noticed that the puzzle associated with the entropyarea law [16] becomes more severe for black holes of larger size (the mismatch
between volume scaling and area scaling is increasingly severe as the size of the
region increases).
But according to the perspective I am here advocating the presence of an horizon
should not be described in terms of the structure of the abstracted spacetime but rather
in terms of the network of emitter/detectors that can be setup. Ultimately horizons
would be described as limitations to the network of detectors that can exchange
information. One cannot include in the analysis the notion of detector inside the
black-hole horizon because of the limitations on signal exchange produced by the
horizon. And holography could reflect the fact that the reduction of meaningful
detectors produced by the event horizon also reduces the amount of possible channels
for information exchange among detectors.
The Possibility of Relative Locality

There are also some results recently obtained in the quantum-gravity literature which,
while not getting rid of spacetime altogether, do already provide frameworks for
describing a spacetime abstraction which is weaker than presently assumed, less
199
capable of organizing comprehensively all phenomena. These are results challenging

the absoluteness of locality.
In our current theories, when observer Alice detects particles from an event, and
uses those particle detections as identification of a distant point in spacetime, then,
and this is absolute locality, all other observers can also analogously determine the
position of the event. The point of spacetime inferred through such procedures carries
different coordinates in the different reference frames of different observers, but the
laws of transformation among reference frames ensure that the different observers
see the same spacetime point.
It is emerging [17, 18] that in theories with certain types of Lie-algebra spacetime
noncommutativity [19, 20] and in the quantum-gravity approaches based on group
field theory [21] this absoluteness of locality might be lost. The feature of these
theories that is primarily responsible [17, 18] for the relativity of spacetime locality
is the fact that the translation-symmetry generators, the total-momentum operators,
are not obtained as a linear sum of single-particle momenta (also see Appendix C).
For example, for processes 1 + 2 3 + 4 (two incoming, two outgoing particles)
one can have conservation laws of the form
p[1] p[2] = p[3] p[4]
with k q = k + q , though of course one does recover k q k + q
when all momenta involved are small with respect to the Planck scale [1721].
Then let us consider the determination of the interaction point obtainable by
finding the intersection of the worldlines of the outgoing particles. And let me denote
by x[,A] the intersection point thereby determined by observer Alice, i.e. there is a
value s of the worldlines affine parameter such that x[3,A] (s ) = x[4,A] (s ) = x[,A] .
If one now acts on Alice with a translation of parameters b , as a way to test how this
point x[,A] is viewed by observers distant from Alice, the mentioned nonlinearities of
the sum rule that gives the total-momentum generators affect the analysis nontrivially
[17, 18]. For example one finds that
x[3,B] (s) = x[3,A] (s) + b { p[3] p[4] , x[3,A] (s)} = x[3,A] (s) + b
and in particular x[4,B] (s )x[4,A] (s ) = x[3,B] (s )x[3,A] (s ) (unless the momenta

are much smaller than the Planck scale). In our current theories a spacetime point
is absolutely marked by an intersection of worldlines. In relative-locality theories
processes are still objective [17, 18] but their association to points of a spacetime
is observer dependent, with the familiar spacetime abstraction emerging only for
processes involving particles of energies much smaller than the Planck scale.
200
G. Amelino-Camelia
A Challenge for Experimentalists

To me it is irresistibly intriguing to speculate that our insistence on the availability
of the spacetime abstraction might at this point be limiting our opportunities for
discovery. And I am contemplating a meaningful question: it is for experiments to
decide whether or not the reliability of our spacetime inferences is truly universal.
I here stressed that opportunities for such experimental scrutiny are found already
within ordinary quantum mechanics. For example, we should continue to investigate
whether indeed the tunneling time can only be determined as time of some specific
clock, with different outcomes for different clocks (and therefore no emerging notion
of a clock-independent time of spacetime).
And if my detectors-first perspective does turn out to be applicable to quantumgravity research, there will be (though perhaps in a distant future [22]) other opportunities for experiments looking for evidence against spacetime.
Appendix A: More on Covariant Quantum Mechanics

Within the manifestly-covariant formulation of special-relativistic quantum mechanics, which matured significantly over the last decade [68], the spatial coordinates
and the time coordinate play the same type of role. And there is no evolution, since
dynamics is codified within Diracs quantization as a constraint, just in the same
sense familiar for the covariant formulation of classical mechanics (see, e.g., Chap. 4
of Ref. [23]).
Spatial, X , and time, T , coordinates are well-defined operators on a kinematical
Hilbert space, which is just an ordinary Hilbert space of normalizable wave functions
[8, 9], where they act multiplicatively: X (x, t) = x(x, t), T (x, t) = t(x, t).
And one has a standard description on this kinematical Hilbert space of their conjugate momenta [8, 9]:
X ] = i , [ P,
T ] = [ P0 , X ] = 0
[ P0 , T ] = i , [ P,
Observable properties of the theory are however formulated on the physical Hilbert
space, obtained from the kinematical Hilbert space by enforcing the constraint
of vanishing covariant-Hamiltonian, which in the case of a free special-relativistic
particle takes the form
( P02 P 2 ) physical = 0
The observables of the theory, the Dirac observables, must commute with the
constraint, and this is where one sees the root of the localization problem for specialrelativistic quantum mechanics, which in different fashion had already been noticed
by Newton and Wigner [3]: both T and X are not good observables, even within the
free-particle theory, since they evidently do not commute with P02 P 2 .
201
Appendix B: More on Quantum Tunneling

The study of quantum tunneling has a very long history. However, the quality of
related experimental results has improved significantly over the last two decades
[12, 13], particularly starting with the measurements reported in Ref. [14], where a
two-photon interferometer was used to measure the time delay for a single photon
(i.e. one photon at a time) to tunnel across a well-measured barrier.
And also the understanding of quantum tunneling, and particularly of the tunneling
time, has improved significantly in recent times. Previously there was much controversy particulary revolving around relativistic issues. Under appropriate conditions
[1214] a particle prepared at time ti in a quantum state with peak of the probability
distribution located at a certain position to one side of a barrier is then found on the
other side of the barrier at time t f with distribution peaked at a distance L from the
initial position, with L bigger than c(t f ti ). It is by now well established that this
apparently superluminal behavior is not in conflict with Einsteins relativity. Key
for this emerging understanding is appreciating that in such measurement setups
at first there is a large peaked distribution approaching the barrier from one side,
and then a different (and much smaller, transmitted) peaked distribution is measured
on the other side of the barrier. Contrary to our classical-limit-based intuition, as a
result of quantum-mechanical effects (such as interference) the peak observed after
the barrier is not some simple fraction of the peak that was approaching the barrier.
This sort of travel times of distribution peaks, are not travel times of any signal, and
indeed it is well known that for smooth, frequency-band limited, distributions the
precursor tail of the distribution allows one to infer by analytic continuation [12,
14, 15] the structure of the peak. Even for free propagation, by the time the peak
reaches a detector it carries no new information [12] with respect to the information
already contained in the precursor tail. An example of new information is present
in modified distributions containing abrupt signals [12, 15], and indeed it is found
that when these new-information features are sufficiently sharp they never propagate
superluminally [15].
It is important for the thesis advocated in these notes that, as these theoretical
issues get clarified, and theoretical results get in better agreement with experiments,
we are also getting more an more robust evidence of the fact that in quantifying
the analysis of quantum tunneling we do not have the luxury of referring to some
objective spacetime picture. In particular, there is no single time of spacetime but
rather several possibilities for a time of a specific clock [12].
Appendix C: Relative Locality and Curved

Momentum Space
As mentioned in the main text of these notes the main sources of interest in relative
locality originate from results obtained in studies of certain types of Lie-algebra
spacetime noncommutativity [19, 20] and of group field theory [21], where the
202
G. Amelino-Camelia
generators of translation-symmetry transformations are not described as linear sums

of single-particle momenta. In this appendix I want to highlight a particularly powerful formulation of relative locality that emerges in these theories by formally taking
the limit [17, 18] of both 0 and G N 0 but keeping their ratio /G N fixed. In
this regime both quantum mechanics and gravity are switched off, but the mentioned
nonlinearities for the composition of momenta are found to survive [17, 18] and to
take the form of a manifestation of a nontrivial geometry for momentum space.
The implications for the phase space associated with each particle in this regime
are found to be rather striking [17, 18]: this phase space is the cotangent bundle over
momentum space, which one may denote by R L = T (P). So this regime is, at least
in this respect, dual to the standard classical-gravity regime, where the single-particle
phase space is the cotangent bundle of the spacetime M, which one may denote by
G R = T (M). And just like in the general-relativistic formulation of classical
gravity momenta of particles at different points of spacetime, x and y, can only be
compared by parallel-transporting along some path from x to y, using the spacetime
connection, one finds on R L an analogous problem for the comparison of spacetime
coordinates on the worldlines of two particles, A and B, with different momenta.
These coordinates x A and x B live in different spaces and they can be compared
only in terms of a parallel transport on momentum space. All this is formalized in
Refs. [17, 18] where the relative-locality features mentioned in the main text of these
notes are shown to admit a fully geometric description (in terms of the geometry of
momentum space).
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
H. Poincar, Science and Hypothesis, Chapter 12 (Walter Scott Publishing, London, 1905)
G. Amelino-Camelia, Spacetime in special-relativistic quantum theory, in preparation
T.D. Newton, E.P. Wigner, Rev. Mod. Phys. 21, 400 (1949)
B. Schroer, arXiv:0711.4600
R.-A. Alema-Berenguer, arXiv:philsci-archive.pitt.edu/4313
J.J. Halliwell, Phys. Rev. D64, 04408 (2001)
R. Gambini, R.A. Porto, Phys. Rev. D63, 105014 (2001)
M. Reisenberger, C. Rovelli, Phys. Rev. D65, 125016 (2002)
L. Freidel, F. Girelli, E.R. Livine, Phys. Rev. D75, 105016 (2007)
L.A. MacColl, Phys. Rev. 40, 621 (1932)
E.P. Wigner, Phys. Rev. 98, 145 (1955)
A.M. Steinberg, Lect. Notes Phys. 734, 333 (2008)
H.G. Winful, Phys. Rep. 436, 1 (2006)
A.M. Steinberg, P. Kwiat, R. Chiao, Phys. Rev. Lett. 71, 708 (1993)
M.D. Stenner, D.J. Gauthier, M.A. Neifeld, Nature 425, 695 (2003)
S. Carlip, Lect. Notes Phys. 769, 89 (2009)
G. Amelino-Camelia, L. Freidel, J. Kowalski-Glikman, L. Smolin, Phys. Rev. D84, 084010
(2011). arXiv:1101.0931
18. G. Amelino-Camelia, L. Freidel, J. Kowalski-Glikman, L. Smolin, Gen. Relativ. Gravit. 43,
2547 (2011). arXiv:1106.0313
19. S. Majid, H. Ruegg, Phys. Lett. B334, 348 (1994)
20. S. Majid, arXiv:hep-th/0604130
203
21. D. Oriti, arXiv:0912.2441

22. G. Amelino-Camelia, Nature 408, 661 (2000). arXiv:gr-qc/0012049
23. M. Henneaux, C. Teitelboim, Quantization of Gauge Systems (Princeton University Press,
1992)
Chapter 14
A Chicken-and-Egg Problem:
Which Came First, the Quantum
State or Spacetime?
Torsten Asselmeyer-Maluga
Abstract In this essay I will discuss the question: Is spacetime quantized, as in

quantum geometry, or is it possible to derive the quantization procedure from the
structure of spacetime? All proposals of quantum gravity try to quantize spacetime
or derive it as an emergent phenomenon. In this essay, all major approaches are
analyzed to find an alternative to a discrete structure on spacetime or to the emergence
of spacetime. Here I will present the idea that spacetime defines the quantum state
by using new developments in the differential topology of 3- and 4-manifolds. In
particular the plethora of exotic smoothness structures in dimension 4 could be the
corner stone of quantum gravity.
Basic Assumptions in Quantum Gravity

General relativity (GR) has changed our understanding of spacetime. In parallel, the
appearance of quantum field theory (QFT) has modified our view of particles, fields
and the measurement process. The usual approach for the unification of QFT and
GR, to a quantum gravity, starts with a proposal to quantize GR and its underlying
structure, spacetime. There is an unique opinion in the community about the relation
between geometry and quantum theory: The geometry as used in GR is classical and
should emerge from a quantum gravity in the limit (Plancks constant tends to zero).
Most theories went a step further and try to get a spacetime from quantum theory. But
what happens if this prerequisite is wrong? Is it possible to derive the quantization
procedure from the structure of space and time? My own research program [113]
seems to imply a simple answer: Yes. But in this essay I will discuss it from a more
general perspective and analyze the basic assumptions of quantum gravity1 first:
There are many books about quantum gravity, for instance [14], and the original papers which
I omit to cite.
T. Asselmeyer-Maluga (B)
German Aerospace Center, Berlin, Germany
205
206
T. Asselmeyer-Maluga
1. Spacetime has dimension 4 (also true in superstring theory after compactification).

2. Classical spacetime, i.e. the spacetime of GR or the spacetime as a limit of quantum gravity, is a smooth, non-compact 4-manifold with Lorentz structure and
(trivial) codimension-1 foliation of the form R ( a smooth 3-manifold, the
space).
3. In the process of quantization, spacetime admits a discrete structure and one
obtains the continuum only in the limit of large scales.
4. The quantum state in quantum gravity is realized as element (or vector) in some
abstract state space (the space of all connections, the space of all spin networks
etc.) with no direct reference to the classical spacetime.
5. Quantum gravity (as containing GR in some limit) must be background-independent, i.e. it does not explicitly depend on a concrete shape of the spacetime
(diffeomorphism invariance).
6. Classical spacetime is induced (as emergent structure) by the quantum state in
quantum gravity in the semi-classical limit.
All other explicit or implicit assumptions in concrete quantum gravity theories (like
action principles or the dynamics) are very specific for the corresponding approach
to quantum gravity. Let us analyze its relevance. The first assumption2 is merely
known from our experience but we must realize that four-dimensional spaces have
an exceptional property among other spaces which is important for physics, exotic
smoothness. The second assumption is motivated by GR and the research on the
implementation of causality in GR. It is the starting point of Loop quantum gravity.
The form of causality which was used to establish R (see [15, 16]) is certainly too strong. There is a unique path into the past but not an unique path in the
future, otherwise one needs the net of many-worlds. So, the second assumptions
can be weakened to consider every kind of codimension-1 foliation on a smooth,
non-compact 4-manifold. It is known that a Lorentz structure will be a result of this
foliation. The third assumption is the starting point of some approaches to quantum gravity (Causal nets, Regge calculus etc.) but all approaches are looking for
discrete structures (spin networks and area quantization of Loop quantum gravity,
finite string length in String theory) or to explicitly construct them (Dynamical triangulation etc.). Interestingly, all these approaches have problems in reaching the
continuum limit. But is it really necessary to reduce the continuous spacetime to
a discrete space? From GR we know that (by using diffeomorphism invariance) a
point has no meaning in GR (see the next section for the whole discussion). But
one can use relative techniques to relate two subspace of the spacetime to each
other. Then only statements like the two submanifolds intersect transversally can
be decided without introducing a concrete coordinate system. But statements of this
kind are part of a mathematical theory called (differential) topology with discrete
structures used to classify topological spaces. I will discuss these in more detail in
Sect. Differential Topology Unveils the Quantum Nature of Spacetime. The fourth
assumption is the most problematic one. Each approach to quantum gravity can be
2
Current experiments at the LHC do not give any sign for extra dimensions (see the Particle Data
Group).
14 A Chicken-and-Egg Problem: Which Came First, the Quantum State
207
distinguished by its state space. So, the question must be: what is the natural state
space of quantum gravity? Geometrodynamics started with the original superspace,
i.e. the space of 3-metrics, with the Wheeler-deWitt equation as dynamics. This program was superseded by the introduction of Ashtekar variables (densitized coframe
and SU (2) connection) leading to Loop quantum gravity. The state space is the
space of spin networks but the solution of the Hamilton constraint (as analog to the
Wheeler-deWitt equation) is an unsolved problem. All state spaces are rather artificial and the relation to the geometry of the spacetime is poor. Part of the problem is
the answer to the question: Is the quantum state real? If the answer is yes one has to
consider quantum states in spacetimes. In particular one has to interpret the superposition of states. I will discuss this problem also in Sect. Differential Topology
Unveils the Quantum Nature of Spacetime. The fifth assumption is induced directly
from GR. It is one of the crucial points in GR: the theory has to be formulated without
the reference to coordinates or the choice of a coordinate system does not influence
the result. The implementation of this assumption is usually done by connecting it
with the sixth assumption: there is no spacetime at the quantum level, the classical
spacetime emerges from the deeper quantum level. Theories with a fuzzy spacetime including also noncommutative geometry are promising candidates for a direct
implementation of quantum properties into spacetime.
In summary, especially the last four assumptions are questioned in this essay. In
the following I will argue that the spacetime has the right properties for a spacetimepicture of QFT. Quantum gravity should be also part of this picture.
The First Sign: Spacetime Is More Than Classical

When Einstein developed GR, his opinion about the importance of general covariance changed over the years. In 1914, he wrote a joint paper with Grossmann.
There, he rejected general covariance by the now famous hole argument. But after
a painful year, he again considered general covariance now with the insight that
there is no meaning in referring to the spacetime point A or the event A, without further specifications. Therefore the measurement of a point without a detailed
specification of the whole measurement process is meaningless in GR. The reason is simply the diffeomorphism-invariance of GR which has tremendous consequences. Physical observables have to be diffeomorphism-invariant expressions. In
most cases, this demand is impossible. The momentum of a moving particle is not
diffeomorphism-invariant but it is physically meaningful. Therefore this momentum
must be seen in the context of a measurement device which includes a concrete
coordinate system. But generally, GR is background-independent of any coordinate
system and, as usually thought, from the topology of the spacetime. So, if one fixes
the topology then GR depends only on the (diffeomorphism-class) metric. But in
Sect. The Secret Revolution: Our Understanding of 3- and 4-Manifolds, I will
discuss another possibility. Let M be a smooth 4-manifold, say M is the topologically R4 but with Lorentz structure, i.e. the Minkowski space. M is flat and the GR
208
vacuum equations are trivially fulfilled. The (smooth) atlas of M is called the smoothness structure unique up to diffeomorphisms. One would expect that there is only
one smooth atlas for M, all other possibilities can be transformed into each other
by a diffeomorphism. But this is not true, see Sect. The Secret Revolution: Our
Understanding of 3- and 4-Manifolds . In fact, there are infinitely many nonequivalent smoothness structures on M with no heuristic to distinguished one above
the others as physically relevant.
The Concept of Spacetime in the Main Approaches

to Quantum Gravity
Quantum gravity as the unification of GR and QFT is currently an open problem but
with a long list of possible candidates. Among them are Superstring theory and Loop
quantum gravity. For each of these there are two classes: background-dependent and
background-independent, respectively. But all proposals of a quantum gravity theory
have one assumption in common: the spacetime has a foam-like structure or is discrete
from the beginning. In string theory, the string has the extension of one Planck length
and all structures below this length are unimportant. In particular the whole spacetime is seen as an emergent phenomenon which must be derived from the full theory.
The background-dependence is a problem but it will be resolved in the conjectured
M-theory. The relation to GR is interesting. At the non-quantized level, the world
surface of the string is embedded in some background. If this embedding is fixed then
the curvature of the world surface is determined by the curvature of the embedding
space and vice versa. On the quantum level, one usually argues with the help of
the function, encoding the dependence of a coupling parameter in the renormalization group. For a conformal invariant theory (conformal invariance of the world
surface), this function has to vanish. In the case of string theory, one obtains the vanishing of the scalar curvature of the embedding space or the Einstein equation for the
vacuum. In Loop quantum gravity, the central theme is background-independence
installed by the vanishing of three constraints. One of the main results of the theory
is the quantization of the area and the volume. It seems that spacetime must have
a discrete structure. But the result was obtained by using an eigenvalue equation.
As in the case of a harmonic oscillator, the underlying space is continuous but the
spectrum of the area or volume operator is discrete. In Sect. Differential Topology
Unveils the Quantum Nature of Spacetime , I will give an example of a class of spaces
(hyperbolic 3-manifolds) with this property. Many other proposals start with a discrete structure from the scratch: Causal Sets, Quantum Causal Histories, Dynamical
Triangulations etc. Therefore in all current quantum gravity proposals, the spacetime
has a discrete structure or emerges from a discrete structure. But is the spacetime
model of a smooth manifold dead? Mathematicians have found many interesting
properties of 3- and 4-manifolds.
209
The Secret Revolution: Our Understanding

of 3- and 4-Manifolds
According to GR, spacetime is a smooth 4-manifold carrying a Lorentz-structure. The
existence of a Lorentz-structure is closely related to the existence of a codimension1 foliation. Therefore one also has to consider the spatial component, a smooth
3-manifold, as leaf of the foliation. From the mathematical point of view, one has to
look into the theory of 3- and 4-manifolds (see [3] for more details and the references).
The central concept is the smooth manifold as a generalization of a surface. Riemann was the first one to do extensive work generalizing the idea of a surface to
higher dimensions. The name manifold comes from Riemanns original German
term, Mannigfaltigkeit. In his Gttingen inaugural lecture (Ueber die Hypothesen,
welche der Geometrie zu Grunde liegen), Riemann described the set of all possible values of a variable with certain constraints as a Mannigfaltigkeit, because
the variable can have many values. By induction, Riemann was able to define an
ntimes extended manifold. Then Poincare gave a definition of a (differentiable)
manifold (varit) as subset of some Euclidean space which served as a precursor
to the modern concept of a manifold. During the 1930s Hassler Whitney and others
clarified the foundational aspects of the subject. In particular, a manifold was defined
intrinsically with charts and transition functions and the equivalence of this definition with Poincares Euclidean subsets was shown (Whitneys embedding theorem).
The development of algebraic topology gave the right frame to investigate the manifolds (classification of surfaces by homology, the fundamental group etc.). The first
surprise came in 1957. John Milnor constructed the first seven exotic 7-spheres (7dimensional spheres which are homeomorphic but not diffeomorphic to each other).
Therefore there must be a difference between a smooth manifold (all charts as maps
to a Euclidean space and transition functions are smooth maps) and a topological
manifold (the maps are now only continuous). The new topic differential topology
was born. The (smooth) atlas (all smooth charts and transitions function to cover the
manifold) was called smoothness structure. But how many in-equivalent smoothness
structures on a nmanifold exists? Using powerful methods (like the h-cobordism
theorem), Kervaire and Milnor were able to determine it for the spheres. Number of
exotic nspheres up to dimension 13: (Table 14.1).
In particular, it was shown that the number of in-equivalent smoothness structure on a nmanifold is only finite for n > 4. The uniqueness in low dimensions
n < 4 is a classical result. Therefore only the 4-dimensional case was open. All
methods developed for 2- and 3-manifolds or for higher-dimensional (n > 4) manifolds were useless in this dimension. In 1973, Andrew Casson gave lectures about
a new construction in dimension four, which he called flexible handles (now known
Table 14.1 Number of exotic nspheres up to dimension 13

Dimension n
1
2
3
4
5
6
7
8
#Diff (S n )
28
10
11
12
13
992
210
as Casson handles). The mimeographed notes of these three lecture came to Micheal
Freedman and occupied him over the next 7 years. Eventually, he successfully classified simply-connected, compact 4-manifolds. Meanwhile, Simon Donaldson (a student of M. Atiyah) started the investigation of anti-self-dual connections of a SU (2)
gauge theory, in physics known as instantons. As a surprising result, he proved that
not all topological, compact 4-manifolds are smoothable. With this result in mind,
Kirby, Freedman and Gompf constructed the first examples of an exotic R4 , i.e. a
space homeomorphic to R4 but not diffeomorphic. The second surprise came in the
form of the number of inequivalent smoothness structures: there are countably infinite many for most compact 4-manifolds and uncountably infinite many for most
non-compact 4-manifolds including R4 . The development of this topic is not complete. In particular for the simplest compact 4-manifold, the 4-sphere, we do not know
the number of in-equivalent smoothness structures. Most mathematicians conjecture
that there are countably infinite many structures on the 4-sphere. So, dimension 4 is
exceptional for smoothness!
One point is also important for quantum gravity. With Milnors discovery of
exotic 7-spheres, one started to look for the existence of other structures on manifolds. In particular the existence question for a triangulation (or piecewise-linear
structure, PL-structure) was very interesting. The surprising result of Cerf for manifolds of dimension smaller than seven was simple: PL-structure (or triangulations)
and smoothness structure are the same. This implies that every PL-structure can
be smoothed to a smoothness structure and vice verse. But also for manifolds of
dimension higher than three, there is a big difference between the topological and
PL/smooth structure, i.e. not every topological manifold is smoothable/triangulable
or uniquely smoothable/triangulable. Therefore the discrete approach (via triangulations) and the smooth manifold are the same!
The theory of 3-manifolds also changed its view in the 80s. The theory of
3-manifolds was strongly influenced by the Poincar conjecture but the progress
was slow before 1957. In 1957, Papakyriakopoulos proved three technical theorems
(the sphere and loop theorem as well Dehns lemma) with tremendous impact on the
whole 3-manifold theory. Next, Milnor proved that every compact 3-manifold can be
split into an unique decomposition of prime manifolds (3-manifolds which are not
splittable by using the connected sum). In 1979, Jaco, Shalen and Johannson found
a finer decomposition of some prime manifolds by cutting them into pieces along
embedded tori (JSJ decomposition). But the real breakthrough came from Thurston
around 1980. His work was inspired by the construction of geometric structures (metric of constant curvature simulated by a homogeneous space) for the complement
of knots in the 3-sphere. Based on this examples, he conjectured that any compact
3-manifold can be cut into pieces so that every piece admits a geometric structure
(Geometrization conjecture). If this conjecture is true then Poincares conjecture is
also settled. So, the topological structure of 3-manifolds has a lot to do with its
geometry! In the 80s, Hamilton developed the Ricci flow technique to prove this
conjecture with very interesting results. But in 2002 and 2003, Perelman submitted three papers to arxiv.org in which he proposed a proof of the Geometrization
conjecture. The arguments were checked by many mathematicians but no error was
211
found. According to Hamilton and Perelman, the a 3-manifold can be determined

topologically by considering the long-time behavior of the Ricci flow or a 3-manifold
is determined by a flow in the space of metrics.
Differential Topology Unveils the Quantum

Nature of Spacetime
As mentioned above, differential topology is the mathematical theory of smooth
manifolds including the (smooth) relations between submanifolds. In the first section,
the basic assumptions of quantum gravity were discussed. The usage of a smooth
4-manifold as spacetime was not questioned. However the choice of the smoothness
structure is not unique and I will discuss it now. As starting point consider a toy
model to visualize the changes of the smoothness structure. Start with a torus (or
doughnut) T 2 = S 1 S 1 . Now one cuts the torus along one circle to obtain a cylinder,
twist one end of the cylinder by 2 and glue the two ends together. This process is
called a Dehn twist and one obtains a torus again but now with a twist. But by the
classification of closed surfaces, the usual and the twisted torus are diffeomorphic to
each other. This fact is amazing but both tori are located in different components of
the diffeomorphism group, i.e. in two different isotopy classes. In particular there is
no coordinate transformation (i.e. a diffeomorphism connected to the identity) which
transforms the twisted torus to the usual torus. This toy model shows the difference
between coordinate transformations (diffeomorphisms connected to the identity) and
global diffeomorphisms. A similar effect is the change of the smoothness structure
(to a non-equivalent one) but it cannot be visualized in this easy way. As an example
of this change I will consider a compact 4-manifold M (topologically complicated
enough, i.e. a K3 surface or more advanced) containing a special torus Tc2 (so called
c-embedded torus). Now cut out a neighborhood D 2 Tc2 of this torus (with boundary
a 3-torus T 3 ) and glue in (S 3 \ (D 2 K )) S 1 (having also the boundary T 3 ) where
S 3 \ (D 2 K ) denotes the complement of a knot K in the 3-sphere S 3 . Then one
obtains

(14.1)
M K = M \ D 2 Tc2 T 3 (S 3 \ (D 2 K )) S 1
a new 4-manifold M K which is homeomorphic to M (Fintushel-Stern knot surgery
[17]). If the knot is non-trivial then M K is not diffeomorphic to M. One calls M K
an exotic 4-manifold, a misleading term. Nothing is really exotic here. Nearly all
smoothness structures on a 4-manifold are exotic.
Now consider the physically significant non-compact examples of exotic
4-manifolds like R4 and S 3 R. Start with S 3 R. This non-compact 4-manifold
has the usual form used in GR. There is a global foliation along R, i.e. S 3 {t}
with t R are the (spatial) leafs. S 3 R with this foliation is called the standard
S 3 R. I will denote an exotic version by S 3 R. The construction of S 3 R is
212
rather complicated (see [18]). As a main ingredient one needs a homology 3-sphere
(i.e. a compact, closed 3-manifold with the homology groups of the 3-sphere)
which does not bound a contractable 4-manifold (i.e. a 4-manifold which can be
contracted to a point by a smooth homotopy). Interestingly, this homology 3-sphere
is smoothly embedded in S 3 R (as cross section, i.e. {0} S 3 R).
From the geometrical point of view, this 3-manifold is also very interesting. One can
choose so that it admits a hyperbolic structure, i.e. a homogeneous metric of constant negative curvature. Hyperbolic 3-manifolds have a special property: Mostow
rigidity [19]. Every diffeomorphism (especially every conformal transformation) of
a hyperbolic 3manifold is induced by an isometry. Therefore the volume and the
curvature are topological invariants for hyperbolic 3-manifolds. In particular there
are surfaces in hyperbolic 3-manifolds (incompressible surfaces) having a special
(not divisible) volume. Then one obtains also a kind of quantized areas by purely
topological methods.
What about the foliation of S 3 R? There is no foliation along R but there is
a codimension-one foliation of the 3-sphere S 3 (see [20] for the construction). So,
S 3 R is foliated along S 3 and the leafs are Si R with the surfaces {Si }iI S 3 .
But what happens with the 3-spheres in S 3 R? There is no smoothly embedded
S 3 in S 3 R (otherwise it would have the standard smoothness structure). But
there is a wildly embedded S 3 ! Let i : K M be an embedding of K (with
dim K < dim M). One calls the embedding i wild if i(K ) is not a finite polyhedron
(or i(K ) is not triangulated by a finite number of simplices). See the example of wildly
embedded circles in the 3-space in Fig. 14.1, the famous Fox-Artin wild knot. In [7],
we considered wildly embedded submanifolds as models of quantum D-branes. The
prominent example of a wildly embedded submanifold is Alexanders horned sphere.
Wild embedded submanifolds are fractals in a generalized manner. Now I will argue
in the following that this wild embedding is a geometric model for a quantum state.
As discussed in the first section, all approaches of quantum gravity have problems
with the construction of the state space. If I assume that the spacetime has the right
properties for a spacetime picture of quantum gravity then the quantum state must be
part of the spacetime or must be geometrically realized in the spacetime. Consider
(as in geometrodynamics) a 3-sphere S 3 with metric g. This metric (as state of GR)
is modeled on S 3 at every 3-dimensional subspace. If g is a metric of a homogeneous
space then one can choose a small coordinate patch. But if g is inhomogeneous then
one can use a diffeomorphism to concentrate the inhomogeneity at a chart. Now
Fig. 14.1 Examples of wild

knots
213
one combines these infinite charts (I consider only metrics up to diffeomorphisms)

into a 3-sphere but without destroying the infinite charts by a diffeomorphism. Wild
embeddings are the right structure for this idea. A wild embedding cannot be undone
by a diffeomorphism of the embedding space. Secondly, this wild embedding of a
3-sphere into S 3 R is determined by its complement. So, if one understands the
complement of the wild embedding then one understands the wild embedding itself.
Interestingly, one can construct an operator algebra and a Hilbert space from a wild
embedding (see Appendix A, below). This operator algebra can be also obtained
from the foliation (by using the noncommutative geometry approach), see [10, 21].
It is the hyperfinite factor III1 von Neumann algebra having the structure of the
local algebras in a relativistic QFT with one vacuum vector. Then one obtains a
background-independent approach but how does the classical spacetime appears? A
wild embedding is defined by the infinite polyhedron. So, if I reduces the wild embedding to a finite polyhedron by contracting the smaller parts to zero then one obtains
a finite polyhedron. But by definition, a finite polyhedron is not a wild embedding
(or comes from a wild embedding). In [7], the process was discussed for quantum
D-branes. In particular, the classical action of a D-brane was obtained. But if we are
able to reduce the wild embedding (quantum state) to a tame embedding (classical
state) then we have to show that a wild embedding is the quantization of a tame
embedding. In [22]we showed even this fact:
The wild embedding can be obtained by a deformation quantization (DrinfeldTuraev quantization) from a tame embedding. Furthermore, we constructed one of
the main ingredients for a quantum field theory (C algebra and the observablen
algebra of von Neumann type) for Alexanders horned sphere and identifying it with
the knwon results (so-called factor III1 ).
Decoherence and Inflation

If our approach to identify the wild embedding with a quantum state will be successful
then we have to present an interpretation of decoherence or the measurement process
in quantum mechanics. As an example [23] we will consider the quantum state at the
beginning of the universe, i.e. a wildly embedded 3-sphere. In our model of the exotic
S 3 R we made a rescaling so that the 3-sphere at t = is the initial state of the
cosmos (at the big bang), i.e. we assume that the cosmos starts as a small 3-sphere (of
Planck radius). But in an exotic S 3 R, every 3-sphere (at every time) is a wildly
embedded 3-sphere S3 . In the model of S 3 R we have a topologically complicated
3-manifold (a homology 3-sphere) at a later time step (say at t = 0). As Freedman
[18] this 3-manifold is smoothly embedded, i.e. it represents a classical state.
Therefore we obtain the transition
decoherence
quantum state S3 classical state
(14.2)
214
and we studied this process in [24] more carefully. This process has an exponential
rate of expansion, a process usually called inflation today. The reason of this inflation
is the transition from the quantum state (wildly embedded 3-sphere) to the classical
state (complicated, smoothly embedded 3-manifold = tame embedding). At this
stage, our model is very explicit: an infinite polyhedron (wildly embedded manifold)
is reduced to a finite polyhedron (tame embedding) which can be part of the infinite
polyhedron. Mathematically we obtain a projection or state reduction or the collapse
of the wave function.
This particular example showed the main features of the decoherence process.
Our model is general enough to explain also the decoherence process for wildly
embedded subsystems. But I will point to one interesting result: the smoothness
structure determines the classical state in our model above. If one generalize this result
then (differential) topology has to be included in the discussion of the measurement
process.
In summary, we obtained a state space as operator algebra of the wild embedding
induced by exotic smoothness structures. The state space is
1. background-independent (diffeomorphism invariant)
2. with countable infinite basis (discrete structure)
3. and contains the classical spacetime as limit.
Of course, the whole approach is very theoretical up to now. For instance I do not
start with a concrete action or list of fields. But sometimes, things went better than
expected. In [12], we considered the Fintushel-Stern knot surgery above to obtain
(14.1), the exotic M K . If one started with the Einstein-Hilbert action on M then we
obtained the combined Einstein-Hilbert-Dirac-Yang-Mills system. The knot complement is directly related to the fermions whereas the bosons appear as torus bundles
(the pieces between the knot complements). In an extension of this work [25], the
Higgs mechanism was also included. A lot of work has to be done but it is a beginning.
Conclusion
I have presented a certain number of ideas and results:
1. There is a freedom in the definition of the spacetime coming from the choice of
the smoothness structure.
2. There are an infinity of exotic smoothness structures to choose from. For example
the foliation of an exotic spacetime like S 3 R can be very complicated.
3. For the usual foliation S 3 {t} with t R of S 3 R the 3-sphere must be a
wildly embedded submanifold (represented by an infinite polyhedron).
4. A quantum state can be defined on the spacetime as wild embedding.
5. A glimpse of an action to obtain a full QFT and quantum gravity is also obtained.
Before concluding, I must add that the views expressed are only partly original. I have
partially drawn from the ideas of Carl H. Brans, Jerzy Krl and Helge Ros.
215
Appendix A: C algebras Associated

to Wild Embeddings
Let I : K n Rn+k be a wild embedding of codimension k with k = 0, 1, 2.
In the following we assume that the complement Rn+k \ I (K n ) is non-trivial, i.e.
1 (Rn+k \ I (K n )) = = 1. Now one defines the C algebra C (G, ) associated
to the complement G = Rn+k \ I (K n ) with group = 1 (G). If is non-trivial then
this group is not finitely generated. The construction of wild embeddings is given
by an infinite construction3 (see Antoines necklace or Alexanders horned sphere).
From an abstract point of view, we have a decomposition of G by an infinite union
G=
Ci
i=0
of level sets Ci . Then every element lies (up to homotopy) in a finite union
of levels.
The basic elements of the C algebra C (G, ) are smooth half-densities with
1/2
compact supports on G, f Cc (G, 1/2 ), where for is the onedimensional complex vector space of maps from the exterior power 2 L, of the
union of levels L representing , to C such that
() = ||1/2 ()
2 L , R .
For f, g Cc (G, 1/2 ), the convolution product f g is given by the equality

f (1 )g(2 )
( f g)() =
1 2 =
with the group operation 1 2 in . Then we define via f () = f ( 1 ) a

operation making Cc (G, 1/2 ) into a algebra. Each level set Ci consists of simple
pieces (for instance tubes in case of the Alexanders horned sphere) denoted by T .
For these pieces, one has a natural representation of Cc (G, 1/2 ) on the L 2 space
over T . Then one defines the representation

(x ( f ))() =
f (1 )(2 )
L 2 (T ), x .
1 2 =
This infinite construction is necessary to obtain an infinite polyhedron, the defining property of a
wild embedding.
216
The completion of Cc (G, 1/2 ) with respect to the norm

|| f || = sup ||x ( f )||
xG
makes it into a C algebra Cc (G, ). The C algebra Cc (K , I ) associated to the

wild embedding I is defined to be Cc (K , j) = Cc (G, ). The GNS representation
of this algebra is called the state space.
References
1. T. Asselmeyer-Maluga, Exotic smoothness and quantum gravity. Class. Quant. Gravity
27:165002 (2010). arXiv:1003.5506v1 [gr-qc]
2. T. Asselmeyer-Maluga, C.H. Brans, Cosmological anomalies and exotic smoothness structures.
Gen. Relativ. Gravit. 34, 597607 (2002)
3. T. Asselmeyer-Maluga, C.H. Brans, Exotic Smoothness and Physics (World Scientific Publication, Singapore, 2007)
4. T. Asselmeyer-Maluga, J. Krl, Small exotic smooth R 4 and string theory, in International
Congress of Mathematicians ICM 2010 Short Communications Abstracts Book, ed. R. Bathia
(Hindustan Book Agency, 2010), p. 400
5. T. Asselmeyer-Maluga, J. Krl, Constructing a quantum field theory from spacetime (2011).
arXiv:1107.3458
6. T. Asselmeyer-Maluga, J. Krl, Exotic smooth R 4 and certain configurations of NS and D
branes in string theory. Int. J. Mod. Phys. A 26, 13751388 (2011). arXiv:1101.3169
7. T. Asselmeyer-Maluga, J. Krl, Topological quantum d-branes and wild embeddings from
exotic smooth R 4 . Int. J. Mod. Phys. A 26, 34213437 (2011). arXiv:1105.1557
8. T. Asselmeyer-Maluga, J. Krl, On topological restrictions of the spacetime in cosmology.
Mod. Phys. Lett. A 27, 1250135 (2012). arXiv:1206.4796
9. T. Asselmeyer-Maluga, J. Krl, Quantum D-branes and exotic smooth R4 . Int. J. Geom. Methods Mod. Phys. 9, 1250022 (2012). arXiv:1102.3274
10. T. Asselmeyer-Maluga, R. Mader, Exotic R 4 and quantum field theory, in 7th International
Conference on Quantum Theory and Symmetries (QTS7), ed. by C. Burdik et al. (IOP Publishing, Bristol, 2012), p. 012011. doi:10.1088/1742-6596/343/1/012011, arXiv:1112.4885
11. T. Asselmeyer-Maluga, H. Ros, Dark energy and 3-manifold topology. Acta Phys. Pol. 38,
36333639 (2007)
12. T. Asselmeyer-Maluga, H. Ros, On the geometrization of matter by exotic smoothness. Gen.
Relativ. Gravit. 44, 28252856 (2012). doi:10.1007/s10714-012-1419-3, arXiv:1006.2230
13. T. Asselmeyer, Generation of source terms in general relativity by differential structures. Class.
Quant. Gravity 14, 749758 (1996)
14. C. Rovelli, Quantum Gravity. Cambridge Monographs on Mathematical Physics (Cambridge
University Press, Cambridge, 2004). www.cpt.univ-mrs.fr/~rovelli/book.pdf
15. A.N. Bernal, M. Sanchez, Smoothness of time functions and the metric splitting of globally
hyperbolic space times. Commun. Math. Phys. 257, 4350 (2005). arXiv:gr-qc/0401112
16. A.N. Bernal, M. Sanchez, Globally hyperbolic spacetimes can be defined as causal instead
of strongly causal. Class. Quant. Gravity 24, 745750 (2007). arXiv:gr-qc/0611138
17. R. Fintushel, R. Stern, Knots, links, and 4-manifolds. Inven. Math. 134, 363400 (1998) (dgga/9612014)
18. M.H. Freedman, A fake S 3 R. Ann. Math. 110, 177201 (1979)
19. G.D. Mostow, Quasi-conformal mappings in n-space and the rigidity of hyperbolic space forms.
Publ. Math. IHS 34, 53104 (1968)
217
20. T. Asselmeyer-Maluga, J. Krl, Abelian Gerbes, generalized geometries and exotic R 4 . J. Math.
Phys. (2009). arXiv:0904.1276
21. T. Asselmeyer-Maluga, J. Krl, Exotic smooth R4 , noncommutative algebras and quantization
(2010). arXiv:1001.0882
22. T. Asselmeyer-Maluga, J. Krl, Quantum geometry and wild embeddings as quantum states.
Int. J. Geom. Methods Mod. Phys. 10(10) (2013). arXiv:1211.3012
23. T. Asselmeyer-Maluga, J. Krl, Decoherence in quantum cosmology and the cosmological constant. Mod. Phys. Lett. A 28, 350158 (2013). doi:10.1142/S0217732313501587,
arXiv:1309.7206
24. T. Asselmeyer-Maluga, J. Krl, Inflation and topological phase transition driven by
exotic smoothness. Adv. HEP 2014, 867460 (2014). http://dx.doi.org/10.1155/2014/867460,
arXiv:1401.4815
25. T. Asselmeyer-Maluga, J. Krl, Higgs potential and confinement in Yang-Mills theory on exotic
R4 (2013). arXiv:1303.1632
Chapter 15
Gravity Can Be Neither Classical

Nor Quantized
Sabine Hossenfelder
Abstract I argue that it is possible for a theory to be neither quantized nor classical.
We should therefore give up the assumption that the fundamental theory which
describes gravity at shortest distances must either be quantized, or quantization must
emerge from a fundamentally classical theory. To illustrate my point I will discuss an
example for a theory that is neither classical nor quantized, and argue that it has the
potential to resolve the tensions between the quantum field theories of the standard
model and general relativity.
To Quantize or not to Quantize Gravity

Gravity stands apart from the other three interactions of the standard model by its
refusal to be quantized. To be more precise, quantizing gravity is not the actual
problem; gravity can be perturbatively quantized. The problem is that the so quantized
theory cannot be used at energies close by and above the Planck energy, and thus
cannot be considered a fundamental theory; it is said to be non-renormalizable,
meaning it has no predictive power in the extremely high energy regime.
This mismatch between the quantum field theories of the standard model and classical general relativity is more than an aesthetic problem: It signifies a severe shortcoming of our understanding of nature. This shortcoming has drawn a lot of attention
because its resolution it is an opportunity to completely overhaul our understanding
of space, time and matter. The search for a consistent theory of quantum gravity
that could be applied also at Planckian energies, or strong curvature respectively,
has thus lead to many proposals. But progress has been slow and in the absence
of experimental evidence, our reasons for the necessity of quantizing gravity are
theoretical:
S. Hossenfelder (B)
Nordita, Roslagstullsbacken 23, 106 91 Stockholm, Sweden
219
220
S. Hossenfelder
1. Classical general relativity predicts the formation of singularities under quite

general circumstances. Such singularities are unphysical and should not occur
in a fundamentally meaningful theory. It is expected that quantum gravity is
necessary to prevent the formation of singularities.
2. Applying quantum field theory in a curved background at small curvature leads
to the evaporation of black holes, as first shown by Hawking [1]. This black hole
evaporation however seems to violate unitary which is incompatible with quantum mechanics. It is widely believed that quantum gravitational effects restore
unitarity and information is conserved.
3. All quantum fields carry energy so they all need to couple to the gravitational
field, but we do not know a consistent way to couple a quantum field to a classical
field. As Hannah and Eppley have argued [2], the attempt to do such a coupling
leads either to a violation of the uncertainty principle (and thus would necessitate
a change of the quantum theory) or to the possibility of superluminal signaling,
which brings more problems than it solves. While Mattingly has argued [3] that
Hannah and Eppleys thought experiment can not be carried out in our universe,
that does not address the problem of consistency.
These issues have all been extensively studied and discussed in the literature and
are familiar ground. The most obvious way to address them seems to be a nonperturbation theory in one or other form, and several attempts to construct one are
under way. I will use the opportunity of the essay contest to stray from the welltrodden ground and argue that we should instead reinvestigate the apparent tension
between the quantized matter and non-quantized gravity. It is worthwhile for the
following to recall the problems with coupling a classical to a quantum field.
The first problem, as illuminated by Hannah and Eppley is that the classical and the
quantum fields would have different uncertainty relations, and their coupling would
require a modification of the quantum theory. Just coupling them as they are leads to
an inconsistent theory. The beauty of Hannah and Eppleys thought argument is its
generality, but that is also its shortcoming, because it does not tell us how a suitable
modification of quantum theory could allow such a coupling to be consistent.
The second problem is that it is unclear how mathematically the coupling should be
realized, as the quantum field is operator-valued and the classical field is a function
on space-time. One possible answer to this is that any function can be identified
with an operator on the Hilbert space by multiplying it with the identity. However,
the associated operators would always be commuting, so they are of limited use to
construct a geometrical quantity that can be set equal to the operator of the stressenergy-tensor (SET) of the quantum fields.
Another way to realize the coupling is to construct classical field from the operator
of the SET by taking the expectation value. The problem with this approach is that
the expectation value may differ before and after measurement, which then conflicts
with the local conservation laws of general relativity. Coupling the classical field to
the SETs expectation value is thus usually considered valid only in approximation
when superpositions carry negligible amounts of energy.
15 Gravity Can Be Neither Classical Nor Quantized
221
Because of these difficulties to make sense of the theory, leaving gravity classical
while the other interactions are quantized is not a very promising option. However,
this theoretical assessment should be supported by experimental test; recent proposals
for this have been put forward in [4, 5].
How to Be Neither Classical Nor Quantized

Let us carefully retrace the logic of the arguments in the previous section.
We have experimental evidence that matter is quantized in the energy regimes
that we have tested. We cannot leave gravity unquantized if it couples to quantized
matter. Thus gravity has to be quantized in the energy regimes we have tested. We can
quantize gravity perturbatively. This theory does make sense in the energy regimes
that we have tested, but does not make sense in the strong curvature regime. We have
no experimental evidence for the existence and properties of singularities or black
hole evaporation, or the behavior of matter in the strong curvature regime.
To conclude from the previous paragraph that we need a non-perturbative completion of quantum gravity necessitates a further assumption, that is that the quantization
procedure itself is independent of the energy range at which we apply the theory. It
is this assumption that I argue should be given up.
We normally think of a theory as either being quantized or classical, but let us
entertain the possibility that quantization is energy-dependent. Concretely, consider
that Plancks constant is a field whose value at high energies goes to zero. In four
space-time dimensions, Newtons constant is G = c/m 2Pl , so if we keep mass units
fix, G will go to zero together with , thereby decoupling gravity. If gravity decouples,
theres no reason for singularities to form. If gravity becomes classical, theres no
problem with the perturbative expansion. So this possibility seems intriguing, if
somewhat vague. I will now make this idea more concrete and then explain how it
addresses the previously listed problems with quantizing gravity.
The starting point is that Plancks constant is a massless scalar field over space time
(x, t), and the equal time commutation relations for all fields, including Plancks
constant itself, are proportional then to (x, t). Since we have no experimental evidence for the variation of Plancks constant, the most conservative assumption is that
the -field is presently in its ground state, and difficult to excite with energies that we
have access to. This suggests that we think about quantization as the consequence
of a spontaneous symmetry breaking, and we have to add a suitable potential for
to the Lagrangian to achieve this. We are presently experiencing (x, t) as having a
non-zero vacuum expectation value that we will denote with 0 . This is the measured
value of Plancks constant. But at high temperature, presumably close by the Planck
energy, the symmetry can be restored, resulting in a classical theory.
Gravity and matter then have a quantized phase and an unquantized phase, and
are fundamentally neither quantized nor classical in the same sense that water is
fundamentally neither liquid nor solid. Quantization, in this case, is also not emergent
from a classical theory because the condition for second quantization does always
contain the (x, t).
222
S. Hossenfelder
A New Look at Old Problems

Let us now come back to the three problems mentioned in the first section that a
theory for quantum gravity should address.
First, there is the formation of singularities. We know of two types of singularities
that we should worry about, the Big Bang singularity and the singularities inside
black holes.
If we move backwards in time towards the early universe, the temperature of
matter increases and will eventually exceed the Planck energy. This is the standard
scenario in which symmetry restoration takes place [6], so the expectation value of
goes to zero, gravity becomes classical, and matter decouples. If matter decouples,
it cannot collapse to a singularity.
Collapse to a black hole is somewhat more complicated because its not a priori
clear that the temperature of the collapsing matter necessarily increases, but it plausibly does so for the following reason.1 If matter collapses to a black hole, it does so
rapidly and after horizon formation lightcones topple inward, so no heat exchange
with the environment can take place and the process is adiabatic. The entropy of the
degenerate Fermi gas is proportional to T n 2/3 , where T is the temperature and n is
the number density. This means that if the number density rises and entropy remains
constant, the temperature has to rise [7]. So again, matter decouples and there is
nothing left to drive the formation of singularities.
Note that the -field makes a contribution to the source term, necessary for energy
conservation.
Second, there is the black hole information loss. It was argued in [8] that the
problem is caused by the singularity, not the black hole horizon, and that removing
the singularity can resolve the information loss problem. This necessitates the weak
interpretation of the Bekenstein-Hawking entropy so that a stable or quasi-stable
Planck scale remnant, or a baby-universe, can store a large amount of information.
There are some objections to the existence of such remnants, but they rely on the
use of effective field theory in strong curvature regimes, the validity of which is
questionable [9]. Thus, unitarity in black hole evaporation can be addressed by the
first point, avoiding the formation of singularities.
Third, the difficulty of coupling a quantum field to a classical field and the nonrenormalizability of perturbatively quantized gravity. In the here proposed scenario,
there is never a classical field coupled to a quantum field. Instead, gravity and matter
are of the same type and together either in a quantum phase or a classical phase. In
the quantum phase, gravity is quantized perturbatively. It then needs to be shown that
the perturbation series cleanly converges for high energy scattering because is no
longer a constant. This is a subtle point and I can here only give a rough argument.
To see how this would work, first note that we can rewrite the equal time commutation relation into a commutation relation for annihilation and creation operators
of the fields. The commutator between annihilation and creation operators is then
1
I acknowledge helpful conversation with Cole Miller on this issue.
15 Gravity Can Be Neither Classical Nor Quantized
223
The same
proportional to the Fourier-transform of (x, t), which I will denote .
is true for the annihilation and creation operators of (x, t) (though the prefactors
differ for dimensional reasons).
Now consider an arbitrary S-matrix transition amplitude with some interaction
vertices. We evaluate it by using the commutation relations repeatedly until annihilation operators are shifted to the very right side, acting on the vacuum, which leaves
c-numbers, or the Feynman rules respectively. If Plancks constant is a field, then
every time we use the commutation relation, we get a power of the -field, and the
S-matrix expansion is a series in expectation value of powers of times the other
factors of the transition amplitudes. Then, we use the commutation relations on , or
its annihilation and creation operators respectively. Now note that exchanging two of
Thus, we can get rid of the expectation value of powthese will only give back one .
ers, so that in the end we will have a series in powers of vacuum expectation values
of (as opposed to a series of expectation values of powers, note the difference).
If we consider the symmetry breaking potential to be induced by quantum
corrections at low order, the transition to full symmetry restoration may be at a finite
value of energy. In this case then, the quantum corrections which would normally
diverge would cleanly go to zero, removing this last problem with the perturbative
quantization of gravity.
Summary
I have argued that the fundamental theory can be neither classical nor quantized, but
that quantization may be a phase that results from spontaneous symmetry breaking.
Needless to say, this proposal is presently very speculative and immature. Some more
details can be found in [10], but open questions remain. However, I hope to have
convinced the reader that giving up the assumption that a theory is either classical or
quantized can be fruitful and offers a new possibility to address the problems with
quantum gravity.
References
1. S.W. Hawking, Particle creation by black holes. Commun. Math. Phys. 43, 199220 (1975)
2. K. Eppley, E. Hannah, The necessity of quantizing the gravitational field. Found. Phys. 7, 5165
(1977)
3. J. Mattingly, Why Eppley and Hannahs thought experiment fails. Phys. Rev. D 73, 064025
(2006) [gr-qc/0601127]
4. D. Giulini, A. Grossardt, Gravitationally induced inhibitions of dispersion according to the
Schrdinger-Newton equation. Class. Quant. Gravity 28, 195026 (2011). arXiv:1105.1921
[gr-qc]
5. J.R. van Meter, Schrodinger-Newton collapse of the wave function, Class. Quant. Gravity.
28, 215013 (2011). arXiv:1105.1579 [quant-ph]
6. J.I. Kapusta, Finite-temperature Field Theory (Cambridge University Press, Cambridge, 1993)
224
S. Hossenfelder
7. D.S. Kothari, Joule-Thomson effect and adiabatic change in degenerate gas. Proc. Natl. Inst.
Sci. India 4, 69 (1938)
8. S. Hossenfelder, L. Smolin, Conservative solutions to the black hole information problem,
Phys. Rev. D 81, 064009 (2010). arXiv:0901.3156 [gr-qc]
9. J.F. Donoghue, When effective field theories fail. PoS EFT 09, 001 (2009). arXiv:0909.0021
[hep-ph]
10. S. Hossenfelder, A possibility to solve the problems with quantizing gravity. arXiv:1208.5874
[gr-qc]
Chapter 16
Weaving Commutators: Beyond

Fock Space
Michele Arzano
Abstract The symmetrization postulate and the associated Bose/Fermi (anti)commutators for field mode operators are among the pillars on which local quantum
field theory lays its foundations. They ultimately determine the structure of Fock
space and are closely connected with the local properties of the fields and with the
action of symmetry generators on observables and states. We here show that the
quantum field theory describing relativistic particles coupled to three dimensional
Einstein gravity as topological defects must be constructed using a deformed algebra of creation and annihilation operators. This reflects a non-trivial group manifold
structure of the classical momentum space and a modification of the Leibniz rule for
the action of symmetry generators governed by Newtons constant. We outline various arguments suggesting that, at least at the qualitative level, these three-dimensional
results could also apply to real four-dimensional world thus forcing us to re-think the
ordinary multiparticle structure of quantum field theory and many of the fundamental
aspects connected to it.
Introduction
Quantum field theory (QFT), the theoretical framework at the basis of our understanding of particle physics, lays its foundations on a set of fundamental assumptions
whose raison detre is intimately related with the existence of a fixed and highly
symmetric background space-time. When gravity enters the quantum stage one is
faced with a series of conceptual tensions which are the basis of the formidable challenge that the formulation of a quantum theory of geometry and matter has posed
to theoretical physicists in the past eighty years [1]. The extent of this tension is
dramatically evident already in the most celebrated effect in semiclassical gravity:
black hole quantum radiance. In this context a free quantum field living on a black
hole background produces a steady thermal emission of quanta from the horizon.
M. Arzano (B)
Dipartimento di Fisica, Sapienza University of Rome, P.le A. Moro 2,
00185 Roma, Italy
225
226
M. Arzano
Assuming an initial pure quantum state for the system, after a crude implementation
of back-reaction, such evaporation would end with a final mixed state thus violating
the basic postulate of unitarity of quantum evolution [2].
This blatant paradox, a quantum phenomenon which predicts a violation of one
of the principles of quantum theory itself, forces us to pass under scrutiny all the tacit
assumptions that enter the derivation of the effect. Factorization of the Hilbert space of
the quantum field states described by a Fock space is essential in the characterization
of field modes inside and outside the horizon (which in turn is closely related to
locality and microcausality) and is at the basis of the assumption that the use of
low-energy effective field theory is reliable in the derivation of such effect [3]. In
this essay we will argue, without making any assumptions about the nature of a yetto-be-formulated theory of quantum gravity, that three-dimensional semiclassical
gravity in the presence of the simplest form of topological back-reaction leads
to the demise of the usual formulation of Fock space. In particular, multiparticle
states will no longer be constructed from (anti)-symmetrized tensor product of oneparticle states but by a coloured generalization of them reflecting a deformed
algebra of creation and annihilation operators. Newtons constant (Plancks mass
in three dimensions) enters as a deformation parameter which governs the nonLeibniz action of translation generators on the quantum multiparticle states. Such
deformation is a consequences of the non-trivial group manifold structure of the
momentum space of the corresponding classical particle which is coupled to gravity
as a topological defect. Such unconventional quantization of the field modes signals a
departure from several assumptions at the basis of ordinary quantum field theory from
additivity of quantum charges associated to space-time symmetries to departures
from locality. The fact that ordinary QFT is smoothly recovered once the Newtons
constant/deformation parameter is set to zero suggests that these models can be
regarded as a natural extension of the conventional field theoretic paradigm which
might open new avenues in attacking the quantum gravity problem from below.
Curved Momentum Space in Flatland

As it is very well known Einstein gravity in three space-time dimensions does not
admit local degrees of freedom [4]. Point particles are coupled to the theory as topological defects [5]. The space-time describing a single particle will be flat everywhere
except at the location of the particle where one has a conical singularity. Indeed the
length of a circular path centered at the location of the particle divided by its radius
will be less than 2. The deficit angle is = 8Gm, proportional to the mass of the
particle m and Newtons constant G. For the description of the phase space of the
particle we need a characterization of its three-positions and three-momenta. This
can be achieved mapping the conical space-time into three-dimensional Minkowski
space with a cylindrical boundary and a wedge cut-off representing the deficit
angle of the cone [6].
16 Weaving Commutators: Beyond Fock Space
227
In ordinary (non-gravitational) relativistic mechanics in three-dimensions,

Minkowski space is isomorphic as a vector space to the Lorentz algebra and the
extended phase space is a vector space given by the direct product of two copies of
sl(2, R) i.e. sl(2, R) sl(2, R) R2,1 R2,1 . Positions and momenta are
parametrized by coordinates in such space i.e. x = x and p = p where a are
2 2 traceless matrices.
When the particle is coupled to gravity we can map the simply connected part of the
conical space-time to Minkowski space and thus coordinates will be given by a map
q from the cone to sl(2, R). To complete the embedding we need to map the local
frames at each point of the cone into a fiducial inertial frame on the Minkowski side.
Thus to each point, besides q, we also need to associate an element U SL(2, R)
which provides the information regarding the type of Lorentz rotation needed to
match the local frame to the background Minkowski frame. The pair (q, U) provides
an isometric embedding of the bundle of local frames on the simply connected part of
the manifold into Minkowski space. The freedom of choosing the background inertial
frame is reflected in the freedom of transforming the functions q and U via a global
Poincar transformation (n, U), respectively a translation and a Lorentz rotation.
Likewise the values of the embedding functions on the two faces of the wedge q
and U should be identified via a generic Poincar (gauge) transformation which we
denote by (v, P).
The three-position of the particle i.e. the location of its worldline in the auxiliary
Minkowski space is given by the values of the function q on the cylindrical boundary
which regularizes the singularity at the tip of the cone. Such function, which we
in principle depends on time and on an angular variable however, if we
denote by q,
want the cylindrical boundary to look like a worldline, we must impose the additional
condition that q depend only on time [6] i.e. q x(t). Thus the three-positions of
the particle will be still given by a vector, namely the coordinates of an element
of sl(2, R). The values of the worldline function q at the left and right boundary
of the wedge, q , are subject to the matching condition q + P1 (q v)P.
Since q + = q = x(t) is the location of the particle, taking the derivative with
respect to time of the equation above we obtain that the velocity of the particle
must commute with the group element P (remember that v is constant). This implies
that three-momentum vectors have to be proportional to the projection of the group
element P SL(2, R) on its Lie algebra sl(2, R), i.e. if we write P in its matrix
expansion P = u1 + 4Gp , we discard the part of P proportional to the identity
and take p = p . Notice that now the components of the momentum vector
are coordinates on a group manifold, indeed the condition det P = 1 implies that
u2 16 2 G2 p 2 = 1, the equation of a hyperboloid embedded in R4 . The phase space
in the presence of topological gravitational backreaction is thus G = sl(2, R)
SL(2, R) R2,1 SL(2, R).
In the following sections we will discuss the dramatic consequences that this
structural modification of the phase space of a relativistic particle has for the
corresponding (quantum) field theory.
228
M. Arzano
From Conical Particles to Braided Commutators

One Particle
We turn now to the description of the field theory describing the quantization of the
relativistic particle coupled to three-dimensional gravity discussed in the Introduction. Before we do that it will be useful to make a short digression on the definition
of the physical phase space and the associated mass-shell relation. In order to determine the mass shell we need to find a characterization of the mass of the particle and
relate it to the notion of generalized momentum. As we saw above the mass of the
particle is proportional to the deficit angle of the conical space. A way to measure
the deficit angle is to transport a vector along a closed path around the boundary, as
a result this will be rotated by the angle = 8Gm. Physical momenta will be thus
characterized by holonomies P which represent a rotation by = 8Gm. Such
requirement imposes the restriction
sin2 (4Gm)
1
Tr(P 2 ) = cos(4Gm) p 2 =
,
2
16 2 G2
(16.1)
on the physical holonomies giving us a deformed mass-shell condition. From a

mathematical point of view such on-shell condition is equivalent to imposing that
physical holonomies/momenta lie in a given conjugacy class of the Lorentz group
(see [7] for a pedagogical discussion). Roughly speaking if in ordinary Minkowski
space the mass-shell hyperboloid describing the physical momenta of a massive
particle is given by an orbit of the Lorentz group in flat momentum space, physical
momenta belonging to a given conjugacy class can be seen as exponentiated orbits
of the group on its manifold.
In analogy with ordinary Minkowski space we can consider complex functions
on the momentum group manifold described above. It turns out that when restricted
to momenta belonging to a given conjugacy class, such space of functions carries
a unitary irreducible representation of the semi-direct product of the momentum
group manifold and the Lorentz group [8] and thus is analogous (modulo a choice
of polarization [9]) to the ordinary one-particle Hilbert space for a quantum field.
Without loss of generality and to keep our considerations at the simplest level we
now switch to Euclidean signature allowing the phase space of the particle to be
G = su(2) SU(2). In analogy with ordinary field theory we consider plane waves
labelled by group elements belonging to a given conjugacy class as representatives of
a one-particle wave function. Momenta are now coordinates on SU(2), in particular
we work with cartesian coordinates1
P(p) = p0 1 + i
p
,
For simplicity we restrict to functions on SO(3) SU(2)/Z2 .
(16.2)
229

2
where = (4G)1 , p0 = 1 p 2 and are Pauli matrices. Plane waves can be
written in terms of a Lie algebra element x = x i i su(2) as
i
eP (x) = e 2 Tr(xP) = eipx .
(16.3)
Tr(P
). The main effect of the group structure of momentum space is
with p = 2i
that the composition of plane waves is non-abelian indeed we can define a -product
for plane waves
i
eP1 (x) eP2 (x) = e 2 Tr(xP1 ) e 2 Tr(xP2 ) = e 2 Tr(xP1 P2 ) ,
(16.4)
differentiating both sides of this relation one can easily obtain a non-trivial commutator for the xs
[xl , xm ] = ilmn xn ,
(16.5)
i.e. the coordinates xi can be seen as equipped with a non-commutative algebra

structure. Functions of these coordinates will inherit a non-abelian algebra structure
and the corresponding field theory will be a non-commutative field theory. Most
importantly momenta will obey a non-abelian composition rule due to the non-trivial
group structure
1
1
p1 p2 = p0 (p2 ) p1 +p0 (p2 ) p2 + p1 p2 = p1 +p2 + p1 p2 +O(1/2 ). (16.6)
Since plane waves are eigenfunctions of translation generators the non-abelian composition of momenta will correspond to a non trivial action of translation generators
on multiparticle states, in particular one can easily derive the following generalization
of the Leibniz rule on the tensor product of two one-particle states
Pa = Pa 1 + 1 Pa +
1
abc Pb Pc + O(1/2 ).
(16.7)
Notice that 1 = 4G can be seen as a deformation parameter and in the limit

one recovers the usual action of translations as an abelian Lie algebra. As discussed in
[10] such behaviour signals a departure from one of the basic postulates of quantum
field theory namely the additivity of the charges associated with space-time symmetry
generators which in turn is deeply connected with the locality properties of the field
operators [11].
230
M. Arzano
More Particles
So far we have seen that at the one particle level the mass-shell condition which in
the ordinary case is given by orbits of the Lorentz group on the vector space R3,1
is replaced now by orbits of the Lorentz group on a non-linear momentum space
or, more properly, conjugacy classes. In analogy with ordinary field theory let us
label one particle states by elements of these conjugacy subspaces of SU(2) and
denote them by |P. As for any quantum system the space of states of a composite
object is built from the tensor product of its constituents Hilbert spaces. Since at the
quantum level we are dealing with indistinguishable particles one postulates [12]
that n-particle states are constructed from (anti)-symmetrized n-fold tensor products
of one-particle Hilbert spaces for particles with (half)-integer spin. Focusing on the
simplest case of a two-particle state one notices that naively adopting a standard
symmetrization prescription (we assume
for definiteness that we are dealing with
a spinless particle) the candidate state 1/ 2 (|P1 |P2 + |P2 |P1 ) is not an
eigenstate of the translation operator due to the non-abelian composition of momenta
(reflecting the non-trivial Leibniz action of translation generators discussed above).
This problem can be bypassed if one resorts to the following momentum dependent
symmetrization [13, 14]

|P1 P2 L 1/ 2 |P1 |P2 + |P1 P2 P11 |P1 .
(16.8)
We used the subscript L above because one could also choose

|P1 P2 R 1/ 2 |P1 |P2 + |P2 |P21 P1 P2 .
(16.9)
Notice how now both two-particle states are eigenstates of the generators Pa and
have total on-shell momentum P1 P2 p1 p2 . In analogy with the standard case
we can introduce creation and annihilation operators so that
aL,R
(P1 )aL,R
(P2 )|0 |P1 P2 L,R .
(16.10)
The action of the Lorentz group on the kets will be given by conjugation [15] i.e.
H |Pi |H1 Pi H.
(16.11)
It is straightforward to check that both left and right symmetrizations above

are covariant under such transformations [14]. Moreover L and R-symmetrized
states are connected by a Lorentz transformation
(P1 P2 ) |P1 P2 L = |P1 P2 R .
(16.12)
231
In order to determine the algebra satisfied by such operators we start by noticing the
useful relation
(16.13)
(P2 1 P1 ) 1 |P1 P2 L = |P2 P1 L .
Defining R1
L (P1 , P2 ) (P2 1 P1 ) 1 we can then write the following braided
commutators
aL (P1 )aL (P2 ) R1

L (P1 , P2 )aL (P2 )aL (P1 ) = 0
aL (P1 )aL (P2 ) RL (P1 , P2 )aL (P2 )aL (P1 ) = 0.
(16.14)
(16.15)
One can proceed in an analogous way for the right operators to find similar commu1
tation relations with R1
L (P1 , P2 ) replaced by RR (P1 , P2 ) 1 (P1 P1 ).
2
The cross-commutators between a(P) and a (P) will be similarly [14] given by
L (P2 )a (P2 )aL (P1 ) = (P1 P2 )
aL (P1 )aL (P2 ) R
L
1
(16.16)
L (P2 ) = 1 P2 and (P1 P2 ) is the Dirac delta function on the group
where R
1
[16]. One can proceed in an analogous way for the R operators and obtain
R (P2 )a (P2 )aR (P1 ) = (P1 P2 )
aR (P1 )aR (P2 ) R
R
1
(16.17)
R (P1 ) = P1 1.
where now R
We arrived to a modification of the usual algebra of creation and annihilation operators which is quite suggestive. It is reminiscent of the algebra of q-deformed oscillators or quons [17] but with colour-dependent q-factors given by RL,R (P1 , P2 )
L,R (P). We leave it open to speculation whether such deformed commutators
and R
can be interpreted as the quantum counterpart of the braiding of the worldlines of
classical point-defects.
Discussion
The familiar form of the algebra of creation and annihilation operators that we are
accustomed to from quantum field theory textbooks is intimately related to the quantization condition one imposes on fields and their conjugate momenta. The latter is
assumed on the basis of the analogy with ordinary quantum mechanical commutators between position and momenta of a non-relativistic particle. The results we
presented show that Einsten gravity in three space-time dimensions clearly indicate
a possible relaxation of such assumption and a departure from the basic structures
underlying our familiar formulation of local quantum field theory. The most immediate consequence of the deformed algebra of oscillators, as we showed above, is that
232
M. Arzano
the Fock space of the theory loses its simple structure in terms of (anti)-symmetrized
tensor products of given one-particle states. It has been suggested [18] that these
types of departures from ordinary Fock space might reflect a new kind of uncertainty
on the measurement of momenta of multiparticle states namely that measuring the
total momentum of a system precludes complete knowledge of the total momenta of
its components and vice-versa. Besides this observation what is evident now is that
due to the braided nature of the multiparticle states the question of decoupling of
the low energy degrees of freedom form the high energy ones must be handled with
care. This could suggest a weak link in the assumptions at the basis of the derivation
leading to the information paradox, namely the use of low energy effective field
theory in the presence of backreaction.
Another key aspect that is put at stake in these models is locality. In the discussion
above we briefly touched upon the fact that the Leibniz action of symmetry generators
on quantum states is deeply connected with the local properties of the fields. It turns
out that allowing a non-trivial geometry for the momentum space of a classical particle has been subject to recent investigations in the context of the relative locality
paradigm [19]. The phase space of a particle coupled to three dimensional gravity
can indeed be seen as an example of a relative locality theory [20]. The conceptual breakthrough of such models lies in the observer-dependent notion of crossing
of particle worldlines. The far reaching implications of this new feature have been
widely discussed in the literature [21, 22]. In the perspective of our discussion it will
be useful to investigate the behaviour of field operators constructed via the deformed
operators above in order to check whether classical relative locality translates at
the quantum level into departures from the ordinary local field paradigm.
Of course all the discussion so far is very specific to three dimensional gravity
and its topological nature. What about the more realistic four-dimensional world?
Obviously in four space-time dimensions Einsteins gravity is not a topological theory
and thus in general similar arguments would not hold. Surprisingly though there exist
suggestive results on Planckian scattering in quantum gravity that appear to hint in
the right direction. Early work by Hooft [23] and by Verlinde and Verlinde in the early
90s [24] showed that forward scattering at Planckian center of mass energies in 3+1
quantum gravity can be treated semiclassically and gravity splits in a weakly coupled
sector and a strongly coupled sector whose quantum dynamics can be described by
a topological field theory. Could we be dealing with a similar state of affairs also in
this four dimensional regime? As of today the question remains open.
The recent framework of piecewise flat gravity in 3 + 1 dimensions [25] proposed
as a model for gravity which displays only a finite number of degrees of freedom per
compact regions of space-time could also provide a bridge to the real four dimensional
world. Indeed this model is based on a straightforward extension of the picture
of a system of particles described as defects which is found in three dimensional
gravity. To our knowledge nobody has attempted a study of the phase space of these
particles/strings in the same spirit of [6]. It would be not surprising if one would end
up finding non-trivial structures analogous to the ones we discussed in this essay.
Finally, following the relative locality framework mentioned above one could
argue that a curved momentum space is just a feature of a regime of four dimensional
233
quantum gravity in which the Planck length is negligible while the Planck mass
remains finite [22]. This formally means that both quantum and local gravitational
effects become negligible, while their ratio remains finite and governs the non-trivial
geometry of momentum space. If this assumptions are correct then our arguments
would qualitatively hold true in four dimensions and they would indicate that first
order quantum gravity corrections to local QFT would be exactly the kind described
above.
In our opinion and in the light of the observations above, large part of the conceptual apparatus of local QFT is ripe for re-thinking and the three dimensional world is
there to point us the way to go beyond the various assumptions that lie their roots in
the very structure of Minkowski space. What we find remarkable is that the simple
combination of ordinary classical gravity and quantum theory (via a topological coupling), without any reference to a specific quantum gravity model, suggests that
departures from local QFT become quite natural when gravity enters the game. This
suggests that the humble framework of semiclassical gravity has still a lot to teach
us on various puzzling aspects of the marriage between gravity and the quantum
world.
Acknowledgments I would like to thank J. Kowalski-Glikman and V. De Carolis for discussions.
This work is supported by a Marie Curie Career Integration Grant within the 7th European Community Framework Programme and in part by a grant from the John Templeton Foundation.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
S. Carlip, Rep. Prog. Phys. 64, 885 (2001). [gr-qc/0108040]

S.W. Hawking, Phys. Rev. D 14, 2460 (1976)
S.D. Mathur, Lect. Notes Phys. 769, 3 (2009). arXiv:0803.2030 [hep-th]
S. Carlip, (Cambridge University Press, Cambridge, 1998), p. 276
S. Deser, R. Jackiw, G. t Hooft, Ann. Phys. 152, 220 (1984)
H.J. Matschull, M. Welling, Class. Quant. Gravity 15, 29813030 (1998). [gr-qc/9708054]
B.J. Schroers, PoS QG -PH, 035 (2007). arXiv:0710.5844 [gr-qc]
T.H. Koornwinder, F.A. Bais, N.M. Muller, Commun. Math. Phys. 198, 157 (1998). [qalg/9712042]
M. Arzano, Phys. Rev. D 83, 025025 (2011). arXiv:1009.1097 [hep-th]
M. Arzano, Phys. Rev. D 77, 025013 (2008). arXiv:0710.1083 [hep-th]
R. Haag, J.T. Lopuszanski, M. Sohnius, Nucl. Phys. B 88, 257 (1975)
A.M.L. Messiah, O.W. Greenberg, Phys. Rev. 136, B248 (1964)
M. Arzano, D. Benedetti, Int. J. Mod. Phys. A 24, 4623 (2009). arXiv:0809.0889 [hep-th]
M. Arzano, J. Kowalski-Glikman, to appear
Y. Sasai, N. Sasakura, Prog. Theor. Phys. 118, 785 (2007). arXiv:0704.0822 [hep-th]
N.J. Vilenkin, A.U. Klimyk, Representation of Lie Groups and Special Functions1991, vols.
3 (Kluwer, 1993)
O.W. Greenberg, Phys. Rev. D 43, 41114120 (1991)
G. Amelino-Camelia, A. Marciano, M. Arzano, in Handbook of Neutral Kaon Interferometry
at a Phi-factory, vol. 22, ed. by A. Di Domenico (2007), pp. 155186
G. Amelino-Camelia, L. Freidel, J. Kowalski-Glikman, L. Smolin, Phys. Rev. D 84, 084010
(2011). arXiv:1101.0931 [hep-th]
234
M. Arzano
20. G. Amelino-Camelia, M. Arzano, S. Bianco, R.J. Buonocore, Class. Quant. Gravity 30, 065012
(2013) arXiv:1210.7834 [hep-th]
21. G. Amelino-Camelia, M. Arzano, J. Kowalski-Glikman, G. Rosati, G. Trevisan, Class. Quant.
Gravity 29, 075007 (2012). arXiv:1107.1724 [hep-th]
22. G. Amelino-Camelia, L. Freidel, J. Kowalski-Glikman, L. Smolin, Gen. Relativ. Gravity 43,
2547 (2011) [Int. J. Mod. Phys. D 20, 2867 (2011)]. arXiv:1106.0313 [hep-th]
23. G. Hooft, Phys. Lett. B 198, 61 (1987)
24. H.L. Verlinde, E.P. Verlinde, Nucl. Phys. B 371, 246 (1992). [hep-th/9110017]
25. G. Hooft, Found. Phys. 38, 733 (2008). arXiv:0804.0328 [gr-qc]
Chapter 17
Reductionist Doubts
Julian Barbour
Abstract According to reductionism, every complex phenomenon can and should

be explained in terms of the simplest possible entities and mechanisms. The parts
determine the whole. This approach has been an outstanding success in science, but
this essay will point out ways in which it could nevertheless be giving us wrong
ideas and holding back progress. For example, it may be impossible to understand
key features of the universe such as its pervasive arrow of time and remarkably high
degree of isotropy and homogeneity unless we study it holisticallyas a true whole.
A satisfactory interpretation of quantum mechanics is also likely to be profoundly
holistic, involving the entire universe. The phenomenon of entanglement already
hints at such a possibility (Somewhat more technical material that appeared as endnotes in my essay entry now appear as footnotes. I have also added some further
footnotes, identified by 2014 at their beginning, to bring my original essay up to
date, and an addendum at the end. In a few cases, I have made very minor changes to
the original text for the sake of clarity or to correct a type. These are not indicated.).
Reductionisms Strengths and Weaknesses

Nature does not begin with elements as we are obliged to begin with them. It is certainly
fortunate 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 left out of account (Ernst Mach, 1883)
To get an idea where reductionisms strong and weak points lie, lets go to its
source in Newtons worldview: utterly simple laws that govern the motions of bodies
in space and time.
The law of inertia is the most basic: a force-free body moves uniformly in a
straight line forever. Things are almost as simple if bodies interact. When far apart,
they move with near perfect inertial motions, but when closer they can, through forces
Julian Barbour (B)
College Farm, The Town, South Newington, Banbury, OX15 4JG, Oxon, UK
235
236
J. Barbour
like gravitation, begin to change each others motions. If body A has a greater mass
than body B, then A affects B more than B does A. With allowance for the masses,
action and reaction are equal.
The behaviour of a large system of many bodies, in principle the whole universe,
can, according to Newton, be entirely explained by the inherent tendency of each
body to move inertially modified merely by the forces exerted by all the other bodies
in the universe. Just as reductionism proposes, the whole truly is the sum of its parts.
Or is it?
The weak spot in Newtons scheme is the very thing that makes it reductionist.
The position and motion of individual bodies such as atoms, the simple entities, are
defined relative to invisible space and time, the framework in which simple laws can
be formulated. Mach [1] and others argued that in reality the position of any one object
is defined relative to every other object in the universe. That is obviously a far more
complicated approach and it is clearly holistic. Mach argued that nevertheless it could
still reproduce all of Newtons successes because force-free bodies are observed to
move rectilinearly relative to the stars, which suggests that, in their totality, they
exert a powerful causal effect on individual objects. Newton could have mistaken
this empirical fact as evidence for what he called absolute space.
It is easy to see how this could have led to seriously wrong ideas about the universe.
Barely over 100 years ago, most scientists thought we lived in an island universe
of a few million stars (our Galaxy) with nothing but space outside it. According to
Newtons laws, such an island universe could exist and have angular momentum,
L, about an axis, making it oblate like the rapidly rotating Jupiter. This was to be
expected.
However, a simple implementation of Machs ideas [2] rules it out. A Machian
island universe must have angular momentum exactly equal to zero: L = 0, even
though subsystems of bodies in small regions of the Galaxy can behave as Newtons
laws predict and have L = 0. It is merely necessary for the values of L for all the
subsystems to add up to zero. Seeing Newtons laws confirmed in the solar system,
which does have L = 0, astronomers had no reason to question any of the predictions
of Newtonian theory, which in no way is capable of forcing an island universe to
have L = 0.
But suppose astronomers had, on Machian grounds, been convinced around 1900
that the universe must have L = 0 and had found this was not so for the Galaxy.
Far from giving up the Machian theory, they would have predicted that our Galaxy
cannot be the entire universe. They would have predicted a universe with more matter
whose angular momentum exactly balances the Galaxys.
How does the universe we can now observe meet this expectation? Extraordinarily
well. It is full of galaxies and clusters of galaxies. Each and every one manifestly
possesses some angular momentum, but there is no indication that the individual
angular momenta add up to give a significant non-zero angular momentum of the
universe as a whole.
In fact, my example is too simplistic, being based on Newtonian theory and not
its superior successor in the form of Einsteins general relativity. But I think it makes
my point. A reductionist standpoint may be very misleading.
17 Reductionist Doubts
237
Einstein, Mach, and General Relativity

Machs idea that Newtons invisible absolute space could be replaced by an effect
of the entire universe had a deep influence on Einstein, who called the idea Machs
principle [3]. It was the single greatest stimulus to the creation of his wonderful
theory of gravity, general relativity (GR).
If GR implements Machs principle, reductionism will be challenged. Simple
parts and their interactions will not determine the whole; the whole will determine
the way the parts behave. In fact, we shall see that the whole to a considerable degree
determines what we perceive as parts.
So does GR implement Machs principle? This has been a matter of controversy
ever since Einstein created GR, mainly because he set about implementing Machs
principle in an indirect manner that left its status in GR obscure [4]. In fact, his
definition of Machs principle [3] was inconsistent [4], p. 93. I have given what
I believe is the correct definition of Machs principle [5] and argued that if the
universe is closed up in three dimensions like the earths surface in two then GR
does implement Machs principle [5]. If the universe is spatially infinite, the answer
is equivocal. It is Machian however far you can imagine, but infinity is unreachable,
and one can never establish a complete sense in which the whole determines the part.
In the previous section I showed how reductionism can mislead if we conceive
the universe as a mere sum of its parts, but it did not suggest any need to change
the conception of the parts, the individual stars, but only the way they interact. The
Machian interpretation of general relativity is much more sophisticated and changes
radically the way we conceive the parts and not just the way they interact. In this
section I show how the standard representation of GR presents the universe as a sum
of parts. In the next section, I give the alternative Machian interpretation, in which
GR appears much more holistic. In the final section, I discuss possible implications
of a holistic quantum view of the universe.
By parts in GR I mean little (infinitesimal) regions of spacetime. Imagine a twodimensional surface that, at each point, is smooth and curved in accordance with
a definite local law: measurements of the curvature in the immediate vicinity of
every point on the surface would confirm the law holds at that point. Any surface for
which this is true may exist. This is a reductionist situation: the parts (infinitesimal
elements of surface) and the local law they satisfy determine what wholes (extended
two-dimensional surfaces) can exist.
In general relativity, in the simplest situation in which no matter is present, the
infinitesimal elements have not two but four dimensions: one of time and three
of space. Otherwise the situation is very similar to what I described. Infinitesimal
regions of spacetime satisfy Einsteins famous field equation R = 0, and his theory
permits any spacetime in which this local law is satisfied everywhere. Presented in
these terms, it is a great triumph of reductionism; the predictions of GR have been
very well confirmed for subsystems of the universe (the solar system and binary
pulsars) and are not in conflict with cosmological observations. But in fact these
present numerous puzzles whose solution may well call for a truly holistic approach
238
J. Barbour
that restricts the solutions GR allows and puts them in a different perspective. I now
turn to that.
Gravity, Angles, and Distances

In the previous section, I identified the parts in GR as infinitesimal pieces of spacetime. Their structure is determined by the fundamental quantity in GR: the metric
tensor g . Being symmetric (g = g ) it has ten independent components, corresponding to the four values the indices and can each take: 0 (for the time
direction) and 1, 2, 3 for the three spatial directions. Of the ten components, four
merely reflect how the coordinate system has been chosen; only six count. One of
them determines the four-dimensional volume, or scale, of the piece of spacetime, the
others the angles between curves that meet in it. These are angles between directions
in space and also between the time direction and a spatial direction.
Now Machs attitude to physics was not only holistic but also that all theoretical
concepts should stay as close as possible to directly observable quantities. Using
grand philosophical terms, the gap between epistemologywhat can be observed
and ontologywhat is assumed to existshould be as small as possible. Ideally,
there should be no gap at all. From this perspective, Einsteins ontology as expressed
through the metric tensor g can be questioned on two grounds.
First, time is taken to be just as real as space. But Mach argued that time cannot
be a primary notion; it must be derived from motion. In the first of these fqxi essay
competitions, I showed how Newtons notion of an independent absolute time can
be derived from motion. In Einsteins theory, Newtons global absolute and nondynamical time is replaced by local proper time, which interacts with space. This is a
major change, but time, like space, is still taken to be a primary ontological concept,
not something to be derived from more basic concepts.
Second, there is a big difference between, on one hand, quantities with scale,
like lengths and volumes, and, on the other, angles: the former cannot be specified
without the human choice of some unitsay metres or yards. That is arbitrary. In
contrast, an angle can be specified in dimensionless terms as a fraction of a radian,
which itself is a dimensionless quantity.
Let me spell this out to underline the Machian aspect. Suppose yourself at the
centre of a circle of radius r looking at two points on the circumference separated
by r . The angle that you see between them is by definition one radian. Key here is that
the angle is the same whatever the radius r . This brings out the difference between
angles and lengths.
Now imagine yourself on a clear night at high altitude in a desert. What do you
see? Thousands of stars studding the black sky, all with definite angles between them.
239
Its beautiful. The scientific point is that you directly see the angles between the stars.
But you cannot tell their distances.1
There are two possible reactions to the difference between dimensionful and
inherently dimensionless quantities like angles. The standard one in science is to
say that it is not a big deal; one must simply express everything in terms of ratios.
Astronomers, for example, express interplanetary distances as ratios, taking the semimajor axis of the Earths orbit as unit. Moreover, these distances have dynamical
effect: Newtons gravitational force is inversely proportional to the square of the
distance.
The alternative attitude to dimensionful quantities is to deny them any fundamental
role. Is this feasible? I think so. Note first that, despite being defined through ratios
of distances, angles can be truly said to belong to a point. The radius of the circle
taken to define a radian can always be taken as small as one likes. In contrast, a
distance is necessarily associated with two points, the ends of the considered interval.
In the standard formulation of GR, angles and distances both have ontological
dynamicalstatus. Distances do because the curvature of spacetime plays a central
role in GR, and curvature is defined by comparisons over distances even though they
can be taken arbitrarily small.2
Curvature, and with it distance, is so deeply embedded in the conceptual and
technical foundations of GR it would seem only a mad man would try to deny it a
deep dynamical significance. General relativity without curvature seems like Hamlet
without the Prince of Denmark. Remarkably, a new formulation of GR, called shape
dynamics [68], suggests that, in a well defined sense, it is only angles that count.
Because it shows how reductionism could be misleading us, I will sketch how the
significance of angles emerged.
Dirac [9] and Arnowitt, Deser, and Misner (ADM) [10] took the first step over 50
years ago by replacing Einsteins vision of spacetime as a four-dimensional block
universe by a picture in which a three-dimensional entity evolves, building up spacetime in the process. This is known as the ADM formalism. The hope, still held by
many, is that the DiracADM approach will one day lead, as its creators intended,
to a quantum theory of gravity.
I need to say what the DiracADM three-dimensional entity is. It is a Riemannian
three-geometry (named after the great mathematician Riemann, who introduced the
concept). A two-geometry is easy to visualize: it is like the two-dimensional curved
surface of the Earth, which closes up on itself. A closed geometry is needed to model
a universe as a whole. A closed three-geometry is much harder to imagine, but is
mathematically possible. In the ADM formalism, angles and infinitesimal distances
are, dynamically, on an equal footing.
However, a hint that angles might have priority emerged from subsequent work
related to the technically important ADM initial-value problem. There is no need
1
Even when distances are determined by parallax, it is changes of observed angles that determine
the distances, which are themselves ratios of the trigonometric base length.
2 Technically, curvature involves second derivatives of the metric tensor, whereas angles are defined
by the undifferentiated metric tensor.
240
J. Barbour
for me to give details except to say that without its solution it is simply impossible
to make any practical use of GR to establish, for example, what happens when two
black holes coalesce. Among other things, this work, by James York [11] with input
from his student Niall Murchadha [12], established the amount of information that
is needed to determine a physically sensible solution of GR.
The required information can be expressed in terms of a 3-geometry and the rate
at which it is changing. As with spacetime, at each point of a 3-geometry there is
encoded angle and scale information, 2 and 1 numbers respectively (compared with
5 + 1 for four-dimensional spacetime). The rates at which these are changing are
also characterized by 2 + 1 numbers. York showed that the purely angle part of this
information, i.e., 2 + 2 = 4 numbers at each point, is sufficient to solve the initial
value problem and hence determine a complete spacetime that satisfies Einsteins
equations.
This suggested that GR is a theory of the way angle data change and that distance
data play no dynamical role. However, this possibility was never taken too seriously
because Yorks technique seemed to violate a foundational principle of GRthat
there is no distinguished definition of simultaneity in the universe. For this reason,
Yorks work, which does single out a notion of simultaneity, has usually been regarded
as a merely technical device, albeit very valuable, for solving problems but not settling
issues of principle.
Shape Dynamics
To what extent does shape dynamics (SD), which I mentioned earlier and for which
the above has been a preparation, change this?
Imagine a toy model universe of just three particles in two dimensions. Picture
them as dots on an infinite sheet of grid paper. Then two coordinates determine the
position of each. The six coordinates together define a Newtonian configuration, the
sheet playing the role of absolute space. But one can ignore the positions of the dots
on the sheet and regard as real only the three distances between each particle pair.
That defines a Machian relative configuration. Defined in its own intrinsic terms,
it does not exist anywhere in space. The final step is to say only the shape of the
triangle, defined by two of its angles, is real.
The progressive elimination of non-Machian data takes us from Newtonian configurations through relative configurations to shapes. Now comes a key fact. Mathematically it is vastly easier to describe change in terms of coordinates, at the level
of Newtonian configurations. That is why Newton introduced absolute spaceand
with it reductionism. Because shapes are irreducibly holistic, it is much harder to
work with them and achieve the gold standard of dynamical theory: the ability to
determine the future from an initial state of the system. In Newtons dynamics, a
configuration and its rate of change define an initial state. In shape dynamics, we
241
(a)
(b)
Fig. 17.1 a An arbitrary relative positioning of two triangles determines an initial Newtonian
configuration (vertices of the grey triangle) and its rate of change (length of the arrows to the
vertices of the dashed triangle). In the best-matched positioning b, shifting, rotating and scaling
has brought the dashed triangle as close as possible to perfect overlap (congruence) with the
grey one. An appropriate mathematical expression [5] measures the residual incongruence, the
determination of which depends only on the shapes, not on their sizes or positions in space. The
procedure is holistic, since the shapes alone determine the outcome, which is moreover expressed
tractably through the coordinates of the vertices of the dashed triangle
want a shape and its rate of change to determine an initial state.3 Thats where the
problem, the passage from reductionism to holism, lies. It is solved by best matching,
which is explained in Fig. 17.1.
Once the shape-dynamic problem of defining initial states has been solved, it is
easy to determine how shapes change according to a law that in no way relies on
absolute duration, on position in an frame like grid paper, or on an external scale that
measures size. The idea can be applied very generally, in particular to Riemannian
geometries from which all information except that relating to angles between curves
has been abstracted. Then we are dealing with shapes of closed geometries, just as
York did, but now with some differences.
First, SD has led to an improved understanding of Yorks technique. This is a
technical matter that I will relegate to a footnote4 for experts.
3 I need to say something about what rate of change means in shape dynamics. It is not with respect
to any external time but relates to the expansion of the universe, which cosmology indicates is a
reality. In the context of shape dynamics, this introduces a single overall scale ratio: one can say
the universe is twice as large now as it was in an earlier epoch. This scale ratio, which is why shape
dynamics is based on volume-preserving conformal transformations, has nothing to do with angles
but provides a single global parameter that defines the rate of change of the angles. Thus, the angles
do not depend on time but on the scale ratio. I do find the need for a scale ratio mysterious. Perhaps
it has to be there to provide a substitute for time. 2014: I discuss this further in the Addendum.
4 York based his technique, for a spatially closed universe, on full conformal transformations (which
change local sizes but leave the angle information unchanged), whereas SD is based on volumepreserving conformal transformations. This is a tiny restriction, but it explains a bizarre feature
of Yorks method that seemed completely ad hoc. I am referring to the scaling law that York adopted
242
J. Barbour
Second, and relevant for us, York did not solve the initial-value problem following
any clearly formulated first principles but by exploring different mathematical techniques until he found one that worked, supremely well in fact. In contrast, SD derives
not only Einsteins equations but also Yorks method for solving the initial-value
problem on the basis of clearly formulated Machian first principles, of which there is
a holy trinity: duration, position, and size must all be relative. The first two of these
were foundational principles of Einsteins original derivation,5 but the third was not.
In the shape-dynamic reformulation of GR, Einsteins relativity of simultaneity is
traded for relativity of size.
The generalization of best matching to shapes of geometries is direct but so far
reaching it is worth outlining. One starts from 2+2 given pieces of angle information
at each point and considers how one could add to them in all possible ways 1 + 1
pieces of trial information that specify distances and their rates of change. The extra
1 + 1 pieces of information are then varied in all possible ways until best matching
is achieved. The mathematics is sophisticated but conceptually fully analogous to
adjusting the dashed triangle to the best-matched positioning in Fig. 17.1. As in that
case, there is always just one unique way, for given 2 + 2 angle information, in
which the best-matching condition can be satisfied. One then has an initial state that
determines the future uniquely.6
Prior to the best matching, there is nothing in the specification of the initial
2 + 2 pieces of angle information at each point of space that determines the distance between nearby points, the rate at which time flows at each space point, or
the path of an inertially moving body. What I find really remarkable is that best
matching applied to the distance-free 2 + 2 pure angle information leads to complete
determination of distances, rates of time flow, and inertial motion everywhere.7 It is
a triumph of Machs intuition, going far beyond his original tentative suggestions.
Best matching is profoundly holistic. The procedure based on it that determines
how the extra 1 + 1 pieces of distance information are added at each point has to take
into account the angle information everywhereat all points. What is fixed here is
determined by everything that is there.8 There is a remarkable delicate interdependence of everything.
(Footnote 4 continued)
for the trace part of the momentum in the LichnerowiczYork equation, which seemed incompatible
with the law adopted for the trace-free part. SD provides a simple explanation for the law [6].
5 2014: I see that what I wrote originally was a bit confusing. Relativity of simultaneity, rather than
of duration, was a foundational principle for Einstein. However, the latter is a consequence of the
former in conjunction with Einsteins other assumptions. The key difference in SD is then, as stated,
that relativity of simultaneity is traded for relativity of size as a foundational principle.
6 My characterization of shape dynamics emphasizes its conceptual aspects. My collaborators
Henrique Gomes, Sean Gryb, and Tim Koslowski also use the name shape dynamics for a specific
gauge-theoretical implementation of the underlying ideas in a form that is likely to be best suited
for technical developments.
7 Soon after its completion, Yorks work suggested to Isenberg and Wheeler [13] a formulation of
Machs principle in somewhat similar terms but without a clear shape-dynamic underpinning.
8 Best matching in the case of shapes of geometries involves the solution of elliptic partial differential equations. As York and Murchadha showed, these have very good existence and uniqueness
243
You may well ask: has anything been achieved, or is SD merely a different way
of interpreting GR? The real test will be whether SD helps in the creation of a
quantum theory of gravity. I shall consider that in the next section. It is however
noteworthy that GR allows a great number of solutions that seem physically most
implausible; this has long been a concern. The ADM formalism, if regarded as the
true form of the theory, already considerably restricted the allowed solutions. Shape
dynamics goes significantly further. It requires the universe to be compatible with
generation from York initial data and to be spatially closed. This is what the truly
holistic approach requires. In this respect (in postulating a spatial closure), it runs
counter to much current thinking in cosmology. It is always good for a theory to
make bold predictions and live dangerously.
The main aim of this section has been to show that a reductionist approach can
give a seriously misleading intuition about what the world is made of and how it
works. In certain key respects, the standard representation of GR remains close to
Newtons world view. Time and space are fused together, but locally their ontological
nature and inertial motion in them are barely changed. Distance in space and duration
in time are both real and foundational. Shape dynamics suggests that only angles are
real.9
Quantum Implications
Consider the archetypal quantum two-slit experiment in which, in Diracs famous
words, each photon interferes with itself. This implies a particle that somehow
splits in two and manages to go through both slits at once. This obviously stretches
the imagination.
Could a reductionist mindset be misleading us in trying to interpret quantum
experiments in terms of particles moving in space and time? Its a very Newtonian
picture. We see the Moon but never a photon. All we see is equipment and an event
that we attribute to a photon. We see a photon-generating laser, a filter to ensure that
properties. In the case of spatially closed universes, there are no boundary conditions to be specified;
everything is determined intrinsically and holistically.
9 I should add two caveats here. First, the discussion so far has ignored matter. That can be treated
by Yorks method and hence added to shape dynamics without difficulty. Indeed it adds conceptual
clarity to the picture. This is because matter fields define curves in space. The angles between any
two curves are real (ontological), but distances along the curves are not; they are gauge. Second,
all relativists would agree that a spacetime is determined by the specification of 2 + 2 pieces of
information at each point in a three-geometry. However, a majority would probably deny that these
pieces of information must be exclusively and exhaustively angle data; they could be a mixture of
angle and distance data. If relativity of simultaneity is taken to be sacrosanct, that is undeniable but
it leaves one with a very indeterminate situation. What, if any, mixture of angle and distance data is
correct? Three arguments speak for pure angle data: angles are conceptually more fundamental, the
choice is clean (no indeterminate mixture), no other general and robust method apart from Yorks
pure-angle method has been found to solve the initial-value problem.
244
J. Barbour
only one photon at a time reaches the slits, and then the screen on which spots appear
one after another.
The observed events are certainly discrete, but are we right to try to interpret them
in terms of point-like particles? We do so because the experiment results in those
spots on the screen or some other localized responses in apparatus. Of course, Bohr
above all emphasized the complementary nature of quantum phenomena: particlelike or wavelike features are manifested depending on the macroscopic arrangement
of the experiment. Change the arrangement and complementary aspects of reality are
manifested. One can measure either momentum or position of a particle but never
both at once.
This led Bohr to insist on the impossibility of formulating any meaningful statements about quantum phenomena without the framework provided by the classical,
non-quantum world of macroscopic instruments. In this view quantum theory is not
about what is but about what we can say.
But perhaps quantum theory is about what is and the problem is that reductionism
has led us to the wrong picture of what is. Instead of thinking of particles in space
and time, we should perhaps be thinking in terms of complete shapes of the universe.
Then a photon would not be a particle that moves from one place to another but a
law telling us how one shape of the universe becomes another. After all, that is what
happens in the two-slit experiment. A spot appears on a screen that was not there
before. In the first place, this tells us the shape of the universe has changed. Of course,
it is only one change among an immense multitude, but it is a part of the totality of
change. We may be doing violence to the universeMachs overpowering unity of
the Allby supposing individual events are brought about by individual things, by
parts. Hard as it may be, I think we need to conceptualize in terms of wholes.
The Machian interpretation of GR gives us hints of how this could change things.
It suggests that what we observe locally has a part that is really herein the first
place angle informationand a framework and basic laws that seem to be here
local distances, times and inertial motionbut are in reality the effect of the complete
universe.
In many ways, Bohr had a Machian mindset: the interpretation of quantum events
depends crucially on the relationships that define the experimental layout. But where
does that end and what determines it? Quantum entanglement measurements are
now performed between the Canary Islands and the west coast of Africa and would
be impossible without clocks that measure photon travel times. But if the Machian
interpretation of GR is correct, the very distances over which travel times can be
measured and even the times themselves are in a real sense created by the universe.
Thus, I think Bohr was on the right track but that one needs to go much further
along it. Where will it take us? I dont know but will hazard a conjecture or two.
First, the quantum mechanics of the universe, like the Machian interpretation of
GR, is holistic: it is about shapes. However, in contrast to shape dynamics, in which
one shape follows another, in the quantum universe all shapes are present with
different probabilities. But then whence comes our tremendously strong impression
that time passes and we have a unique history?
245
In The End of Time [14], I attempted to answer this question with the notion of a
time capsule: a very special configuration of the universe whose structure suggests
that it is the outcome of a process that has taken place in time in accordance with
definite laws. In [14] I still took distance to be real. The ideas that later led to
shape dynamics were only briefly mentioned (in Box 3). I now distinguish carefully
between relative configurations, in which both distances and angles are fundamental,
and shapes defined by angles alone.
In [14] I conjectured a timeless quantum competition between all possible configurations of the universe in which time capsules get the greatest quantum probabilities.10 We experience them as instants of time bearing every appearance of the
transition from a definite past to an indefinite future. I justified this conjecture by the
pronounced asymmetry of what I called Platonia, the collection of all possible relative configurations of the universe. But with distance removed from the foundations,
we are left with shapes. Platonia is shape space.
In this view, the quantum universe is doubly holistic: the possible shapes of the
universe are wholes, each a unity in itself, but there is also shape space, the higher
whole formed by the totality of possible wholes. Shape space is the ultimate quantum
arena.11 That is my second conjecture.
Seen from a gods eye perspective, the striking asymmetry of Platonia that I
noted in [14] becomes even more pronounced and suggestive: in any space of shapes
there is always one unique shape, appropriately called Alpha, that is more uniform
than any other possible shape.12
It interesting to note in this connection that the theory of cosmological inflation
has some significant achievements to its credit but nevertheless must assume, as a
10 2014: As I will explain in the Addendum below, I still believe that the concept of time capsules
is correct, but am now inclined to believe that one can speak meaningfully of evolution of the wave
function of the universe.
11 In the ADM approach to quantum gravity, the wave function of the universe for vacuum gravity
is defined on superspace, the space of Riemannian three-geometries. The problem with this arena
is the supreme difficulty, unresolved despite over 50 years of attempts, of implementing the ADM
Hamiltonian constraint. I suspect that the root of the problem is the indeterminate mixture of angle
and distance information encoded in superspace, which, as noted in the main text, is unavoidable as
long as relativity of simultaneity (refoliation invariance) is held to be sacrosanct. In shape dynamics,
the wave function of the universe is defined on conformal superspace. This brings undoubted conceptual clarity, as York already pointed out, but severe problems still remain. The ADM Hamiltonian
may be conceptually hybrid but it is at least local; the shape-dynamic Hamiltonian is non-local.
12 2014: During work on the paper [15], my collaborators Tim Koslowski and Flavio Mercati
pointed out to me that this statement is not true for shape space, on which a very natural measure of
uniformity (and, simultaneously, complexitysee footnote 13) puts a plateau-like structure within
which there are narrow infinitely deep potential wells. A uniquely distinguished peak Alpha is not
present. Fortunately, my slip in the essay does not affect my argument. Indeed, in [15] we present
encouraging support for it. In an addendum at the end, I comment on the main results of [15], which
follow already from the plateau-like structure.
246
J. Barbour
hitherto unexplained initial condition, that the presently observed universe emerged
from a state that was highly isotropic and homogeneous: in a word, very uniform.
This suggests to me, as my third conjecture, that we should not be looking to explain
the isotropy and homogeneity by a special initial condition but through the dominant
status of Alpha in the universes shape space. This is reflected in the fact that every
shape has a degree of complexity that ranges from the most uniform possible to
shapes of ever increasing complexity.13 One never reaches Omega.
If there is no Omega in shape space, there must still be an end to this essay. Let
it be my final conjecture, which sharpens the one in [14]. Our experiential life is
dominated by time and a great asymmetry: the transition from past to future, from
birth to death. Science has so far failed to explain the asymmetry and has been forced
to attribute it to a specialhighly uniforminitial state of the universe. I suspect we
need look no further than the structure of shape space. Its highly uniform Alpha,14
which holistic shape dynamics tells us must be deeply significant, seems to match
perfectly what we take to be the uniform initial state of the universe.
Mathematics, the tool of theoretical physics, can only describe becoming through
differences of structures that simply arethey exist in a timeless Platonic realm.
Becoming is the ever mysterious moving reflection of being. We cannot explain the
indubitable asymmetry of becoming that we experience unless being is asymmetric.
It is.
Addendum (2014)
Since this essay was completed in 2012, my collaborators and I have had several
insights that, I believe, significantly strengthen the plausibility of the conjectures
made in my essay. These are set out in detail in [15], which builds on [17, 18]. The
first key insight, achieved in [17, 18], relates to the nature of time and its relation
to scale. It was in part anticipated in the submitted essay in footnote 3 and can be
explained in simple Newtonian terms.
Consider N point particles with masses m i and suppose they represent an island
universe. If we treat it in Machian shape-dynamic terms as suggested in the text, we
13
Complexity is a surprisingly difficult notion to capture in a definition. However, in the case

of the Newtonian N -body problem, there is a uniquely appropriate measure of the complexity
of an N -body system that I call the scale-invariant gravitational potential Vsi . It is the product
of the ordinary
Newtonian potential and the square root of the centre-of-mass moment of inertia
Icm := (1/M) i< j m i m j ri2j , where ri j is the distance between particles i and j and M is the total
mass. It is a function on shape space and takes its maximum on the most uniform shape the system
can have. For large N , the corresponding distributions of the particles are remarkably uniform
[16]. As the distributions become more clustered, Vsi takes ever larger negative values. Given the
close analogy between the Newtonian N -body problem when recast in Machian terms and GR,
I anticipate the existence of a function on conformal superspace (the shape space for GR) that is
analogous to Vsi .
14 2014: In line with the correction noted in footnote 12, this should be replaced by Its striking
plateau-like structure.
247
would like to insist that only the shapes formed by the set of points have dynamical
significance. Within Newtonian theory, this is almost the case if, as is reasonable, we
insist that the universe as a whole has vanishing total energy and angular momentum.
Then position and orientation in space of the universe as a whole have no meaning.
However, if one sticks with the Newtonian gravitational potential energy, it is not
possible to eliminate completely the role of scale. At a given instant, it has no physical
significanceit would be changed by the human convention of measuring in inches
rather than centimetres. However, a ratio of sizes at different instants, which is a
dimensionless number and independent of the choice of units, does have meaning.
Now recall what I said about looking at the stars: you can directly see angles
between stars but not their individual distances. This is closely related to my conviction, implicit in the distinction that I made between angles and distance, that shape
is something far more fundamental than size. One sees this already in the case of a
triangle, for which its shape is intuitively different from its size and entirely independent of any units of measurement. What is particularly striking is that one needs
3N 7 numbers to fix the shape of N points embedded in three-dimensional space
but only one to fix its size, which, moreover, depends on the unit of measurement.
One can have arbitrarily many shape degrees of freedom, but only one size degree of
freedom. In the N -body problem, the Newtonian theory of N point particles interacting gravitationally, there is a (mass-weighted) measure of size that is uniquely
distinguished by the dynamics. It is the (centre-of-mass) moment of inertia I .
In [15], my collaborators and I draw attention to a fact discovered well over a
century ago by Lagrange, which is this. If the total energy E of the system is nonnegative, E 0, then (except in very special cases that need not concern us here)
the curve of I (as a function of the Newtonian time) is U-shaped upward. It passes
through a minimum and tends to infinity in both time directions. Now suppose you
start at infinity on one side of the U. It has a slope, we call it D in [15], which is
negative and infinitely large in magnitudethe curve points vertically downward.
As you move further down the U, the slope (with allowance for its sign) steadily
increases from until it reaches the bottom of the U, where it passes through
zero, D = 0, after which it goes on increasing all the way to + at the top of the
other side of the U. Thus, the slope of D is always increasing. Moreover, if you track
the behaviour of D going in the opposite direction, you find exactly the same thing:
with allowance for its sign, the slope of D always increases. To find such behaviour
of a degree of freedom in any dynamical system is remarkable. It is given a special
name and called a Lyapunov variable.
In [18], we draw attention to an almost uncanny similarity between the time t
in Newtonian theory and this behaviour of D. Both are invisible, and their role in
dynamics can only be inferred from the behaviour of things that we can see. Moreover,
both t and D are monotonic: they either always increase or decrease.15 In fact, in
15
Of course, one normally thinks that t always increases, but Newtons equations take the same
form if one reverses the direction of time, which corresponds to reversing all the velocities in a
given solution. The form of the law does not imprint a direction of time on the solution. Newtons
laws, like Einsteins, are time-reversal symmetric.
248
J. Barbour
[18] we argue that, if one is considering a complete Newtonian universe, a variable

related to D should be regarded as time and not t. It is particularly interesting
that in general relativity there is no analogue of Newtonian time but there is a close
analogue of D called the York time. It emerges naturally from the method, discussed
in Sect. Gravity, Angles, and Distances, by which York solved the initial-value
problem in GR.
The recognition of the special status of D and its strong similarity to Newtons
time might have great significance for the creation of quantum gravity, which has
long been plagued by the so-called problem of time. This arises because quantum
mechanics seems to need a well-defined time variable with properties like Newtons t.
But in general relativity as created by Einstein with relativity of simultaneity taken as
a fundamental postulate it is not possible to identify such a time. The case is altered
if one is prepared, as we do in shape dynamics, to trade relativity of simultaneity for
relativity of size. In the classical theory, this does not lead to any conflict with known
observations, but it could have far-reaching implications in the quantum theory. In
particular, if the York time is identified as the time that quantum gravity seems to
need, one would no longer have a frozen wave function of the universe and face the
acute problem of how to recover our vivid experience of the passage of time.16
For me, one of the most interesting things about the U-shaped curve of the moment
of inertia is the light it may cast on the origin of the arrow of time. As noted in footnote
15, the known laws of nature do not distinguish a direction of time. However, the
great bulk of the actual processes that we observe taking place in the universe exhibit
a marked directionality [19]. It is particularly striking that qualitatively different
kinds of processes all have arrows that point in the same direction. Moreover, all
observations suggest that the direction is the same everywhere in the universe and
at all times in its history. It has been a mystery for well over a century how this
behaviour can arise from time-symmetric laws.
The most common suggestion is known as the past hypothesis, which is that at
the big bang the universe was created in a special state of extremely low entropy. If
this was so, then there would be no conflict with the second law of thermodynamics, according to which entropy always increases. It started very low and has been
increasing ever since. Such a scenario would also be compatible with the growth of
structure that we see everywhere on the earth and in the universe: there is no conflict with the second law provided the decrease of entropy associated with growth of
structure in one region is offset by growth of entropy elsewhere.
The past hypothesis relies on a fundamental fact about physical laws: they all
allow a great many different solutions. In a laboratory, the solution one actually
encounters is determined by some initial (and boundary) conditions. There is, thus,
16
Readers familiar with my earlier work, especially The End of Time [14], will see that I have
moved from advocacy of the position that there is nothing in the external universe corresponding
to change and the passage of time to serious consideration of the possibility that difference of scale
could play a role somewhat like, but certainly not identical to, Newtons time. While it is a bit
embarrassing to make such a major change, I am happy that it came about through systematic study
of the role of scale in gravitational dynamics very much in line with relational principles of the kind
advocated by Leibniz and Mach.
249
a dual explanation for any given phenomenon: the underlying law and an initial
condition. The past hypothesis implies that one of the most all-pervasive phenomena
we observe both near and far in the universethe arrow of timeis not a consequence
of the law that governs the universe but of a very special initial condition. Why such
an initial condition should have been realized is a complete mystery. This is hardly
a satisfactory state of affairs, and in [15] we question it.
Our key observation relates to the fact that effectively all solutions of the
N -body problem with zero energy and angular momentum divide into the two halves
either side of the bottom of the U-shaped curve described above. Reflecting the timereversal symmetry of Newtons law, there is qualitative symmetry of each solution
about the central point. However, when one examines the halves in detail, they are
quite different. Moreover, in all the typical solutions one finds that near the bottom
of the U, where D = 0, the motion is basically chaotic but that with increasing
distance from D = 0 structures, especially orbiting pairs of particles, tend to form.
This is clearly reflected in the behaviour of the complexity/uniformity function that
is defined in footnote 13 as a measure of how structured the system of particles is.
Its value fluctuates but grows steadily between bounds. If, as is very natural, one
defines a direction of time to match the growth of structure, then time increases in
both directions away from the point D = 0. One can say that each solution consists
of two histories that have a single common past, at D = 0, but separate futures, at
the two infinite tops of the U. Moreover, as we argue in [15], any observer living
in such a solution must be in one of the halves and will take that to be effectively
a single history which emerges from a relatively uniform but rather chaotic initial
state.
It is far too soon to claim that our observation will help to crack the great problem
of the origin of the arrow of time. However, to us at least, it does suggest that the law
which governs the universe might well, by virtue of its form and without any special
initial conditions, create manifest arrows of time for observers within the universe.
References
1. E. Mach. Die Mechanik in ihrer Entwicklung Historisch-Kritisch Dargestellt (1883)
2. J. Barbour, B. Bertotti, Machs principle and the structure of dynamical theories (downloadable
from platonia.com). Proc. R. Soc. Lond. A 382, 295306 (1982)
3. A. Einstein, Prinzipielles zur allgemeinen Relativittstheorie. Annalen der Physik 55, 241244
(1918)
4. J. Barbour, H. Pfister (eds.), Machs Principle: From Newtons Bucket to Quantum Gravity,
Einstein Studies, vol. 6 (Birkhuser, Boston, 1995)
5. J. Barbour, Machs principle: a response to Mashhoon and Wessons paper Annalen der Physik
(2011). arXiv:1106.6036, arXiv:1108.3057
6. E. Anderson, J. Barbour, B.Z. Foster, B. Kelleher, N.. Murchadha, The physical gravitational
degrees of freedom. Class. Quant. Gravity, vol. 22:17951802, (2005), gr-qc/0407104
7. J. Barbour, Shape dynamics: an introduction. in Quantum Field Theory and Gravity. Proceedings Conference at Regensburg 2010, ed. by F. Finster et al. (Birkhuser, 2012)
8. H. Gomes, S. Gryb, T. Koslowski, Einstein gravity as a 3D conformally invariant theory. Class.
Quant. Gravity 28, 045005 (2011) (arXiv:1010.2481)
250
J. Barbour
9. P.A.M. Dirac, Fixation of coordinates in the Hamiltonian theory of gravitation. Phys. Rev. 114,
924930 (1959)
10. R. Arnowitt, S. Deser, C.W. Misner, The dynamics of general relativity, in Gravitation: An
Introduction to Current Research, ed. by L. Witten (Wiley, New York, 1962), pp. 227265
11. J.W. York, The role of conformal 3-geometry in the dynamics of gravitation. Phys. Rev. Lett.
28, 10821085 (1972)
12. N.. Murchadha, J. York, Existence and uniqueness of solutions of the Hamiltonian constraint
of general relativity on compact manifolds. J. Math. Phys. 4, 15511557 (1973)
13. J. Isenberg, J.A. Wheeler, Inertia here is fixed by mass-energy there in every W model universe,
in Relativity, Quanta, and Cosmology in the Development of the Scientific Thought of Albert
Einstein, ed. by M. Pantaleo, F. deFinis (Johnson Reprint Corporation, New York, 1979), pp.
267293
14. J. Barbour, The End of Time (Oxford University Press, Oxford, 1999)
15. J. Barbour, T. Koslowski, F. Mercati, A gravitational origin of the arrows of time.
arXiv:1310.5167 [gr-qc]
16. R. Battye, G. Gibbons, P. Sutcliffe, Central configurations in three dimensions. Proc. Roy. Soc.
Lond. A459, 911943 (2003)
17. J. Barbour N.. Murchadha, Conformal superspace: the configuration space of general relativity. arXiv:1009.3559
18. J. Barbour, T. Koslowski, F. Mercati, The solution to the problem of time in shape dynamics
(Accepted for publication in Class. Quant. Gravity 2014). (2013) arXiv:1302.6264 [gr-qc]
19. H.D. Zeh, The Physical Basis of the Direction of Time (Springer, New York, 2007)
Chapter 18
Rethinking the Scientific Enterprise:

In Defense of Reductionism
Ian T. Durham
Abstract In this essay, I argue that modern science is not the dichotomous pairing
of theory and experiment that it is typically presented as, and I offer an alternative
paradigm defined by its functions as a human endeavor. I also demonstrate how
certain scientific debates, such as the debate over the nature of the quantum state,
can be partially resolved by this new paradigm.
The Scientific Enterprise

I have begun to enter into companionship with some few men who bend their minds to the
more solid studies, rather than to others, and are disgusted with Scholastic Theology and
Nominalist Philosophy. They are followers of nature itself, and of truth, and moreover they
judge that the world has not grown so old, nor our age so feeble, that nothing memorable
can again be brought forth. Henry Oldenberg, as quoted in [4].
Science is a living, breathingand very humanenterprise. As such, it has always

been a malleable process. Indeed, that is one of its enduring traits: not only does
science prescribe a system by which its predictions may be refined by additional
knowledge, but its very nature changes as our understanding of the world and ourselves broadens. Nevertheless, there is an over-arching paradigm to modern science
whose origins are rooted in the works of Alhazen1 who flourished during the Islamic
Golden Age, circa 1000 CE. In its simplest form, this paradigm consists of the posing
of questions and the subsequent testing of those questions [12]. This process is, of
course, cyclic, as the testing of the original questions very often leads to new ones.
But the asking of a question is really at the root of all scientific endeavors and stems
from humanitys innate curiosity about itself and the world around us. In a sense, we
all remain children, continually asking why? In more modern scientific terms, the
act of questioning forms the basis of a scientific theory that is a well-substantiated
Abu Ali al-H.asan ibn al-H.asan ibn al-Haytham (965 CEc. 1040 CE), also known as Ibn
al-Haytham and sometimes al-Basri.
1
I.T. Durham (B)

Department of Physics, Saint Anselm College, Manchester, NH 03102, USA
251
252
I.T. Durham
explanation of some aspect of the natural world, based on a body of facts that have
been repeatedly confirmed through observation and experiment [13]. In other words,
Alhazens paradigm breaks science into two equal parts: theory and experiment.
While there have been modern refinements to Alhazens basic framework, notably
the adoption of the hypothetico-deductive2 model, the basic division into theory and
experiment remains. Victoria Stodden has recently proposed that computational science be recognized as a third division and, indeed, this is an attractive suggestion [14].
But it would fail to address certain persistent problems with both theory and experiment that raise deeper questions about the overall methodology of science. Clues to
a solution to these problems can be found in the origins of that methodology.
While a precise formulation of the history of modern scientific methodology is
not only lengthy but somewhat subjective, it is generally agreed that the revolution
it sparked began in 17th century Europe and many of its principles were codified
in the early documents and practices of the Royal Society of London, arguably
the worlds oldest scientific organization3 [4]. As Thomas Sprat wrote, the Royal
Societys purpose was not the Artifice of Words, but a bare knowledge of things
expressed through Mathematical plainness [4]. This early scientific community
developed a highly mechanistic approach to science that, while applied with equal
vigor to anything tangible (and thus encompassing the modern fields of astronomy,
chemistry, biology, physiology, et. al.), was decidedly grounded in the physical. The
modern field that we recognize as physics has been called the most fundamental
and all-inclusive of the sciences [5]. Arguably a portion of that inclusivity stems
from the fact that all the other sciences are constrained by physical laws. This is one
way in which scientific reductionism can be interpreteda reduction of the other
sciences to physics. But physics is also inclusive by dint of its methods. Physics,
throughout its history, has hewn most closely to the mechanistic approach developed
in the 17th century and, indeed, this is the other way in which scientific reductionism
is traditionally interpreteda reduction of a system to its constituent parts in an
effort to better comprehend the whole.
This interpretation of reductionism is closely related to the notion of causality
and, as a view of science, has been challenged in recent years as a result of work on
emergence and complex systems [1, 9, 10, 16]. As Jonah Lehrer4 wrote in a recent
article
[t]his assumptionthat understanding a systems constituent parts means we also understand
the causes within the systemis not limited to the pharmaceutical industry or even to biology.
It defines modern science. In general, we believe that the so-called problem of causation can
be cured by more information, by our ceaseless accumulation of facts. Scientists refer to
2
The term hypothetico-deductive has been attributed to William Whewell, though evidence for
this is lacking as the term does not appear in any of his works on the inductive sciences.
3 The history of the Royal Society is tightly linked with a number of organizations that arose
in the mid17th century including Acadmie Monmor, the Acadmie des sciences, and Gresham
College [4].
4 The ideas for the present essay were in large part developed as a rejoinder to Lehrer prior to his
resignation from the New Yorker after admitting to fabricating quotes. That incident should have
no bearing on what is written and discussed here.
18 Rethinking the Scientific Enterprise: In Defense of Reductionism
253
this process as reductionism. By breaking down a process, we can see how everything fits
together; the complex mystery is distilled into a list of ingredients [10].
Lehrers article, however, focused almost exclusively on a single aspect of scientific

methodology that is not necessarily mechanistic and that is misunderstood, even by
scientists themselves: statistics and mathematical modeling. If reductionism is indeed
what Lehrer claims it is, then statistical methods and mathematical modeling are
most definitely not reductionist since they only seek to find mathematical structures
that explicitly match existing data. This point is perhaps the most misunderstood in
all of science. As an example, consider first the relationship between statistics and
probability.
Mathematical Models
Statistics often accompanies probability (at least in textbook titles and encyclopedia
entries). But this belies a subtle but important difference between the two. Both are
indeed disciplines in their own right that fall under the larger umbrella of mathematics and logic. But only statistics is an actual tool of science. Probability is a
logico-mathematical description of random processes. Statistics, on the other hand,
is a methodology by which aggregate or bulk information may be analyzed and
understood. It loses its meaning and power when applied to small sample sizes.
And theres the rub. If reductionism is the act of breaking down a process in order
to understand its constituent parts, as Lehrer claims, statistics is the antithesis of
reductionism because it makes no such effort.
Why then do we stubbornly persist in thinking that statistical methods in science
can masquerade as some kind of stand-in for reductionism? Why do we expect
more from statistics than we have a right to? Statistics is a veryveryimportant
tool in science, but it is often misapplied and its results are often misinterpreted.
Few understood this better than E.T. Jaynes. Jaynes spent the better part of his career
attempting to correct one of the more egregious misconceptions, one that is intimately
related to the difference between probability and statistics.
Roughly speaking, statistics generally describe information we already know or
data weve already collected, whereas probability is generally used to predict what
might happen in the future. As Jaynes astutely noted, if we imagine data sampling
as an exchangeable sequence of trials,
the probability of an event at one trial is not the same as its frequency in many trials;
but it is numerically equal to the expectation of that frequency; and this connection holds
whatever correlation may exist between different trials The probability is therefore the
best estimate of the frequency, in the sense that it minimizes the expected square of the
error [7].
In other words, probabilities can only be accurately formulated from statistical data if
that data arose from a perfectly repeatable series of experiments or observations. This
is the genesis of the interpretational debate over the meaning of the word probability,
254
I.T. Durham
with the frequentists on one side claiming a probability assignment is really nothing
more than an assignment of the frequency of occurrence of a given outcome of a
trial, and the Bayesians on the other side claiming that a probability assignment is a
state of knowledge. As Jaynes clearly notes, the frequency interpretation is only valid
under strictly enforceable conditions whereas the Bayesian view is more general.
What does the Bayesian interpretation of probability tell us about reductionism?
The key to the Bayesian interpretation is the notion that, if probabilities represent our
states of knowledge, measurements update these states of knowledge. Thus knowledge is gained in an incremental manner5 which is the essence of reductionism. Thus
probabilities, in a Bayesian context, are absolutely reductionist. As Jaynes points out,
it is possible to give probabilities a frequentist interpretation, in which case they connect to the more aggregate descriptions provided by statistics, but only under certain
strict conditions.
All of this does not necessarily obviate the need for the broader generalizations
provided by statistics. In fact, as the foundational basis for thermodynamics, statistics
as understood in the sense of distributions of measured quantities, has been very
successful in explaining large-scale phenomena in terms of the bulk behavior of
microscopic processes. Similar arguments can be made in terms of fluid dynamics,
atmospheric physics, and like fields. As Jaynes pointed out,
[i]n physics, we learn quickly that the world is too complicated for us to analyze it all at
once. We can make progress only if we dissect it into little pieces and study them separately.
Sometimes, we can invent a mathematical model which reproduces several features of one of
these pieces, and whenever this happens we feel that progress has been made [8], [emphasis
added].
Thus statistics is one of the primary methods by which larger-scale patterns are
discovered. These patterns as such emerge in aggregate behavior from the underlying pieces. However, it is wrong to assume that such patterns can emerge completely independently of the underlying processes. This is tantamount to assuming
that macroscopic objects can exist independently of their underlying microscopic
structure. The melting of an ice cube clearly refutes this notion.
Of course, very few true anti-reductionists would argue this fairly extreme
view. Instead they argue an intermediate position such as that proposed by P.W.
Andersen [1]. Andersen fully accepts reductionism, but argues that new principles
appear at each level of complexity that are not merely an extension of the principles
at the next lower level of complexity. In another words, Andersen is suggesting that
were we to be endowed with a sufficiently powerful computer and were we to have
a full and complete understanding of, say, particle physics, we still would not be
able to derive a human being, for example, or, at the very least, the basic biological
laws governing human beings. Biology and chemistry, to Andersen, are more than
just applied or extended physics. This is precisely the point Lehrer is trying to make.
But there is at least one fundamental problem with this argument: it assumes that no
amount of additional knowledge can bridge the gap between levels of complexity,
5
This is not necessarily the same thing as sequential, as is clearly demonstrated by certain quantum
states.
255
i.e. it takes as a priori that reductionism (or constructionism, as Andersen calls it)
is either wrong or incomplete. But this is logically unprovable. As Carl Sagan wrote,
[y]our inability to invalidate my hypothesis is not at all the same thing as proving
it true [11]. In fact, this is precisely the same argument that proponents of creationism and intelligent design employ in claiming the universe (and, in particular, the
biological life therein) is too complex to arise from simpler, less complex rules [2].
This does not mean that there arent fundamental limits to our ability to know the
universe. For example, Gdels incompleteness theorems place an inescapable limit
on our ability to mathematically describe any physical system representable in an
axiomatic manner [6]. Consider two physical systems, X and Y , each independently
described by the same set of mathematical structures, M, that we take to be the minimum set that fully describes each system. Now suppose that completely combining
these physical systems gives rise to a third physical system, Z , that is described by a
set of mathematical structures, N , where M = N . We assume that N is taken to be
the minimum set of structures that fully describes Z . In this scenario, X and Y are
more fundamental than Z and thus M must necessarily be a more restrictive set of
structures than N . If M and N are formally distinct then Gdels theorems tell us
that there will exist truths in M, for instance, that cannot be proven from within M
itself. Likewise for N . Thus it might be that N cannot be derived from M alone. In
fact, it implies that there are structural elements of N that cannot be derived from any
more primitive set of structures. Is this evidence for the anti-constructionist view?
Not necessarily. While this is a valid limit to our knowledge, it only applies to any
complete axiomatic description of the systems M and N . The universe as a whole
may not be fully axiomatic. In fact, in the history of science, axiomatization was
largely the realm of deductivism. But science is equal parts deductive and inductive
and Gdels theorems say nothing about inductive methods. In other words, the limits
on our knowledge apply to certain methods. As yet, there does not appear to be a
known limit to all methods. Thus it may be more instructive to think about science
in terms of methodologies.
A New Scientific Paradigm

Recall that Alhazens paradigm breaks science into two equal parts: theory and experiment. In this paradigm, experiments describe the universe and theories explain
it. In this light, consider the development of Newtonian gravity in the 17th century.
We can assign Galileo the role of experimenter/observer for his work with falling
bodies, bodies on an inclined plane, and his observations of the moons of Jupiter, the
latter of which importantly showed that celestial objects could orbit other celestial
objects aside from the earth. This final point emphasizes the fact that a full theory of
gravity had to take into account the movement of celestial bodies as well as terrestrial
ones. Where, then, in this historical context, can we place Kepler? The data used by
Kepler in the derivation of his three laws of planetary motion was largely taken by
Tycho Brahe. They were not explained until nearly six decades after Keplers death
256
I.T. Durham
(in 1630) when Newton published his Philosophi Naturalis Principia Mathematica
in 1687.6 Thus, Kepler was neither the one who performed the original observations
nor was he the one who discovered the explanation for the patterns exhibited by the
observational data. He was, in fact, performing precisely the same general function
as statisticians, climate scientists, and anyone performing clinical drug trials: he was
fitting the data to a mathematical structure; he was modeling. This is neither theory
nor experiment. It is a methodology for ascertaining how the universe works. In a
sense, it is a function.
To some extent we have, as scientists, successfully ignored this problem for four
centuries largely because it didnt seem to matter. After all, the dichotomy of theory
and experiment was only a rough guide anyway and didnt have much of an impact
(if any) on the science itself. But now, in certain areas of science and particularly
in physics, this dichotomy does not appear to be working as it should. The most
obvious example of this may be quantum mechanics where we have more than
a centurys worth of reliable experimental data, a well-established mathematical
structure fit to that data, but no universally agreed upon interpretation of this data and
its mathematical structure. Conversely, with string theory we have a well-established
mathematical structure and a generally agreed-upon theory, but no data. In climate
science, on the other hand, we have a consensus theory concerning climate change
and we have a vast amount of experimental data, but we have no universally agreed
upon mathematical model taking all of this data into account (i.e. we havent reduced
climate change to a self-contained set of equations yet). These examples appear to
suggest that Stodden is on the right track in suggesting that there is a third division
to science.
But how would adding a third division of science to the usual two solve the problems raised by Lehrer, Andersen, and others? To answer this question, let us first
re-examine the purpose of each divisions methods. What is it that experimentalists
are really doing? Are they actually describing the universe or is their aim something
else? I would argue that the aim of experimental science is, in fact, not to merely
describe the universe. Even Aristotle described the universe. What Aristotle didnt
do was describe it in a precise and consistent manner. His interpretation of what he
saw had to fit pre-conceived philosophical notions. The revolution that marked the
advent of modern experimental science aimed at measuring quantities free from preconceived notions of what those quantities should be. In other words, experimental
science does not describe things, it measures things. Inherent in this aim is precision since measurement without precision is meaningless. Achieving a measure of
precision itself requires repeatabilityexperimental results must be repeatable and
independently verifiable. In fact, this latter point is so crucial that it is often more
important for experimentalists to describe their procedures as opposed to their data.
The data will often speak for itself but the procedure must be comprehensible if it is
to be repeated and verified.
6 Robert Hooke famously claimed priority in the formulation of the inverse square law, but, as
Alexis Clairaut wrote in 1759 concerning this dispute, there is a difference between a truth that is
glimpsed and a truth that is demonstrated (quoted and translated in [3]).
257
The aim of theory, on the other hand, has always been to explain the world around
us and not merely to describe it. What sets modern theoretical science apart from
Aristotelianism and other historical approaches is that it aims for logical selfconsistency with the crucial additional assumption that science, as a whole, is ultimately universal. This last point implies that all of science is intimately connected.
Thus we fully expect that biological systems, for example, will still obey physical
and chemical laws. Crucially, modern theoretical science also aims to predict the
future behavior of systems. Thus a good scientific theory is both explanatory as
well as predictive.
Description, then, is the realm of mathematics. Mathematics is ultimately how
we describe what we see in the experimental data. However, since mathematics
is such an integral part of science, neither theorists nor experimentalists can carry
out their work entirely free of it. It is this all-pervasive nature of mathematics that
then leads to confusions and mis-attributions of the kind argued by Lehrer as well as
interpretational problems vis--vis probability theory and its relation to statistics. As
we noted earlier, roughly speaking, statistics generally is applied to prior knowledge
(collected data) whereas probability theory is predictive in nature. As such, statistics
is generally descriptive whereas probability theory is predictively explanatory. Thus
I would argue that some of these issues could be cleared up if, rather than thinking
of science in the way Alhazen did, perhaps with the added third division suggested
by Stodden, we instead should think of science as being divided into three functions:
measurement, description, and predictive explanation. These functions, of course,
are the essence of reductionism.
Now consider the rather sticky example of quantum mechanics which appears
to be lacking a single, unifying interpretation (i.e. theory in the sense we have
discussed above). In our parlance, it would seem that there are multiple predictive
explanations that exist for quantum mechanics. But, in fact, most of the differences
in the various interpretations of quantum mechanics differ in their interpretation of
the quantum state. Thus consider a generic quantum state,
| = c1 |1 + c2 |2 .
If we interpret this statistically, then the values c1 and c2 are arrived at only by making
repeated measurements. Instead, we can interpret this as a state of knowledge about
the system that can be updated with a subsequent measurement. In other words, it
can be interpreted as being predictive, at least in a probabilistic sense. On the other
hand, if we take the state to be ontological, then it actually exists in the form given
by | and thus the state is merely descriptive. Thus these three interpretations of
the quantum state correspond exactly to the three functions of science and, when
viewed in that light, do not necessarily contradict one another. Perhaps, instead
of requiring no interpretation, as Brukner has suggested [15], quantum mechanics
actually requires multiple interpretations.
What does this suggested shift in the description of science imply for complexity
and emergence? If science is to be considered universal, connective, and
self-consistent, perhaps the problem is not that reductionism is a broken paradigm,
258
I.T. Durham
but rather that we are mis-ascribing some of our activities to the wrong scientific
function, e.g. perhaps some of our so-called theories are actually more descriptive
than predictively explanatory. Or perhaps theyre built on the wrong description.
Either way, without a formal proof that reductionism is incapable of certain descriptions of nature, it would seem a bit premature to declare one of the most successful
methods of human inquiry dead. In fact it may simply be that, since the time of
Alhazen, we have simply been missing a key ingredient. In order to maintain science
as a productive, respected, and vital discipline we must ensure that it remains true to
its foundational functions while always allowing room for introspection. Otherwise,
science risks being ignored and too much is at stake for us to let that happen.
Acknowledgments The ideas discussed in this essay were tested on a few unsuspecting audiences
over the course of a little more than a month. Thus, for helping me feel my way through these
ideas, I would like to thank the following for giving me a pulpit from which to preach: the Clemson
University Symposium for Introduction to Research in Physics and Astronomy (SIRPA); the Kennebunk Free Librarys Astronomy Nights, co-hosted by the Astronomical Society of Northern
New England; and the Saint Anselm College Philosophy Club.
References
1. P.W. Andersen, More is different: broken symmetry and the nature of the hierarchical structure
of science. Science 177(4047), 393396 (1972)
2. P. Atkins, Atheism and science, in The Oxford Handbook of Religion and Science, ed. by
P. Clayton, Z. Simpson (Oxford University Press, Oxford, 2006)
3. W.W. Rouse Ball, An Essay on Newtons Principia (Macmillan and Company, London, 1893)
4. D.J. Boorstin, The Discoverers: A History of Mans Search to Know his World and Himself
(Vintage Books, New York, 1983)
5. R.P. Feynman, R.B. Leighton, M. Sands, The Feynman Lectures on Physics, vol. I (AddisonWesley, Reading, 1963)
6. Kurt Gdel, ber formal unentscheidbare stze der principia mathematica und verwandter
systeme, i. Monatshefte fr Mathematik und Physik 38, 173198 (1931)
7. E.T. Jaynes, Where do we stand on maximum entropy? in Papers on Probability, Statistics
and Statistical Physics, ed. by R.D. Rosencrantz, E.T. Jaynes (Kluwer Academic Publishers,
Dordrecht, 1989), pp. 211314
8. E.T. Jaynes, Probability Theory: The Logic of Science (Cambridge University Press, Cambridge, 1998)
9. S.A. Kauffman, Beyond reductionism: reinventing the sacred. Edge.org (2006)
10. J. Lehrer, Trials and errors: why science is failing us. Wired, December 2011
11. C. Sagan, The dragon in my garage, in The Demon-Haunted World: Science as a Candle in the
Dark (Ballantine, New York, 1996)
12. S. Sambursky, Physical Thought from the Presocratics to the Quantum Physicists: An Anthology
(Pica Press, New York, 1974)
13. Steering Committee on Science and Creationism, National Academy of Sciences. Science and
Creationism: A View from the National Academy of Sciences (The National Academies Press,
Washington, 1999)
14. V. Stodden, The Scientific Method in Practice: Reproducibility in the Computational Sciences
(MIT Sloan Research Papers, 2010), Working paper no. 477310
15. B. Swarup, The end of the quantum road? Interview with Caslav Brukner (2009). http://fqxi.
org/community/articles/display/114
16. Robert E. Ulanowicz, Ecology: The Ascendent Perspective (Columbia University Press, New
York, 1997)
Chapter 19
Is Life Fundamental?
Sara Imari Walker
One can best feel in dealing with living things how primitive
physics still is.
Albert Einstein
Although there has been remarkable progress in understanding some pieces of the
puzzle, the emergence of life is still a mystery, presenting what is arguably one of
the greatest unsolved questions in science. For the physicist or biologist, this may
seem a problem for chemistry, and that the difficulty is simply that we dont have the
know-how to engineer chemical networks quiet as complex as life, at least not yet.
However, current challenges and limitations in chemical synthesis and the design
of complex chemical networks may be only part of the story. The central challenge
is that we dont know whether life is just very complex chemistry,1 or if there is
something fundamentally distinct about living matter. Until this issue is resolved,
real progress in understanding how life emerges is likely to be limited.
Whats at stake here is not merely an issue of how chemical systems complexify;
the question of whether life is fully reducible to just the rules chemistry and physics
(albeit in a very complicated manner) or is perhaps something very different forces
us to assess precisely what it is that we mean by the very nature of the question of the
emergence of life [1]. Stated most acutely, if a fully reductionist account is sufficient,
1
This is not to imply that life is any less remarkable if a full account of biological organization
turns out to indeed reduce to nothing more than the underlying rules of chemistry and physics
subject to the appropriate boundary conditions and no additional principles are needed.
S.I. Walker (B)

School of Earth and Space Exploration and Beyond Center for Fundamental
Concepts in Science, Arizona State University,Tempe, AZ, USA
S.I. Walker
Blue Marble Space Institute of Science, Seattle, WA, USA
259
260
S.I. Walker
and life is completely describable as the nothing other than very complicated sets
of chemical reactions, what then can we say originated? Taken to the extreme, the
all life is just chemistry viewpoint advocates in a very real sense that life does
not exist and as such that there is no transition to be defined. While this may very
well be the case, when cast in these terms, even the avid reductionist might be
unwilling, or at least hesitant, to accept such an extreme viewpoint. At the very least,
although it is an open question whether this viewpoint is fundamentally correct,
it is counterproductive to think in such termswithout a well-defined distinction
between the two, there is no constructive mode of inquiry into understanding the
transition from nonliving to living matter. As much as (or perhaps more than) any
other area of science, the study of the emergence of life forces us to challenge our
basic physical assumptions that a fully reductionist account is adequate to explain
the nature of reality.
An illustrative example may be in order. It is widely appreciated that the known
laws of physics and chemistry do not necessitate that life should exist. Nor do they
appear to explain it [2]. Therefore in lieu of being able to start from scratch, and
reconstruct life from the rules of the underlying physics and chemistry, most are
happy to avert the issue nearly entirely. We do so by applying the Darwinian criterion
and assuming that if we can build a simple chemical system capable of Darwinian
evolution the rest will follow suit and the question of the origin of life will be solved
[3]. Accordingly, the problem of the origin of life has effectively been reduced to
solving the conceptually simpler problem of identifying the origin of Darwinian
evolution. Although this methodology has been successful in addressing specific
aspects of the puzzle, it is unsatisfactory in resolving the central issue at hand by
stealthily avoiding addressing when and how the physical transition from nonlife to
life occurs. Therefore, although few (barring the exception of our avid reductionist)
are likely to be willing to accept a simple molecular self-replicator as living, the
assumption goes that Darwinian evolution will invariably lead to something anyone
would agree is alive. The problem is that the Darwinian criteria is simply too
general, applying to any system (alive or not) capable of replication, selection, and
heritage (e.g. memes, software programs, multicellular life, non-enzymatic template
replicators, etc.). It therefore provides no means for distinguishing complex from
simple, let alone life from non-life. In the example above, the Darwinian paradigm
applies to both the precursor of life (i.e. a molecular self-replicator) and the living
system it is assumed to evolve into, yet most might be hesitant to identify the former
as living. It is easy to see why Darwin himself was trepidatious in applying his
theory to explain the emergence of life.2 If we are satisfied to stick with our current
picture decreeing that all life is chemistry with the caveat subject to Darwinian
evolution, we must be prepared to accept that we may never have a satisfactory
answer to the question of the origin of life and in fact that the question itself may not
be well-posed.
Darwin is famously quoted as stating, It is mere rubbish thinking, at present, of the origin of life;
one might as well think of the origin of matter [4].
19 Is Life Fundamental?
261
The central argument of this is essay is that we should not be satisfied with this
fully reductionist picture. If we are going to treat the origin of life as a solvable
scientific inquiry (which we certainly can and should), we must assume, at least
on phenomenological grounds, that life is nontrivially different from nonlife. The
challenge at hand, and I believe this is a challenge for the physicist, is therefore
to determine whatif anythingis truly distinctive about living matter. This is a
tall order. As Anderson put it in his essay More is Different, The ability to reduce
everything to simple fundamental laws does not imply the ability to start from those
laws and reconstruct the universe [5]. From this perspective, although an explanation
of the physics and chemistry underlying the components of living systems is fully
reducible to known physics, for all practical purposes we just cant work in the other
direction and expect to really nail the issue down. If we cant work from the bottomup, then we must work from the top-down by identifying the most distinctive features
of the organizational and logical architecture of known living systems, which set them
apart from their nonliving counterparts. We must therefore assume, right at the outset,
that the all life is chemistry picture is inadequate to explain the phenomenon of
life. We must ask, if life is not just complex chemistry, then what is life?
Despite the notorious difficulty in identifying precisely what it is that makes life
seem so unique and remarkable, there is a growing consensus that its informational
aspect is one key property, and perhaps the key property. If life is more than just
complex chemistry, its unique informational aspects may therefore be the crucial
indicator of this distinction. The remainder of this essay focuses on an illustrative
example of how treating the unique informational narrative of living systems as more
than just chemistry may open up new avenues for research in investigations of the
emergence of life. I conclude with a discussion of the potential implications of such
a phenomenological frameworkif successful in elucidating the emergence of life
as a well-defined transitionon our interpretation of life as a fundamental natural
phenomenon.
It from Bit from It

Wheeler is quite famously quoted as suggesting that all of reality derives its existence
from information, captured cleverly by his aphorism it from bit [6]. If Wheelers
aphorism applies anywhere in physics, it certainly applies to life, albeit in a very
different context than what Wheeler had originally intended. Over the past several
decades the concept of information has gained a prominent role in many areas of
biology. We routinely use terminology such as signaling, quorum sensing and
reading and writing genetic information, while genes are described as being
transcribed, translated, and edited, all implying that the informational narrative
is aptly applied in the biological realm. The manner in which information flows
through and between cells and sub-cellular structures is quiet unlike anything else
we observe in the natural world.
262
S.I. Walker
As we now learn it in school, the central dogma of molecular biology states that
information flows from DNA RNA protein. In reality the situation is much
more complicated than this simple picture suggests. The central dogma captures only
the bit-by-bit transfer of Shannon (sequential) information. However, biology seems
to employ a richer and more challenging concept of information than that tackled
by Shannon, to the point that it is hotly debated what is even meant by the term
biological information. Consider as an example DNA, which acts as a digital storage repository for the cell. The human genome, for instance, contains roughly 3.2 billion base pairs, corresponding to roughly 800 MB of stored data. Compare this to rare
Japanese plant Paris Japonica, with a genomic size of a whopping 150 billion base
pairs or 37.5 GB of dataone of the largest genomes known [7]. Paris Japonica therefore vastly outstrips humans in terms of its genomes Shannon information content.
Does this somehow imply that this slow-growing mountain flower is more complex
(i.e. processes more information) than a human? Of course the answer is no. Across
the tree of life, genome size does not appear to readily correlate with organismal
complexity. This is because the genome is only a small part of the story: DNA is not
a blueprint for an organism,3 but instead provides a database for transcribing RNA,
some (but by no means all) of which is then translated to make proteins.
The crucial point here is the action is not in the DNA, no information is actively
processed in the DNA itself [8]. A genome provides a (mostly) passive access on
demand database, which contributes biologically meaningful information by being
read-out to produce functional (non-coding) RNAs and proteins. The biologically
relevant information stored in DNA therefore has nothing to do with the chemical
structure of DNA (beyond the fact that it is a digital linear polymer). The genetic
material could just as easily be another variety of nucleic acid and accomplish the
same task [9]. What is important is the functionality of the expressed RNAs and
proteins. Functional information is a very strange beast, being dictated in part by the
global context rather than just the local physics [10]. For example, the functionality
of expressed RNA and proteins is context-dependent, and is meaningful only in the
larger biochemical network of a cell, including other expressed proteins, RNAs, the
spatial distribution of metabolites, etc. Sometimes very different biochemical structures (in terms of chemical composition, for example) will fill the same exact functional rolea phenomenon known as functional equivalence (familiar from cases of
convergent evolution) where sets of operations perform the same functional outcome
[11]. Only small subsets of all possible RNA and protein sequences are biologically
functional. A priori, it is not possible to determine which will be functional in a cell
based purely on local structure and sequence information alone (although some algorithms are becoming efficient at predicting structure, functionality is still determined
by insertion in a cell, or inferred by comparison to known structures). Biologically
functional information is therefore not an additional quality, like electric charge or
spin, painted onto matter and fixed for all time. It is only definable in a relational
sense, and thus must be defined only within a wider context.
3
Here a blueprint is defined as providing a one-to-one correspondence between the symbolic

representation and the actual object it describes.
263
One is left to conclude that the most important features of biological information,
such as functionality, are inherently nonlocal. Biological information is clearly not
solely in the DNA, or any other biochemical structure taken in isolation, and therefore must somehow be stored in the current state of the system (e.g. the level of gene
expression and the instantaneous biochemical interaction network). Moreover, molecular biologists are continuing to uncover a huge variety of regulatory RNAs and
proteins, which acting in concert with other cellular components, dictate the operating mode (e.g. phenotype) of a cell. Therefore, not only is the information specifying
functional roles distributed, but information control is also a widely distributed and
context-dependent feature of biological organization [12].
Superficially this may not seem to be anything particularly insightful or illuminating. One might argue that such distribution of information and control is an
inevitable consequence of the complexity of biochemical networks. However, on
closer inspection this state of affairs is really quiet remarkable for a physical system
and potentially hints at something fundamentally different about how living systems
process information that separates them from their nonliving counterparts. Cutting
straight to the point, in biology information appears to have causal efficacy [11, 13].
It is the information encoded in the current state that determines the dynamics and
hence the future state(s) and vice versa [14].
Consider a simplified example: the case of the genome and proteome systems,
where the current state of the systemi.e. the relative level of gene expression
depends on the composition of the proteome, environmental factors, etc. that in turn
regulate the switching on and off of individual genes. These then in turn dictate the
future state of the system. An important point is that these two subsystems cannot
function in isolation. Colloquially, this dynamic is often referred to as a chickenor-egg problem, where neither the genotype nor the phenotype can exist without
the other. Such a dynamic is well-known from the paradoxes of self-reference [15];
picture for example Eschers Drawing Hands where each of a pair of hands is drawing
the other with no possibility of separating the two: it is unclear which hand is the
cause and which the effect.
In biology, we cannot disentangle the genotype and phenotype because causation
is distributed within the state of the system as a whole (including the relations among
all of the subcomponents). Similar dynamics are at play throughout the informational
hierarchies of biological organization, from the epigenome [16], to quorum sensing
and inter-cellular signaling in biofilms [17], to the use of signaling and language to
determine social group behavior [18]. In all of these cases where the informational
narrative is utilized, we observe context (state) dependent causation, with the result
that the update rules change in a manner that is both a function of the current state
and the history of the organism [14]. Here casting the problem in the context of an
informational narrative is crucialthe foregoing discussion may be formalized by
stating that the algorithm describing the evolution of a biological system changes with
the information encoded in the current state and vice versa. Contrast this with more
traditional approaches to dynamics where the physical state of a system at time t1 is
mapped into the state at a later time t2 in accordance with a fixed dynamical law and
imposed boundary conditions. Thus, for example, Newtonian mechanics provides
264
S.I. Walker
the algorithm that maps the state of the solar system today onto its state tomorrow by
specifying a trajectory through phase space. The key distinction between this situation
and that observed in biology is that information doesnt push back and actively
influence the ensuing rules of dynamical evolution as it does in living systems. This
feature of dynamical laws changing with states as far as we know, seems to be
unique to biological organization and is a direct result of the peculiar nature of
biological information (although speculative examples from cosmology have also
been discussed, see e.g. [19]). It therefore serves as a contender for defining living
matter.
Wheelers dictum, as applied to the biological realm should therefore read more
as it from bit from it,4 where lower of levels of matter dictate the informational state
of a system which then in turn dictates its future evolution. In this picture, life is a
dynamical phenomenon that emerges when information gains causal efficacy over the
matter it is instantiated in [20]. A situation made possible by the separation of information from its physical representation (i.e. through functional equivalence, coded
intermediates, etc.). Thus, in biology the informational narrative is freed up to be
almost independent of the material one and we may sensibly discuss cell-sell signaling, or sense data flowing along nerves, without specific reference to the underlying
activity of electrons, protons, atoms or molecules. Of course all information requires
a material substrate, but the important point here is that life cannot be understood in
terms of the substrate alone. Thus it is meaningless to say that any single atom in a
strand of DNA is alive. Yet, it is meaningful to state that the organism as a whole is
living. Aliveness is an emergent global property.
Informational Efficacy and the Origin of Life

The liberation of the informational narrative from the material one potentially elicits
a well-defined physical transition (even if currently not well-understood), which may
be identifiable with the physical mechanism driving the emergence of life. In this
picture, the origin of life effectively mediates the transition whereby information a
high-level phenomenon gains causal efficacy over matter in a top-down manner5
[20]. In physics we are used to the idea of bottom-up causation, where all causation
stems from the most fundamental underlying layers of material reality. In contrast,
top-down-causation is characterized by a higher level in an organizational hierarchy
influencing a lower level by setting a context (for example, by changing some physical
constraints) by which the lower level actions take place. In such cases, causation can
also run downward in organizational hierarchies [21, 22]. Thus, top-down causation
4
Perhaps an even better dictum might be it from bit from it from bit ad infinitum to capture the
self-referential nature of dynamical laws changing with states.
5 In practice, top and bottom levels are typically not easily identified in hierarchical systems.
Conceptually one may view both top-down and bottom-up causal effects as inter-level phenomenon, occurring between neighboring levels in a hierarchy, a phenomenon referred to as levelentanglement by Davies (not to be confused with entanglement in quantum systems) [19].
265
opens up the possibility that high-level non-physical entities (i.e. information) may
have causal efficacy in their own right [19, 23].
There is a vast literature suggesting top-down causation as a unifying mechanistic
principle underlying emergence across the sciences, from quantum physics to computer science, to evolutionary biology, to physiology and the cognitive and social
sciences (see e.g. [22]). In some areas of science, such as physiology, the existence
of top-down causal effects is taken as self-evident and essential to making scientific
progress. For example, it is not even a subject of debate that information control
is widely distributed within living organisms (and thus that causation is also distributed). In other areas of science, such as chemistry and physics, which may be
more familiar to the reader, top-down causation is not nearly as widely accepted.
In particular, its role in chemistry is not well understood at all [24]. Poised at the
intersection of the domains of science where top-down causation is widely accepted
(biology) and where its role is not readily apparent (chemistry and physics) sits the
emergence of life, suggesting that some very interesting physics may be occurring
at this transition, and it may have everything to do with the appearance of genuinely
new high-level causes.
Adopting this picture as constructive scientific inquiry into the emergence of life,
an important question immediately presents itself: if a transition from bottom-up
causation only (e.g. at the level of chemistry), to top-down (intermingled with bottomup) causation may be identifiable with the emergence of life, what sets the origin of
life apart from other areas of science where the role of top-down causation is clearly
evident? As outlined by Ellis, there may in fact be several different mechanisms
for top-down causation, which come into play at different hierarchical scales in
nature [13]. In this regard, there may in fact be something unique to the emergence
of life, which stems from the unique informational narrative of living systems as
described in the previous section. Namely, biological systems (and other physical
systems derivative of the biosphere such as computers and societies) seem to be
unique in their implementation of top-down causation via information control [11,
13]. According to Auletta et al. who have rigorously defined this concept in the
biological realm Top-down causation by information control is the way a higher
level instance exercises control of lower level causal interactions through feedback
control loops, making use of functional equivalence classes of operations [11]. The
key distinction between the origin of life and other realms of science is therefore due
to the onset of distributed information control, enabling context-dependent causation,
where informationa high level and abstract entity, effectively becomes a cause. Cast
in the language of the previous section this is just another way of stating that the
origin of life might be associated with the onset of dynamical laws changing with
states [20].
In contrast to other quantities attempting to capture the role of information in
living systems, such as functional or semantic information, or even dynamical laws
changing with states (e.g. self-referential dynamics), causality is readily definable,
and in principle measureable (although often difficult in practice). This is a primary
reason why top-down causation is widely heralded as one of the most productive
formalisms for thinking about emergence [22]. This framework therefore potentially
266
S.I. Walker
enables a methodology for identifying a non-trivial distinction between life and

nonlife, delineated by a fundamental difference in how information is processed. For
the later, information is passive, whereas for the former information plays an active
role and is therefore causally efficacious. The catch is that one must be willing to
accept (at the very least on phenomenological grounds) the causal role of information
as a defining feature in the story of life right along side the substrate narrative of
the underlying chemistry. This forces new thinking in how life might have arisen
on lifeless planet, by shifting emphasis to the origins of information control, rather
than the onset of Darwinian evolution or the appearance of autocatalytic sets (that
lack control) for example, which do not rigorously define how/when life emerges. It
permits a more universal view of life, where the same underlying principles would
permit understanding of living systems instantiated in different substrates (either
artificial or in alternative chemistries). It may also encourage new thinking about the
emergence of the apparent arrow of time in the biosphere, trending in a direction
of increasing complexity with time: dynamical evolution where laws change with
states is likely to not be time-reversal invariant (although this remains to be rigorously
demonstrated). Once life emerges, we might therefore expect it to complexify and
diversify over time, particularly as information gains causal efficacy over increasingly
higher-levels of organization through major evolutionary innovations [25].
In practice, utilizing this framework as a productive paradigm for addressing the
emergence of life will likely be very difficult. We currently dont have any good measures this transition. Although there is a vast literature in top-down causation, the role
of a possible shift in informational efficacy (control) and thus causal structure as the
key transition mediating the emergence of life has been absent in nearly all discussions of lifes origins (see e.g. [20] for an exception relevant to this discussion). Part
of the challenge is that we do not have the proper tools yet. Walker et al. proposed one
possible measure, applying transfer entropy to study the flow of information from
local to global and from global to local scales in a lattice of coupled logistic maps
[25]. Nontrivial collective behavior was observed each time the dominant direction
of information flow shifted from bottom-up to top-down (meant to act a toy model
for the transition from independent replicators to collective reproducers characteristic of many major evolutionary transitions). However, this measure falls far short
of being satisfactory. In particular, it doesnt capture true emergence where the parts
do not exist without the whole (i.e. the cells in your body cannot exist outside of
the multicellular aggregate that is you). Furthermore, it does not capture the causal
relations among lower level entities and therefore is incapable of quantifying how
the informational state of a system influences these lower level causal relations. In
fact, it is not even a causal measure. In a very different context, a step in this direction
may be provided by Tononis measure of integrated information , which has been
proposed as a way to quantify consciousness by measuring causal architecture based
on network topology [26]. This measure effectively captures the information generated by the causal interactions of the sub-elements of a system beyond that which is
generated independently by its parts. It therefore provides a measure of distributed
information generated by the network as a whole due to its causal architecture. A
version of the theory whereby itself is treated as a dynamical variable that has
267
causal power in its own right might provide a way of quantifying the causal efficacy
of information in the context that has been discussed here. Additional formalisms will
need to also account for reliable encodings, where the same high-level phenomenon
is reliably produced. In biology we have the example of the genetic code, but are
far from decoding more distributed aspects of algorithmic information processing as
occurs in the epigenome or the connectome.
It is an open question what will ultimately provide a useful phenomenological
formalism for understanding the emergence of life. At the minimum the framework
presented here provides a non-trivial distinction between life and nonlife and thus
formulates the origin of life as a well-defined scientific problem, a key requirement for
rigorous inquiry into lifes emergence as discussed in the introduction. Life may be
identified as fundamentally distinct from just complex chemistry due to its causal
structure dictated by the causal efficacy of information. This immediately suggests
several lines of inquiry into the emergence of life (which may or may not be practical
at present). A top-down approach is to identify the causal architecture of known
biochemical networks by applying measures (such as , or other measures of causal
relationships [27]), for example by focusing on regulatory networks (information
control networks). A bottom-up approach is to determine how information control
emerges ab initio from chemical kinetics as well as how control evolves once this
information takeover has occurred. Some of these principles will likely be testable
in simple laboratory systems. A third line of inquiry could focus on the fundamental
aspects of the problem, such as state-dependent dynamical laws, or the reproducibility
of high-level outcomes via reliable encodings.
This is only a place to start, and it is entirely possible that additional and/or
other novel physical principles will be required to pin-down what really drove the
emergence of life. Whatever proper formalism emerges, we should not shy away
from treating life as a distinct and novel physical phenomenon when addressing its
origins. If this line of inquiry provides a productive framework for addressing the
origin of life, a question, which must eventually be asked, is: Is life fundamental?
For example, characterizing the emergence of life as a shift in causal architecture due
to information gaining causal efficacy over the matter it is instantiated would mark
the origin of life as a unique transition in the physical realm. Life would therefore be
interpreted as logically and organizationally distinct from other kinds of dynamical
systems,6 and thus be a novel state of matter emerging at higher levels of reality.
Our usual causal narrative, consisting of the bottom-up action of material entities
only, would therefore be only a subset of a broader class of phenomenaincluding
lifewhich admit immaterial causes in addition to material ones and which are
characterized by their causal architecture. We would therefore have to consider that
higher levels of reality admit the emergence of novel fundamental phenomena.
Note this does not preclude that there may exist a gradation of states which are almost life with
properties somewhere between completely passive and active informational dynamics, i.e. some
parts might exist autonomouslyan interesting question to consider in the context of astrobiology.
268
S.I. Walker
References
1. C.E. Cleland, C.F. Chyba, Defining life. Orig. Life Evol. Biosph. 32, 387393 (2002)
2. P.C.W. Davies, The Fifth Miracle: The Search for the Origin and Meaning of Life (Simon and
Schuster, New York, 1999)
3. G. Joyce, Bit by bit: the Darwinian basis of life. PLoS Biol. 418, 214221 (2012)
4. C. Darwin, Letter to J.D. Hooker, in The Correspondence of Charles Darwin 1863, vol. 11, ed.
by F. Burkhardt, S. Smith (1999), p. 278. (29 March 1863)
5. P.W. Anderson, More is different. Science 177, 393396 (1972)
6. J.A Wheeler, Sakharov revisited: It from bit, ed. by M. Manko In: Proceedings of the First
International A.D Sakharov Memorial Conference on Physics, Moscow, USSR (Nova Science
Publishers, Commack, New York, 1991)
7. J. Pellicer, M.F. Fay, I.J. Leitch, The largest Eukaryotic genome of them all? Bot. J. Linn. Soc.
164(1), 10 (2010)
8. D. Noble, Genes and causation. Philos. Trans. R. Soc. A 366, 30013015 (2008)
9. V.B. Pinhero et al., Synthetic genetic polymers capable of heredity and evolution. Science 336,
341344 (2012)
10. G. Auletta, Cognitive Biology: Dealing with Information from Bacteria to Minds (Oxford
University Press, Oxford, 2011)
11. G. Auletta, G.F.R. Ellis, L. Jaeger, Top-down causation by information control: from a philosophical problem to a scientific research programme. J. R. Soc. Interface 5, 11591172 (2008)
12. U. Alon, An Introduction to Systems Biology: Design Principles of Biological Circuits (CRC
Press Taylor & Francis, 2006)
13. G.F.R. Ellis, Top-down causation and emergence: some comments on mechanisms. J. R. Soc.
Interface 2(1), 126140 (2012)
14. N. Goldenfeld, C. Woese, Life is physics: evolution as a collective phenomenon far from
equilibrium. Annu. Rev. Condens. Matter Phys. 2(1), 375399 (2011)
15. D. Hofstadter, Godel, Escher, Bach: An Eternal Golden Braid (Basic Books Inc., New York,
1979)
16. P.C.W. Davies, The epigenome and top-down causation. J. R. Soc. Interface 2(1), 4248 (2012)
17. M.R. Parsek, E.P. Greenberg, Sociomicrobiology: the connections between quorum sensing
and biofilms. Trends Microbiol. 13, 2733 (2005)
18. J.C. Flack, F. de Waal, Context modulates signal meaning in primate communication. In: Proc.
Nat. Acad. of Sci. USA 104(5) 15811586 (2007)
19. P.C.W. Davies, The physics of downward causation, in The Re-emergence of Emergence, ed.
by P. Clayton, P.C.W. Davies (Oxford University Press, Oxford, 2006), pp. 3552
20. S.I. Walker, P.C.W. Davies, The algorithmic origins of life (2012). arXiv:1207.4803
21. D.T. Campbell, Levels of organization, downward causation, and the selection-theory approach
to evolutionary epistemoloty, ed. by G. Greenber, E. Tobach. Theories of the Evolution of
Knowing. T.C. Schneirla Conference Series, (1990), pp. 115
22. G.F.R. Ellis, D. Noble, T. OConnor, Top-down causation: an integrating theme within and
across the sciences? J. R. Soc. Interface 2, 13 (2011)
23. G.F.R. Ellis, On the nature of emergent reality, in The Re-emergence of Emergence, ed. by
P. Clayton, P.C.W. Davies (Oxford University Press, Oxford, 2006), pp. 79107
24. E.R. Scerri, Top-down causation regarding the chemistry-physics interface: a sceptical view.
Interface Focus 2, 2025 (2012)
25. S.I. Walker, L. Cisneros, P.C.W. Davies, Evolutionary transitions and top-down causation, in
Proceedings of Artificial Life XIII, pp. 283290 (2012)
26. G. Tononi, An information integration theory of consciousness. BMC Neurosci. 5, 42 (2004)
27. J. Pearl, Causality (Cambridge University Press, Cambridge, 2000)
Titles in this Series
Quantum Mechanics and Gravity

By Mendel Sachs
Quantum-Classical Correspondence
Dynamical Quantization and the Classical Limit By Josef Bolitschek
Knowledge and the World: Challenges Beyond the Science Wars
Ed. by M. Carrier, J. Roggenhofer, G. Kppers and P. Blanchard
Quantum-Classical Analogies
By Daniela Dragoman and Mircea Dragoman
LifeAs a Matter of Fat
The Emerging Science of Lipidomics
By Ole G. Mouritsen
Quo Vadis Quantum Mechanics?
Ed. by Avshalom C. Elitzur, Shahar Dolev and Nancy Kolenda
Information and Its Role in Nature
By Juan G. Roederer
Extreme Events in Nature and Society
Ed. by Sergio Albeverio, Volker Jentsch and Holger Kantz
The Thermodynamic Machinery of Life
By Michal Kurzynski
Weak Links
The Universal Key to the Stability of Networks and Complex Systems
By Csermely Peter
The Frontiers Collection, DOI 10.1007/978-3-319-13045-3
269
270
The Emerging Physics of Consciousness

Ed. by Jack A. Tuszynski
Quantum Mechanics at the Crossroads
New Perspectives from History, Philosophy and Physics
Ed. by James Evans and Alan S. Thorndike
Mind, Matter and the Implicate Order
By Paavo T.I. Pylkknen
Particle Metaphysics
A Critical Account of Subatomic Reality
By Brigitte Falkenburg
The Physical Basis of The Direction of Time
By H. Dieter Zeh
Asymmetry: The Foundation of Information
By Scott J. Muller
Decoherence and the Quantum-To-Classical Transition
By Maximilian A. Schlosshauer
The Nonlinear Universe
Chaos, Emergence, Life
By Alwyn C. Scott
Quantum Superposition
Counterintuitive Consequences of Coherence, Entanglement, and Interference
By Mark P. Silverman
Symmetry Rules
How Science and Nature Are Founded on Symmetry
By Joseph Rosen
Mind, Matter and Quantum Mechanics
By Henry P. Stapp
Entanglement, Information, and the Interpretation of Quantum Mechanics
By Gregg Jaeger
Relativity and the Nature of Spacetime
By Vesselin Petkov
The Biological Evolution of Religious Mind and Behavior

Ed. by Eckart Voland and Wulf Schiefenhvel
Homo NovusA Human Without Illusions
Ed. by Ulrich J. Frey, Charlotte Strmer and Kai P. Willfhr
Brain-Computer Interfaces
Revolutionizing Human-Computer Interaction
Ed. by Bernhard Graimann, Brendan Allison and Gert Pfurtscheller
Extreme States of Matter on Earth and in the Cosmos
By Vladimir E. Fortov
Searching for Extraterrestrial Intelligence
SETI Past, Present, and Future
Ed. by H. Paul Shuch
Essential Building Blocks of Human Nature
Ed. by Ulrich J. Frey, Charlotte Strmer and Kai P. Willfhr
Mindful Universe
Quantum Mechanics and the Participating Observer
By Henry P. Stapp
Principles of Evolution
From the Planck Epoch to Complex Multicellular Life
Ed. by Hildegard Meyer-Ortmanns and Stefan Thurner
The Second Law of Economics
Energy, Entropy, and the Origins of Wealth
By Reiner Kmmel
States of Consciousness
Experimental Insights into Meditation, Waking, Sleep and Dreams
Ed. by Dean Cvetkovic and Irena Cosic
Elegance and Enigma
The Quantum Interviews
Ed. by Maximilian Schlosshauer
Humans on Earth
From Origins to Possible Futures
By Filipe Duarte Santos
271
272
Evolution 2.0
Implications of Darwinism in Philosophy and the Social and Natural Sciences
Ed. by Martin Brinkworth and Friedel Weinert
Probability in Physics
Ed. by Yemima Ben-Menahem and Meir Hemmo
Chips 2020
A Guide to the Future of Nanoelectronics
Ed. by Bernd Hoefflinger
From the Web to the Grid and Beyond
Computing Paradigms Driven by High-Energy Physics
Ed. by Ren Brun, Federico Carminati and Giuliana Galli Carminati
Why More Is Different
Philosophical Issues in Condensed Matter Physics and Complex Systems
Ed. by Brigitte Falkenburg and Margaret Morrison
Appendix
List of Winners
First Prize
Robert Spekkens: The paradigm of kinematics and dynamics must yield to causal
structure1
Second Prizes
George Ellis: Recognising Top-Down Causation
Steve Weinstein: Patterns in the Fabric of Nature
Third Prizes
Julian Barbour: Reductionist Doubts
Giacomo DAriano: Quantum-informational Principles for Physics
Benjamin Dribus: On the Foundational Assumptions of Modern Physics
Sabine Hossenfelder: Gravity can be neither classical nor quantized
Ken Wharton: The Universe is not a Computer
Fourth Prizes
Giovanni Amelino-Camelia: Against spacetime
Michele Arzano: Weaving commutators: Beyond Fock space
1
From the Foundational Questions Institute website: http://www.fqxi.org/community/essay/

winners/2012.1.
The Frontiers Collection, DOI 10.1007/978-3-319-13045-3
273
274
Appendix: List of Winners
Torsten Asselmeyer-Maluga: A chicken-and-egg problem: Which came first, the

quantum state or spacetime?
Olaf Dreyer: Not on but of.
Ian Durham: Rethinking the scientific enterprise: In defense of reductionism
Sean Gryb & Flavio Mercati: Right about time?
Daryl Janzen: A Critical Look at the Standard Cosmological Picture
Israel Perez: The Preferred System of Reference Reloaded
Angelo Bassi, Tejinder Singh & Hendrik Ulbricht: Is quantum linear superposition
an exact principle of nature?
Sara Walker: Is Life Fundamental?

Questioning Basics of Physics

Uploaded by

Copyright:

Available Formats

Questioning Basics of Physics

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Questioning Basics of Physics

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the topics covered in the book?

What are some of the topics covered in the book?

T H E

THE FRONTIERS COLLECTION

THE FRONTIERS COLLECTION

More information about this series at http://www.springer.com/series/5342

Anthony Aguirre Brendan Foster

This book is a collaborative project between Springer and The Foundational

The Paradigm of Kinematics and Dynamics Must Yield

Recognising Top-Down Causation . . . . . . . . . . . . . . . . . . . . . . . .

On the Foundational Assumptions of Modern Physics . . . . . . . . .

The Preferred System of Reference Reloaded . . . . . . . . . . . . . . . .

Right About Time? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

A Critical Look at the Standard Cosmological Picture . . . . . . . . .

Patterns in the Fabric of Nature . . . . . . . . . . . . . . . . . . . . . . . . .

Is Quantum Linear Superposition an Exact Principle

Quantum-Informational Principles for Physics . . . . . . . . . . . . . .

The Universe Is Not a Computer . . . . . . . . . . . . . . . . . . . . . . . . .

A Chicken-and-Egg Problem: Which Came First,

Gravity Can Be Neither Classical Nor Quantized . . . . . . . . . . . . .

Weaving Commutators: Beyond Fock Space . . . . . . . . . . . . . . . .

Rethinking the Scientic Enterprise: In Defense

Titles in this Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appendix: List of Winners. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Our conceptions of Physical Reality can never be definitive; we

So which of our current underlying preconceptionstacit or explicitneed

The Paradigm of Kinematics and Dynamics

2 The Paradigm of Kinematics and Dynamics

Different Formulations of Classical Mechanics

On the Possibility of Violating Unitarity

2 The Paradigm of Kinematics and Dynamics

Freedom in the Choice of Kinematics

Kinematical Locality and Dynamical Locality

The assumption has also been called separability [19].

2 The Paradigm of Kinematics and Dynamics

It is possible to derive Bell inequalities starting from the assumption of kinematical

Summary of the Argument

From Kinematics and Dynamics to Causal Structure

2 The Paradigm of Kinematics and Dynamics

Fig. 2.1 The causal graph

measurement on A (respectively B) is denoted S (respectively T ) and the variable

Fig. 2.2 Causal graphs for

2 The Paradigm of Kinematics and Dynamics

characterize equivalence classes of causal models while allowing for restrictions on

Recognising Top-Down Causation

3 Recognising Top-Down Causation

Two Basic Issues

Hierarchy of Scales and Causation

A fuller description is given here: http://www.mth.uct.ac.za/~ellis/cos0.html.

3 Recognising Top-Down Causation

Table 3.1 The hierarchy of physical matter

Complex Structures: Digital Computers

The Software Hierarchy

3 Recognising Top-Down Causation

3 Recognising Top-Down Causation

you change font (Times New Roman to Helvetica)

Complex Structures: Life and the Brain

3 Recognising Top-Down Causation

where is the Christofel symbol and s a scalar parameter of motion, it follows